Splunk Enterprise version 8.2 is no longer supported as of September 30, 2023. See the Splunk Software Support Policy for details. For information about upgrading to a supported version, see How to upgrade Splunk Enterprise.
This documentation does not apply to the most recent version of Splunk® Enterprise.
For documentation on the most recent version, go to the latest release.
server.conf
The following are the spec and example files for server.conf.
server.conf.spec
# Version 8.2.12 #
OVERVIEW
# This file contains settings and values to configure server options # in server.conf. # # Each stanza controls different search commands settings. # # There is a server.conf file in the $SPLUNK_HOME/etc/system/default/ directory. # Never change or copy the configuration files in the default directory. # The files in the default directory must remain intact and in their original # location. # # To set custom configurations, create a new file with the name server.conf in # the $SPLUNK_HOME/etc/system/local/ directory. Then add the specific settings # that you want to customize to the local configuration file. # For examples, see server.conf.example. You must restart the Splunk instance # to enable configuration changes. # # To learn more about configuration files (including file precedence) see the # documentation located at # http://docs.splunk.com/Documentation/Splunk/latest/Admin/Aboutconfigurationfiles # #
GLOBAL SETTINGS
# Use the [default] stanza to define any global settings. # * You can also define global settings outside of any stanza at the top # of the file. # * Each configuration file should have at most one default stanza. # If you have multiple default stanzas, settings are combined. If you # have multiple definitions of the same settings, the last definition # in the file wins. # * If a setting is defined at both the global level and in a specific # stanza, the value in the specific stanza takes precedence.
General Server Configuration
[general]
serverName = <ASCII string>
* The name that identifies this Splunk software instance for features such as
distributed search.
* Cannot be an empty string.
* Can contain environment variables.
* After any environment variables are expanded, the server name
(if not an IPv6 address) can only contain letters, numbers, underscores,
dots, and dashes. The server name must start with a letter, number, or an
underscore.
* Default: <hostname>-<user_running_splunk>
hostnameOption = [ fullyqualifiedname | clustername | shortname ]
* The type of information to use to determine how splunkd sets the 'host' value for a Windows
Splunk platform instance when you specify an input stanza with 'host = $decideOnStartup'.
* Applies only to Windows hosts, and only for input stanzas that use the
"host = $decideOnStartup" setting and value.
* Valid values are "fullyqualifiedname", "clustername", and "shortname".
* The value returned for the 'host' field depends on Windows DNS, NETBIOS,
and what the name of the host is.
* 'fullyqualifiedname' uses Windows DNS to return the fully qualified host name as the value.
* 'clustername' also uses Windows DNS, but sets the value to the domain and machine name.
* 'shortname' returns the NETBIOS name of the machine.
* Cannot be an empty string.
* Default: shortname
sessionTimeout = <nonnegative integer>[s|m|h|d]
* The amount of time before a user session times out, expressed as a
search-like time range.
* Examples include "24h" (24 hours), "3d" (3 days),
"7200s" (7200 seconds, or two hours)
* Default: "1" (1 hour)
invalidateSessionTokensOnLogout = <boolean>
* A value of "true" means the SHC invalidates any tokens associated with a logged-out session
across all nodes in the cluster.
* This setting has an effect only if search head clustering and App Key Value store are enabled.
* Splunkd on each node tries to keep the logout information in sync with other nodes in the
cluster within the specified 'logoutCacheRefreshInterval'.
* Default: false
logoutCacheRefreshInterval = <nonnegative integer>[s|m|h|d]
* This setting controls how often splunkd on a given node updates its local cache from the
App Key Value store when 'invalidateSessionTokensOnLogout' is enabled.
* This setting has no effect when 'invalidateSessionTokensOnLogout' is disabled.
* In normal scenarios, maximum time for changes to propogate across the cluster can be upto
this interval, plus a few seconds; minimum can be a second or two.
* There is no guarantee that this sync will always happen within this time. If the system is
blocked because of load or other issues like network partition, the information may not be
propogated within the specified interval.
* Default: 30s
trustedIP = <IP address>
* All logins from specified IP addresses are trusted. This means a
password is no longer required.
* Only set this if you are using Single Sign-On (SSO).
allowRemoteLogin = always|never|requireSetPassword
* Controls remote management by restricting general login. Note that this
does not apply to trusted SSO logins from a trustedIP.
* When set to "always", all remote login attempts are allowed.
* When set to "never", only local logins to splunkd are allowed. Note that this
still allows remote management through Splunk Web if Splunk Web is on
the same server.
* If set to "requireSetPassword":
* In the free license, remote login is disabled.
* In the pro license, remote login is disabled for the "admin" user if
the default password of "admin" has not been changed.
* NOTE: As of version 7.1, Splunk software does not support the use of default
passwords. The "requireSetPassword" value is deprecated and might be removed
in the future.
* Default: requireSetPassword
tar_format = gnutar|ustar
* Sets the default TAR format.
* Default: gnutar
access_logging_for_phonehome = <boolean>
* Enables/disables logging to the splunkd_access.log file for client phonehomes.
* Default: true (logging enabled)
hangup_after_phonehome = <boolean>
* Controls whether or not the deployment server hangs up the connection
after the phonehome is done.
* By default, persistent HTTP 1.1 connections are used with the server to
handle phonehomes. This might show higher memory usage if you have a large
number of clients.
* If you have more than the maximum recommended concurrent TCP connection
deployment clients, persistent connections can not help with the reuse of
connections. Setting this attribute to true helps bring down memory usage.
* Default: false (persistent connections for phonehome)
pass4SymmKey = <password>
* Authenticates traffic between:
* License master and its license slaves.
* Members of a cluster.
* Deployment server (DS) and its deployment clients (DCs).
* When authenticating members of a cluster, clustering might override the
passphrase specified in the clustering stanza. A clustering searchhead
connecting to multiple managers might further override in the
[clustermanager:stanza1] stanza.
* When authenticating deployment servers and clients, by default, DS-DCs
passphrase authentication is disabled. To enable DS-DCs passphrase
authentication, you must also add the following line to the [broker:broker]
stanza in the restmap.conf file: requireAuthentication = true
* In all scenarios, every node involved must set the same passphrase in
the same stanzas. For example in the [general] stanza and/or
[clustering] stanza. Otherwise, the respective communication does not proceed:
- licensing and deployment in the case of the [general] stanza
- clustering in case of the [clustering] stanza)
* Unencrypted passwords must not begin with "$1$". This is used by
Splunk software to determine if the password is already encrypted.
pass4SymmKey_minLength = <integer>
* The minimum length, in characters, that a 'pass4SymmKey' should be for a particular stanza.
* When you start the Splunk platform, if the 'pass4SymmKey' is shorter in length than
what you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* If you use the CLI to modify 'pass4SymmKey' to a value that is shorter than what
you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* Default: 12
listenOnIPv6 = no|yes|only
* By default, splunkd listens for incoming connections (both REST and
TCP inputs) using IPv4 only.
* When you set this value to "yes", splunkd simultaneously listens for
connections on both IPv4 and IPv6.
* To disable IPv4 entirely, set listenOnIPv6 to "only". This causes splunkd
to exclusively accept connections over IPv6. You might need to change
the mgmtHostPort setting in the web.conf file. Use '[::1]' instead of
'127.0.0.1'.
* Any setting of SPLUNK_BINDIP in your environment or the
splunk-launch.conf file overrides the listenOnIPv6 value.
In this case splunkd listens on the exact address specified.
connectUsingIpVersion = auto|4-first|6-first|4-only|6-only
* When making outbound TCP connections for forwarding event data, making
distributed search requests, etc., this setting controls whether the
connections are made using IPv4 or IPv6.
* Connections to literal addresses are unaffected by this setting. For
example, if a forwarder is configured to connect to "10.1.2.3" the
connection is made over IPv4 regardless of this setting.
* "auto:"
* If listenOnIPv6 is set to "no", the Splunk server follows the
"4-only" behavior.
* If listenOnIPv6 is set to "yes", the Splunk server follows "6-first"
* If listenOnIPv6 is set to "only", the Splunk server follow
"6-only" behavior.
* "4-first:" If a host is available over both IPv4 and IPv6, then
the Splunk server connects over IPv4 first and falls back to IPv6 if the
connection fails.
* "6-first": splunkd tries IPv6 first and fallback to IPv4 on failure.
* "4-only": splunkd only attempts to make connections over IPv4.
* "6-only": splunkd only attempts to connect to the IPv6 address.
* Default: auto. This means that the Splunk server selects a reasonable
value based on the listenOnIPv6 setting.
guid = <globally unique identifier for this instance>
* This setting (as of version 5.0) belongs in the [general] stanza of
SPLUNK_HOME/etc/instance.cfg file. See the .spec file of instance.cfg for
more information.
useHTTPServerCompression = <boolean>
* Specifies whether the splunkd HTTP server should support gzip content
encoding. For more info on how content encoding works, see Section 14.3
of Request for Comments: 2616 (RFC2616) on the World Wide Web Consortium
(W3C) website.
* Default: true
defaultHTTPServerCompressionLevel = <integer>
* If the useHTTPServerCompression setting is enabled (it is enabled
by default), this setting controls the compression level that the
Splunk server attempts to use.
* This number must be between 1 and 9.
* Higher numbers produce smaller compressed results but require more CPU
usage.
* Default: 6 (This is appropriate for most environments)
skipHTTPCompressionAcl = <network_acl>
* Lists a set of networks or addresses to skip data compression.
These are addresses that are considered so close that network speed is
never an issue, so any CPU time spent compressing a response is wasteful.
* Note that the server might still respond with compressed data if it
already has a compressed version of the data available.
* These rules are separated by commas or spaces.
* Each rule can be in the following forms:
1. A single IPv4 or IPv6 address, for example: "10.1.2.3", "fe80::4a3"
2. A CIDR block of addresses, for example: "10/8", "fe80:1234/32"
3. A DNS name, possibly with a '*' used as a wildcard, for example:
"myhost.example.com", "*.splunk.com")
4. A single '*' which matches anything
* Entries can also be prefixed with '!' to negate their meaning.
* Default: localhost addresses
legacyCiphers = decryptOnly|disabled
* This setting controls how Splunk software handles support for
legacy encryption ciphers.
* If set to "decryptOnly", Splunk software supports decryption of
configurations that have been encrypted with legacy ciphers.
It encrypts all new configurations with newer and stronger cyphers.
* If set to "disabled", Splunk software neither encrypts nor decrypts
configurations that have been encrypted with legacy ciphers.
* Default: decryptOnly
site = <site-id>
* Specifies the site that this Splunk instance belongs to when multisite is
enabled.
* Valid values for site-id include site0 to site63
* The special value "site0" can be set only on search heads or on forwarders
that are participating in indexer discovery.
* For a search head, "site0" disables search affinity.
* For a forwarder participating in indexer discovery, "site0" causes the
forwarder to send data to all peer nodes across all sites.
useHTTPClientCompression = true|false|on-http|on-https
* Specifies whether gzip compression should be supported when splunkd acts
as a client (including distributed searches). Note: For the content to
be compressed, the HTTP server that the client is connecting to should
also support compression.
* If the connection is being made over https and
"useClientSSLCompression=true", then setting "useHTTPClientCompression=true"
results in double compression work without much compression gain. To
mitigate this, set this value to "on-http" (or to "true", and
useClientSSLCompression to "false").
* Default: true
embedSecret = <string>
* When using report embedding, normally the generated URLs can only
be used on the search head that they were generated on.
* If "embedSecret" is set, then the token in the URL is encrypted
with this key. Then other search heads with the exact same setting
can also use the same URL.
* This is needed if you want to use report embedding across multiple
nodes on a search head pool.
parallelIngestionPipelines = <integer>
* The number of discrete data ingestion pipeline sets to create for this
instance.
* A pipeline set handles the processing of data, from receiving streams
of events through event processing and writing the events to disk.
* An indexer that operates multiple pipeline sets can achieve improved
performance with data parsing and disk writing, at the cost of additional
CPU cores.
* For most installations, the default setting of "1" is optimal.
* Use caution when changing this setting. Increasing the CPU usage for data
ingestion reduces available CPU cores for other tasks like searching.
* NOTE: Enabling multiple ingestion pipelines can change the behavior of some
settings in other configuration files. Each ingestion pipeline enforces
the limits of the following settings independently:
1. maxKBps (in the limits.conf file)
2. max_fd (in the limits.conf file)
3. maxHotBuckets (in the indexes.conf file)
4. maxHotSpanSecs (in the indexes.conf file)
* Default: 1
pipelineSetSelectionPolicy = <round_robin | weighted_random>
* Specifies the pipeline set selection policy to use while selecting pipeline
sets for new inputs.
* If set to round_robin, the incoming inputs are assigned to pipeline sets in a
round robin fashion.
* If set to weighted_random, the incoming inputs are assigned to pipeline sets
using a weighted random scheme designed to even out the CPU usage of each
pipeline set.
* NOTE: This setting only takes effect when parallelIngestionPipelines is
greater than 1.
* Default: round_robin
pipelineSetWeightsUpdatePeriod = <number>
* The interval, in seconds, when pipeline set weights are recalculated for the
weighted_random pipeline set selection policy.
* Reducing this interval causes pipeline set weights to be re-evaluated more
frequently, thereby enabling the system to react more quickly to changes in
dutycycle estimation.
* Increasing this interval causes pipeline set weights to be re-evaluated less
frequently, thereby reducing the likelihood of the system responding to
bursty events.
* Default: 30
pipelineSetNumTrackingPeriods = <number>
* The number of look-back periods, of interval pipelineSetWeightsUpdatePeriod,
that are used to keep track of incoming ingestion requests for pipeline sets.
* This information is used as a heuristic to calculate the pipeline set weights
at every expiry of pipelineSetWeightsUpdatePeriod.
* Default: 5
pipelineSetChannelSetCacheSize = <integer>
* Maximum number of inactive channels to be stored in the per-pipeline set
cache to reduce load in the configuration management system.
* Currently only affects ingestion via the HTTP Event Collector.
* Increasing this setting should reduce the number of created channels
reported in metrics.log under the 'channel_cache' group. If neither that
group nor the 'created' field exists in metrics.log, increasing this
value has no effect.
* Default: 12
instanceType = <string>
* Should not be modified by users.
* Informs components (such as the Splunk Web Manager section) which
environment the Splunk server is running in, to allow for more
customized behaviors.
* Default: download
requireBootPassphrase = <boolean>
* Prompt the user for a boot passphrase when starting splunkd.
* Splunkd uses this passphrase to grant itself access to platform-provided
secret storage facilities, like the GNOME keyring.
* For more information about secret storage, see the [secrets] stanza in
$SPLUNK_HOME/etc/system/README/authentication.conf.spec.
* Default (if Common Criteria mode is enabled): true
* Default (if Common Criteria mode is disabled): false
numThreadsForIndexInitExecutor = <positive integer>
* Number of threads that can be used by the index init thread pool.
* Maximum accepted value for this setting is 32.
* Default: 16
remoteStorageRecreateIndexesInStandalone = <boolean>
* Controls re-creation of remote storage enabled indexes in standalone mode.
* Default: true
cleanRemoteStorageByDefault = <boolean>
* Allows 'splunk clean eventdata' to clean the remote indexes when set to true.
* Default: false
recreate_index_fetch_bucket_batch_size = <positive_integer>
* Controls the maximum number of bucket IDs to fetch from remote storage
as part of a single transaction for a remote storage enabled index.
* Only valid for standalone mode.
* Default: 500
recreate_bucket_fetch_manifest_batch_size = <positive_integer>
* Controls the maximum number of bucket manifests to fetch in parallel
from remote storage.
* Only valid for standalone mode.
* Default: 100
splunkd_stop_timeout = <positive_integer>
* The maximum time, in seconds, that splunkd waits for a graceful shutdown to
complete before splunkd forces a stop.
* Default: 360 (6 minutes)
decommission_search_jobs_wait_secs = <unsigned integer>
* The maximum time, in seconds, that splunkd waits for running searches to complete
during a shutdown_decommission_search.
* To trigger this type of shudown, post to 'services/server/control/shutdown_decommission_search'
* If set to 0, splunkd does not wait, and all searches in progress will fail.
* If this search head is a member of a search head cluster, use 'decommission_search_jobs_wait_secs'
in the [shclustering] stanza instead.
* NOTE: If this search head is a node of an indexer cluster, use 'decommission_search_jobs_wait_secs'
in the [clustering] stanza instead.
* Default: 0
python.version = {python2|python3|force_python3}
* For Python scripts only, sets the default Python version to use.
* Can be overridden by other 'python.version' values elsewhere, with the
following exception:
* If you set to "force_python3", the system always uses Python 3, and ignores
'python.version' values that you set elsewhere.
* Default: python3
roll_and_wait_for_uploads_at_shutdown_secs = <non-negative number>
* Currently not supported. This setting is related to a feature that is
still under development.
* Default: 0 (disabled)
preShutdownCleanup = <bool>
* Currently not supported. This setting is related to a feature that is
still under development.
* Specifies if indexer waits to complete any indexing activities before continuing with shutdown.
* Default: true
reset_manifests_on_startup = <boolean>
* Whether or not the Splunk platform instance regenerates size retention
information for index bucket summaries that have been stored in the
manifest.csv files.
* Configuring this setting lets the platform instance have the most
up-to-date size retention information immediately after startup.
* When set to true, the size retention information for summaries stored
in the manifest.csv files are removed and regenerated during startup.
* When set to false, manifest.csv files are not reset during startup.
* Default: true
percent_manifests_to_reset = <percentage>
* In order to minimize the cost of resetting all manifest.csv files at once
the manifest.csv files are separated in groups that are processed separately.
* This percentage defines how many manifest.csv files each group will reset.
* For example, a setting of 20 means each group resets 20% of all manifests
resulting in 5 groups with 20% each.
* The minimum of one manifest.csv file will be processed per group.
* Legal values are between 0% and 100%.
* Default: 10%
regex_cache_hiwater = <integer>
* A threshold for the number of entries in the regex cache. If the regex cache
grows larger than this, splunkd server will purge some of the older entries.
* When set to a negative value, no purge occurs, no matter how large
the cache.
* Default: 2500
Configuration Change Audit
[config_change_audit] disabled = <boolean> * Whether or not splunkd writes configuration changes to the configuration change log at $SPLUNK_HOME/var/log/splunk/configuration_change.log. * If set to "false", configuration changes are captured in $SPLUNK_HOME/var/log/splunk/configuration_change.log. * If set to "true", configuration changes are not captured in $SPLUNK_HOME/var/log/splunk/configuration_change.log. * Default: true mode = [auto|track-only] * Set to "auto" or "track-only" to get log of .conf file changes under $SPLUNK_HOME/etc/system, $SPLUNK_HOME/etc/apps, $SPLUNK_HOME/etc/users, $SPLUNK_HOME/etc/slave-apps or changes to $SPLUNK_HOME/etc/instance.cfg. * The values "auto" and "track-only" are identical in their effects. Set mode to "auto" to auto-enroll this deployment into all the latest features. * CAUTION: This setting is experimental and is related to a feature that is still under development. Using the setting might increase resource usage. * Default: auto
Deployment Configuration details
[deployment]
pass4SymmKey = <passphrase string>
* Authenticates traffic between the deployment server (DS) and its
deployment clients (DCs).
* By default, DS-DCs passphrase authentication key is disabled. To enable
DS-DCs passphrase authentication, you must *also* add the following
line to the [broker:broker] stanza in the restmap.conf file:
requireAuthentication = true
* If the key is not set in the [deployment] stanza, the key is looked
for in the [general] stanza.
* NOTE: Unencrypted passwords must not begin with "$1$", because this is
used by Splunk software to determine if the password is already
encrypted.
pass4SymmKey_minLength = <integer>
* The minimum length, in characters, that a 'pass4SymmKey' should be for a particular stanza.
* When you start the Splunk platform, if the 'pass4SymmKey' is shorter in length than
what you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* If you use the CLI to modify 'pass4SymmKey' to a value that is shorter than what
you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* Default: 12
SSL Configuration details
[sslConfig]
* Set SSL for communications on Splunk back-end under this stanza name.
* NOTE: To set SSL (for example HTTPS) for Splunk Web and the browser,
use the web.conf file.
* Follow this stanza name with any number of the following attribute/value
pairs.
* If you do not specify an entry for each attribute, the default value
is used.
enableSplunkdSSL = <boolean>
* Enables/disables SSL on the splunkd management port (8089) and KV store
port (8191).
* NOTE: Running splunkd without SSL is not recommended.
* Distributed search often performs better with SSL enabled.
* Default: true
useClientSSLCompression = <boolean>
* Turns on HTTP client compression.
* Server-side compression is turned on by default. Setting this on the
client-side enables compression between server and client.
* Enabling this potentially gives you much faster distributed searches
across multiple Splunk instances.
* CAUTION: There are known performance issues due to SSL compression.
Confirm that 'conf_deploy_precompress_bundles',
'precompress_cluster_bundle', 'precompress_artifacts',
'preCompressKnowledgeBundlesClassicMode', 'preCompressKnowledgeBundlesCascadeMode',
and 'useHTTPClientCompression' are set to "false" before setting
'useClientSSLCompression' to "true" to avoid double compression.
* Default: false
useSplunkdClientSSLCompression = <boolean>
* Controls whether SSL compression is used when splunkd is acting as
an HTTP client, usually during certificate exchange, bundle replication,
remote calls, etc.
* This setting is effective if, and only if, useClientSSLCompression
is set to "true".
* NOTE: splunkd is not involved in data transfer in distributed search, the
search in a separate process is.
* Default: true
sslVersions = <versions_list>
* Comma-separated list of SSL versions to support for incoming connections.
* The versions available are "ssl3", "tls1.0", "tls1.1", and "tls1.2".
* The special version "*" selects all supported versions.
The version "tls" selects all versions tls1.0 or newer.
* If a version is prefixed with "-" it is removed from the list.
* SSLv2 is always disabled; "-ssl2" is accepted in the version
list but does nothing.
* When configured in FIPS mode, "ssl3" is always disabled regardless
of this configuration.
* Default: The default can vary (see the 'sslVersions' setting in
the $SPLUNK_HOME/etc/system/default/server.conf file for the
current default)
sslVersionsForClient = <versions_list>
* Comma-separated list of SSL versions to support for outgoing HTTP connections
from splunkd. This includes distributed search, deployment client, etc.
* This is usually less critical, since SSL/TLS always picks the highest
version both sides support. However, you can use this setting to prohibit
making connections to remote servers that only support older protocols.
* The syntax is the same as the 'sslVersions' setting above.
* NOTE: For forwarder connections, there is a separate 'sslVersions'
setting in the outputs.conf file. For connections to SAML servers, there
is a separate 'sslVersions' setting in the authentication.conf file.
* Default: The default can vary (see the 'sslVersionsForClient' setting in
the $SPLUNK_HOME/etc/system/default/server.conf file for the
current default)
supportSSLV3Only = <boolean>
* DEPRECATED. SSLv2 is disabled. The exact set of SSL versions
allowed is configurable using the 'sslVersions' setting above.
sslVerifyServerCert = <boolean>
* This setting is used by distributed search and distributed
deployment clients.
* For distributed search: Used when making a search request
to another server in the search cluster.
* For distributed deployment clients: Used when polling a
deployment server.
* If set to true, make sure that the connected server is
authenticated. Both the common name and the alternate name
of the server are checked for a match if they are specified
in this configuration file. A certificate is considered
verified if either is matched.
* Default: false
sslCommonNameToCheck = <commonName1>, <commonName2>, ...
* If set, and 'sslVerifyServerCert' is set to "true",
splunkd limits most outbound HTTPS connections to hosts which
use a certificate with one of the listed common names.
* The most important scenario is distributed search.
* Optional.
* No default (no common name checking.)
sslCommonNameList = <commonName1>, <commonName2>, ...
* DEPRECATED. Use the 'sslCommonNameToCheck' setting instead.
sslAltNameToCheck = <alternateName1>, <alternateName2>, ...
* If this value is set, and 'sslVerifyServerCert' is set to true,
splunkd also verifies certificates which have a so-called
"Subject Alternate Name" that matches any of the alternate
names in this list.
* Subject Alternate Names are effectively extended descriptive
fields in SSL certificates beyond the commonName. A common
practice for HTTPS certificates is to use these values to
store additional valid hostnames or domains where the
certificate should be considered valid.
* Accepts a comma-separated list of Subject Alternate Names to
consider as valid.
* Items in this list are never validated against the SSL Common Name.
* Optional.
* No default (no alternate name checking.)
requireClientCert = <boolean>
* Requires that any HTTPS client that connects to a splunkd
internal HTTPS server has a certificate that was signed by a
CA (Certificate Authority) specified by the 'sslRootCAPath' setting.
* Used by distributed search: Splunk indexing instances must be
authenticated to connect to another splunk indexing instance.
* Used by distributed deployment: The deployment server requires that
deployment clients are authenticated before allowing them to poll
for new configurations/applications.
* If set to "true", a client can connect ONLY if a certificate
created by our certificate authority was used on that client.
* Default: false
cipherSuite = <cipher suite string>
* If set, Splunk uses the specified cipher string for the HTTP server.
* If not set, Splunk uses the default cipher string provided by OpenSSL.
This is used to ensure that the server does not accept connections using
weak encryption protocols.
* Must specify 'dhFile' to enable any Diffie-Hellman ciphers.
* Default: The default can vary (See the 'cipherSuite' setting in
the $SPLUNK_HOME/etc/system/default/server.conf file for the
current default)
ecdhCurveName = <string>
* DEPRECATED.
* Use the 'ecdhCurves' setting instead.
* This setting specifies the Elliptic Curve Diffie-Hellman (ECDH) curve to
use for ECDH key negotiation.
* Splunk only supports named curves that have been specified by their
SHORT name.
* The list of valid named curves by their short and long names
can be obtained by running this CLI command:
$SPLUNK_HOME/bin/splunk cmd openssl ecparam -list_curves
* Default: empty string.
ecdhCurves = <comma-separated list>
* A list of ECDH curves to use for ECDH key negotiation.
* The curves should be specified in the order of preference.
* The client sends these curves as a part of an SSL Client Hello.
* The server supports only the curves specified in the list.
* Splunk software only supports named curves that have been specified
by their SHORT names.
* The list of valid named curves by their short and long names can be obtained
by running this CLI command:
$SPLUNK_HOME/bin/splunk cmd openssl ecparam -list_curves
* Example setting: "ecdhCurves = prime256v1,secp384r1,secp521r1"
* Default: The default can vary (See the 'ecdhCurves' setting in
the $SPLUNK_HOME/etc/system/default/server.conf file for the
current default)
serverCert = <path>
* The full path to the PEM (Privacy-Enhanced Mail) format server
certificate file.
* Certificates are auto-generated by splunkd on starting Splunk Enterprise.
* You can replace the default certificate with your own PEM
format file.
* Default: $SPLUNK_HOME/etc/auth/server.pem
sslKeysfile = <filename>
* DEPRECATED. Use the 'serverCert' setting instead.
* This file is in the directory specified by the 'caPath' setting
(see below).
* Default: server.pem
sslPassword = <password>
* Server certificate password.
* Default: password
sslKeysfilePassword = <password>
* DEPRECATED. Use the 'sslPassword' setting instead.
sslRootCAPath = <path>
* Full path to the root CA (Certificate Authority) certificate store
on the operating system.
* The <path> must refer to a PEM (Privacy-Enhanced Mail) format
file containing one or more root CA certificates concatenated
together.
* Required for Common Criteria.
* This setting is valid on Windows machines only if you have not set
'sslRootCAPathHonoredOnWindows' to "false".
* No default.
sslRootCAPathHonoredOnWindows = <boolean>
* DEPRECATED.
* Whether or not the Splunk instance respects the 'sslRootCAPath' setting on
Windows machines.
* If you set this setting to "false", then the instance does not respect the
'sslRootCAPath' setting on Windows machines.
* This setting is valid only on Windows, and only if you have set
'sslRootCAPath'.
* When the 'sslRootCAPath' setting is respected, the instance expects to find
a valid PEM file with valid root certificates that are referenced by that
path. If a valid file is not present, SSL communication fails.
* Default: true
caCertFile = <filename>
* DEPRECATED. Use the 'sslRootCAPath' setting instead.
* Used only if 'sslRootCAPath' is not set.
* File name (relative to 'caPath') of the CA (Certificate Authority)
certificate PEM format file containing one or more certificates
concatenated together.
* Default: cacert.pem
dhFile = <path>
* PEM (Privacy-Enhanced Mail) format Diffie-Hellman(DH) parameter file name.
* DH group size should be no less than 2048bits.
* This file is required in order to enable any Diffie-Hellman ciphers.
* No default.
caPath = <path>
* DEPRECATED. Use absolute paths for all certificate files.
* If certificate files given by other settings in this stanza are not absolute
paths, then they are relative to this path.
* Default: $SPLUNK_HOME/etc/auth
certCreateScript = <script name>
* Creation script for generating certificates on startup of Splunk Enterprise.
sendStrictTransportSecurityHeader = <boolean>
* If set to "true", the REST interface sends a "Strict-Transport-Security"
header with all responses to requests made over SSL.
* This can help avoid a client being tricked later by a
Man-In-The-Middle attack to accept a non-SSL request.
However, this requires a commitment that no non-SSL web hosts
ever run on this hostname on any port. For
example, if Splunk Web is in default non-SSL mode this can break the
ability of a browser to connect to it.
* NOTE: Enable with caution.
* Default: false
allowSslCompression = <boolean>
* If set to "true", the server allows clients to negotiate
SSL-layer data compression.
* KV Store also observes this setting.
* If set to "false", KV Store disables TLS compression.
* Default: true
allowSslRenegotiation = <boolean>
* In the SSL protocol, a client may request renegotiation of the
connection settings from time to time.
* If set to "false", causes the server to reject all renegotiation
attempts, breaking the connection. This limits the amount of CPU a
single TCP connection can use, but it can cause connectivity problems
especially for long-lived connections.
* Default: true
sslClientSessionPath = <path>
* Path where all client sessions are stored for session re-use.
* Used if 'useSslClientSessionCache' is set to "true".
* No default.
useSslClientSessionCache = <boolean>
* Specifies whether to re-use client session.
* When set to "true", client sessions are stored in memory for
session re-use. This reduces handshake time, latency and
computation time to improve SSL performance.
* When set to "false", each SSL connection performs a full
SSL handshake.
* Default: false
sslServerSessionTimeout = <integer>
* Timeout, in seconds, for newly created session.
* If set to "0", disables Server side session cache.
* The openssl default is 300 seconds.
* Default: 300 (5 minutes)
sslServerHandshakeTimeout = <integer>
* The timeout, in seconds, for an SSL handshake to complete between an
SSL client and the Splunk SSL server.
* If the SSL server does not receive a "Client Hello" from the SSL client within
'sslServerHandshakeTimeout' seconds, the server terminates
the connection.
* Default: 60
Splunkd http proxy configuration
[proxyConfig]
http_proxy = <string>
* If set, splunkd sends all HTTP requests through the proxy server
that you specify.
* No default.
https_proxy = <string>
* If set, splunkd sends all HTTPS requests through the proxy server
that you specify.
* If not set, splunkd uses the 'http_proxy' setting instead.
* No default.
proxy_rules = <string>
* One or more host names or IP addresses for which splunkd should route
HTTPS requests only through the proxy server.
* If set, splunkd uses the proxy server only for endpoints that match the
hosts or IP addresses in this value.
* Splunkd does not route requests to either the localhost or loopback addresses
through the proxy server.
* Separate multiple entries with commas.
* This setting accepts the following values:
* '*' (asterisk): Proxy all requests. This is the only wildcard, and it can
be used only by itself.
* <IPv4 or IPv6 address>: Route the request through the proxy if the
request is intended for that address.
* <hostname>/<domain name>: Route the request through the proxy if
the request is intended for that host name or domain name.
* Examples:
* proxy_rules = "wimpy": This matches the host name "wimpy".
* proxy_rules = "splunk.com": Matches all host names in the splunk.com
domain (such as apps.splunk.com, www.splunk.com, etc.)
* Default: *
no_proxy = <string>
* One or more host names or IP addresses for which splunkd should
explicitly bypass the proxy server for HTTPS requests.
* If set, splunkd does not route requests to matching host names and
IP addresses through the proxy server.
* This setting overrides the 'proxy_rules' setting. If a host name or IP
address is in both settings, splunkd does not route requests for that
host name or IP address through the proxy server.
* Splunkd does not route requests to either the localhost or loopback addresses
through the proxy server.
addresses.
* Separate multiple entries with commas.
* This setting accepts the following values:
* '*' (asterisk): Proxy all requests. This is the only wildcard, and it can
be used only by itself.
* <IPv4 or IPv6 address>: Route the request through the proxy if the
request is intended for that address.
* <hostname>/<domain name>: Route the request through the proxy if
the request is intended for that host name or domain name.
* Examples:
* no_proxy = "wimpy": This matches the host name "wimpy".
* no_proxy = "splunk.com": Matches all host names in the splunk.com
domain (such as apps.splunk.com, www.splunk.com, etc.)
* Default: localhost, 127.0.0.1, ::1
Splunkd HTTP server configuration
[httpServer]
* Set stand-alone HTTP settings for splunkd under this stanza name.
* Follow this stanza name with any number of the following attribute/value
pairs.
* If you do not specify an entry for each attribute, splunkd uses the default
value.
atomFeedStylesheet = <string>
* Defines the stylesheet relative URL to apply to default Atom feeds.
* Set to 'none' to stop writing out xsl-stylesheet directive.
* Default: /static/atom.xsl
max-age = <nonnegative integer>
* Set the maximum time, in seconds, to cache a static asset served off of
the '/static' directory.
* This value is passed along in the 'Cache-Control' HTTP header.
* Default: 3600 (60 minutes)
follow-symlinks = <boolean>
* Specifies whether the static file handler (serving the '/static'
directory) follows filesystem symlinks when serving files.
* Default: false
disableDefaultPort = <boolean>
* If set to "true", turns off listening on the splunkd management port,
which is 8089 by default.
* NOTE: Changing this setting is not recommended.
* This is the general communication path to splunkd. If it is disabled,
there is no way to communicate with a running splunk instance.
* This means many command line splunk invocations cannot function,
Splunk Web cannot function, the REST interface cannot function, etc.
* If you choose to disable the port anyway, understand that you are
selecting reduced Splunk functionality.
* Default: false
acceptFrom = <network_acl> ...
* Lists a set of networks or addresses from which to accept connections.
* Separate multiple rules with commas or spaces.
* Each rule can be in one of the following formats:
1. A single IPv4 or IPv6 address (examples: "10.1.2.3", "fe80::4a3")
2. A Classless Inter-Domain Routing (CIDR) block of addresses
(examples: "10/8", "192.168.1/24", "fe80:1234/32")
3. A DNS name, possibly with a "*" used as a wildcard
(examples: "myhost.example.com", "*.splunk.com")
4. "*", which matches anything
* You can also prefix an entry with '!' to cause the rule to reject the
connection. The input applies rules in order, and uses the first one that
matches.
For example, "!10.1/16, *" allows connections from everywhere except
the 10.1.*.* network.
* Default: "*" (accept from anywhere)
streamInWriteTimeout = <positive number>
* The timeout, in seconds, for uploading data to the http server.
* When uploading data to http server, if the http server is unable
to write data to the receiver for the specified value, the operation
aborts.
* Default: 5
max_content_length = <integer>
* Maximum content length, in bytes.
* HTTP requests over the size specified are rejected.
* This setting exists to avoid allocating an unreasonable amount
of memory from web requests.
* In environments where indexers have enormous amounts of RAM, this
number can be reasonably increased to handle large quantities of
bundle data.
* Default: 2147483648 (2GB)
maxSockets = <integer>
* The number of simultaneous HTTP connections that Splunk Enterprise accepts
simultaneously. You can limit this number to constrain resource usage.
* If set to "0", Splunk Enterprise automatically sets maxSockets to
one third of the maximum allowable open files on the host.
* If this number is less than 50, it is set to 50.
* If this number is greater than 400000, it is set to 400000.
* If set to a negative number, no limit is enforced.
* Default: 0
maxThreads = <integer>
* The number of threads that can be used by active HTTP transactions.
You can limit this number to constrain resource usage.
* If set to 0, Splunk Enterprise automatically sets the limit to
one third of the maximum allowable threads on the host.
* If this number is less than 20, it is set to 20. If this number is
greater than 150000, it is set to 150000.
* If maxSockets is not negative and maxThreads is greater than maxSockets, then
Splunk Enterprise sets maxThreads to be equal to maxSockets.
* If set to a negative number, no limit is enforced.
* Default: 0
keepAliveIdleTimeout = <integer>
* How long, in seconds, that the Splunkd HTTP server allows a keep-alive
connection to remain idle before forcibly disconnecting it.
* If this number is less than 7200, it is set to 7200.
* Default: 7200 (120 minutes)
busyKeepAliveIdleTimeout = <integer>
* How long, in seconds, that the Splunkd HTTP server allows a keep-alive
connection to remain idle while in a busy state before forcibly
disconnecting it.
* Use caution when configuring this setting as a value that is too large
can result in file descriptor exhaustion due to idling connections.
* If this number is less than 12, it is set to 12.
* Default: 12
forceHttp10 = auto|never|always
* When set to "always", the REST HTTP server does not use some
HTTP 1.1 features such as persistent connections or chunked
transfer encoding.
* When set to "auto" it does this only if the client sent no
User-Agent header, or if the user agent is known to have bugs
in its HTTP/1.1 support.
* When set to "never" it always allows HTTP 1.1, even to
clients it suspects may be buggy.
* Default: auto
crossOriginSharingPolicy = <origin_acl> ...
* List of the HTTP Origins for which to return Access-Control-Allow-* (CORS)
headers.
* These headers tell browsers that web applications are trusted at those sites
to make requests to the REST interface.
* The origin is passed as a URL without a path component (for example
"https://app.example.com:8000").
* This setting can take a list of acceptable origins, separated
by spaces and/or commas.
* Each origin can also contain wildcards for any part. Examples:
*://app.example.com:* (either HTTP or HTTPS on any port)
https://*.example.com (any host under example.com, including
example.com itself)
* An address can be prefixed with a '!' to negate the match, with
the first matching origin taking precedence. For example,
"!*://evil.example.com:* *://*.example.com:*" to not avoid
matching one host in a domain
* A single "*" can also be used to match all origins
* No default.
crossOriginSharingHeaders = <string>
* A list of the HTTP headers to which splunkd sets
"Access-Control-Allow-Headers" when replying to
Cross-Origin Resource Sharing (CORS) preflight requests.
* The "Access-Control-Allow-Headers" header is used in response to
a CORS preflight request to tell browsers which HTTP headers can be
used during the actual request.
* A CORS preflight request is a CORS request that checks to see if
the CORS protocol is understood and a server is aware of using
specific methods and headers.
* This setting can take a list of acceptable HTTP headers, separated
by commas.
* A single "*" can also be used to match all headers.
* Default: Empty string.
x_frame_options_sameorigin = <boolean>
* Adds a X-Frame-Options header set to "SAMEORIGIN" to every response
served by splunkd.
* Default: true
allowEmbedTokenAuth = <boolean>
* If set to false, splunkd does not allow any access to artifacts
that previously had been explicitly shared to anonymous users.
* This effectively disables all use of the "embed" feature.
* Default: true
cliLoginBanner = <string>
* Sets a message which is added to the HTTP reply headers
of requests for authentication, and to the "server/info" endpoint
* This is printed by the Splunk CLI before it prompts
for authentication credentials. This can be used to print
access policy information.
* If this string starts with a '"' character, it is treated as a
CSV-style list with each line comprising a line of the message.
For example: "Line 1","Line 2","Line 3"
* No default.
allowBasicAuth = <boolean>
* Allows clients to make authenticated requests to the splunk
server using "HTTP Basic" authentication in addition to the
normal "authtoken" system
* This is useful for programmatic access to REST endpoints and
for accessing the REST API from a web browser. It is not
required for the UI or CLI.
* Default: true
allowWwwAuthHeader = <boolean>
* Describes whether or not Splunk Web can include a "www-authenticate" header
in a response to a request from a web client to access a management endpoint.
* When Splunk Web sends the "www-authenticate" header in response to such
a request, the client forces its user to provide credentials to authenticate.
* A value of "true" means that Splunk Web sends a "www-authenticate" header
in its response to the web client request. This means that the user of that
client will be prompted to enter valid credentials to access the instance,
even if they provide those credentials as part of the request.
* A value of "false" means that Splunk Web does not send the "www-authenticate"
header in its response to the web client request. This means that the
user of that client will not be prompted to provide valid credentials to
access the instance.
* Giving this setting a value of "false" reduces the attack surface in the
management API when you access it through Splunk Web.
* This setting is not valid for the CLI. It works only with Splunk Web.
* Default: true
basicAuthRealm = <string>
* When using "HTTP Basic" authentication, the 'realm' is a
human-readable string describing the server. Typically, a web
browser presents this string as part of its dialog box when
asking for the username and password.
* This can be used to display a short message describing the
server and/or its access policy.
* Default: /splunk
allowCookieAuth = <boolean>
* Allows clients to request an HTTP cookie from the /services/auth/login
endpoint which can then be used to authenticate future requests
* Default: true
cookieAuthHttpOnly = <boolean>
* When using cookie based authentication, mark returned cookies
with the "httponly" flag to tell the client not to allow javascript
code to access its value
* NOTE: has no effect if allowCookieAuth=false
* Default: true
cookieSameSiteSecure = <boolean>
* Describes whether or not the Splunk web server will set all Splunk cookies
in the browser with the "SameSite=None;Secure" attribute.
* Use this setting to toggle when cookies are to get this value as part of
embedding dashboards into third-party web applications outside of the
Splunk platform instance.
* A value of "true" means that the Splunk web server sets the "SameSite=None,Secure"
attribute for cookies to work with modern browsers. A cookie with this
attribute lets the browser embed a Splunk dashboard into a third-party <iframe>.
* <iframe> stands for "inline frame", and is a web page component within
which you can embed a dashboard.
* A value of "false" means that the web server does not set cookies with
"SameSite=None,Secure" attribute.
* If you want to embed a Splunk dashboard into an outside web application,
you must give this setting a value of "true". Otherwise, the third-party
<iframe> won't let the user authenticate and use the dashboard that
you embedded.
* Default: false
cookieAuthSecure = <boolean>
* When using cookie based authentication, mark returned cookies
with the "secure" flag to tell the client never to send it over
an unencrypted HTTP channel
* NOTE: has no effect if allowCookieAuth=false OR the splunkd REST
interface has SSL disabled
* Default: true
dedicatedIoThreads = [<integer>|auto]
* The number of threads that splunkd dedicates to handling HTTP I/O requests.
* This setting controls thread usage for all HTTP requests through splunkd,
including SSL encryption.
* If you set this to "0", splunkd uses the same thread that accepted the initial
connection over TCP to perform the HTTP I/O.
* If you set this to a number other than "0", splunkd creates that number of
threads to handle HTTP I/O.
* If you set this to "auto", splunkd uses the number of CPU cores on the
machine to determine the number of threads available for HTTP I/O as
follows:
* Number of CPU cores available | 'dedicatedIoThreads'
0 - 16 | 0
17 - 48 | 2
49 - 128 | 4
129 - 192 | 6
193 and higher | 8
* You do not usually need to change this setting.
* Default: auto
dedicatedIoThreadsSelectionPolicy = <round_robin | weighted_random>
* Specifies the I/O threads selection policy to use while selecting I/O thread
for new connection.
* If set to "round_robin", the incoming connections are assigned to I/O threads
in a round robin fashion.
* If set to "weighted_random", the connections are assigned to I/O threads using
a weighted random scheme designed to even out the CPU usage of each I/O thread.
* NOTE: This setting only takes effect when dedicatedIoThreads is greater than 1.
* Default: round_robin
dedicatedIoThreadsWeightsUpdatePeriod = <number>
* The interval, in seconds, when I/O thread weights are recalculated for the
"weighted_random" selection policy.
* Reducing this interval causes the weights to be re-evaluated more
frequently, thereby enabling the system to react more quickly to changes
in relative thread load.
* Increasing this interval causes the weights to be re-evaluated less
frequently, thereby reducing the ability of the system to respond to
bursty events.
* Default: 30
replyHeader.<name> = <string>
* Add a static header to all HTTP responses this server generates
* For example, "replyHeader.My-Header = value" causes the
response header "My-Header: value" to be included in the reply to
every HTTP request to the REST server
Splunkd HTTPServer listener configuration
[httpServerListener:<ip:><port>]
* Enable the splunkd REST HTTP server to listen on an additional port number
specified by <port>. If a non-empty <ip> is included (for example:
"[httpServerListener:127.0.0.1:8090]") the listening port is
bound only to a specific interface.
* Multiple "httpServerListener" stanzas can be specified to listen on
more ports.
* Normally, splunkd listens only on the single REST port specified in
the web.conf "mgmtHostPort" setting, and none of these stanzas need to
be present. Add these stanzas only if you want the REST HTTP server
to listen to more than one port.
ssl = <boolean>
* Toggle whether this listening ip:port uses SSL or not.
* If the main REST port is SSL (the "enableSplunkdSSL" setting in this
file's [sslConfig] stanza) and this stanza is set to "ssl=false" then
clients on the local machine such as the CLI may connect to this port.
* Default: true
listenOnIPv6 = no|yes|only
* Toggle whether this listening ip:port listens on IPv4, IPv6, or both.
* If not present, the setting in the [general] stanza is used
acceptFrom = <network_acl> ...
* Lists a set of networks or addresses from which to accept connections.
* Separate multiple rules with commas or spaces.
* Each rule can be in one of the following formats:
1. A single IPv4 or IPv6 address (examples: "10.1.2.3", "fe80::4a3")
2. A Classless Inter-Domain Routing (CIDR) block of addresses
(examples: "10/8", "192.168.1/24", "fe80:1234/32")
3. A DNS name, possibly with a "*" used as a wildcard
(examples: "myhost.example.com", "*.splunk.com")
4. "*", which matches anything
* You can also prefix an entry with '!' to cause the rule to reject the
connection. The input applies rules in order, and uses the first one that
matches.
For example, "!10.1/16, *" allows connections from everywhere except
the 10.1.*.* network.
* Default: The setting in the [httpServer] stanza
Static file handler MIME-type map
[mimetype-extension-map]
* Map filename extensions to MIME type for files served from the static file
handler under this stanza name.
<file-extension> = <MIME-type>
* Instructs the HTTP static file server to mark any files ending
in 'file-extension' with a header of 'Content-Type: <MIME-type>'.
* Default:
[mimetype-extension-map]
gif = image/gif
htm = text/html
jpg = image/jpg
png = image/png
txt = text/plain
xml = text/xml
xsl = text/xml
Log rotation of splunkd_stderr.log & splunkd_stdout.log
# These stanzas apply only on UNIX. splunkd on Windows has no # stdout.log or stderr.log files. [stderr_log_rotation] * Controls the data retention of the file containing all messages written to splunkd's stderr file descriptor (fd 2). * Typically this is extremely small, or mostly errors and warnings from linked libraries. maxFileSize = <bytes> * When splunkd_stderr.log grows larger than this value, it is rotated. * maxFileSize is expressed in bytes. * You might want to increase this if you are working on a problem that involves large amounts of output to the splunkd_stderr.log file. * You might want to reduce this to allocate less storage to this log category. * Default: 10000000 (10 si-megabytes) BackupIndex = <non-negative integer> * How many rolled copies to keep. * For example, if this setting is 2, the splunkd_stderr.log.1 and splunkd_stderr.log.2 file might exist. Further rolls delete the current splunkd_stderr.log.2 file. * You might want to increase this value if you are working on a problem that involves large amounts of output to the splunkd_stderr.log fils * You might want to reduce this to allocate less storage to this log category. * Default: 2 checkFrequency = <seconds> * How often. in seconds, to check the size of splunkd_stderr.log * Larger values may result in larger rolled file sizes but take less resources. * Smaller values may take more resources but more accurately constrain the file size. * Default: 10 [stdout_log_rotation] * Controls the data retention of the file containing all messages written to splunkd's stdout file descriptor (fd 1). * Almost always, there is nothing in this file. * This stanza can have the same settings as the [stderr_log_rotation] stanza with the same defaults. See above for definitions. maxFileSize = <bytes> BackupIndex = <non-negative integer> checkFrequency = <seconds>
Remote applications configuration (e.g. SplunkBase)
[applicationsManagement] * Set remote applications settings for Splunk under this stanza name. * Follow this stanza name with any number of the following attribute/value pairs. * If you do not specify an entry for each attribute, Splunk uses the default value. allowInternetAccess = <boolean> * Allow Splunk to access the remote applications repository. url = <URL> * Applications repository. * Default: https://apps.splunk.com/api/apps loginUrl = <URL> * Applications repository login. * Default: https://apps.splunk.com/api/account:login/ detailsUrl = <URL> * Base URL for application information, keyed off of app ID. * Default: https://apps.splunk.com/apps/id useragent = <splunk-version>-<splunk-build-num>-<platform> * User-agent string to use when contacting applications repository. * <platform> includes information like operating system and CPU architecture. updateHost = <URL> * Host section of URL to check for app updates, e.g. https://apps.splunk.com updatePath = <URL> * Path section of URL to check for app updates For example: /api/apps:resolve/checkforupgrade updateTimeout = <time range string> * The minimum amount of time Splunk software waits between checks for app updates. * Examples include '24h' (24 hours), '3d' (3 days), '7200s' (7200 seconds, or two hours) * Default: 24h sslVersions = <versions_list> * Comma-separated list of SSL versions to connect to 'url' (https://apps.splunk.com). * The versions available are "ssl3", "tls1.0", "tls1.1", and "tls1.2". * The special version "*" selects all supported versions. The version "tls" selects all versions tls1.0 or newer. * If a version is prefixed with "-" it is removed from the list. * SSLv2 is always disabled; "-ssl2" is accepted in the version list but does nothing. * When configured in FIPS mode, ssl3 is always disabled regardless of this configuration. * Default: The default can vary (See the 'sslVersions' setting in the $SPLUNK_HOME/etc/system/default/server.conf file for the current default) sslVerifyServerCert = <boolean> * If this is set to true, Splunk verifies that the remote server ( specified in 'url') being connected to is a valid one (authenticated). Both the common name and the alternate name of the server are then checked for a match if they are specified in 'sslCommonNameToCheck' and 'sslAltNameToCheck'. A certificate is considered verified if either is matched. * Default: true caCertFile = <path> * Full path to a CA (Certificate Authority) certificate(s) PEM format file. * The <path> must refer to a PEM format file containing one or more root CA certificates concatenated together. * Used only if 'sslRootCAPath' is not set. * Used for validating SSL certificate from https://apps.splunk.com/ sslCommonNameToCheck = <commonName1>, <commonName2>, ... * If this value is set, and 'sslVerifyServerCert' is set to true, splunkd checks the common name(s) of the certificate presented by the remote server (specified in 'url') against this list of common names. * Default: apps.splunk.com sslCommonNameList = <commonName1>, <commonName2>, ... * DEPRECATED. Use the 'sslCommonNameToCheck' setting instead. sslAltNameToCheck = <alternateName1>, <alternateName2>, ... * If this value is set, and 'sslVerifyServerCert' is set to true, splunkd checks the alternate name(s) of the certificate presented by the remote server (specified in 'url') against this list of subject alternate names. * Default: splunkbase.splunk.com, apps.splunk.com cipherSuite = <cipher suite string> * Uses the specified cipher string for making outbound HTTPS connection. * The default can vary. See the 'cipherSuite' setting in the $SPLUNK_HOME/etc/system/default/server.conf file for the current default. ecdhCurves = <comma separated list of ec curves> * ECDH curves to use for ECDH key negotiation. * The curves should be specified in the order of preference. * The client sends these curves as a part of Client Hello. * Splunk software only supports named curves specified by their SHORT names. * The list of valid named curves by their short/long names can be obtained by executing this command: $SPLUNK_HOME/bin/splunk cmd openssl ecparam -list_curves * e.g. ecdhCurves = prime256v1,secp384r1,secp521r1 * Default: The default can vary (See the 'ecdhCurves' setting in the $SPLUNK_HOME/etc/system/default/server.conf file for the current default)
Misc. configuration
[scripts] initialNumberOfScriptProcesses = <num> * The number of pre-forked script processes that are launched when the system comes up. These scripts are reused when script REST endpoints *and* search scripts are executed. The idea is to eliminate the performance overhead of launching the script interpreter every time it is invoked. These processes are put in a pool. If the pool is completely busy when a script gets invoked, a new processes is fired up to handle the new invocation - but it disappears when that invocation is finished.
Disk usage settings (for the indexer, not for Splunk log files)
[diskUsage]
minFreeSpace = <num>|<percentage>
* Minimum free space for a partition.
* Specified as an integer that represents a size in binary
megabytes (ie MiB) or as a percentage, written as a decimal
between 0 and 100 followed by a '%' sign, for example "10%"
or "10.5%"
* If specified as a percentage, this is taken to be a percentage of
the size of the partition. Therefore, the absolute free space required
varies for each partition depending on the size of that partition.
* Specifies a safe amount of space that must exist for splunkd to continue
operating.
* Note that this affects search and indexing
* For search:
* Before attempting to launch a search, Splunk software requires this
amount of free space on the filesystem where the dispatch directory
is stored, $SPLUNK_HOME/var/run/splunk/dispatch
* Applied similarly to the search quota values in authorize.conf and
limits.conf.
* For indexing:
* Periodically, the indexer checks space on all partitions
that contain splunk indexes as specified by indexes.conf. Indexing
is paused and a ui banner + splunkd warning posted to indicate
need to clear more disk space.
* Default: 5000 (approx 5GB)
pollingFrequency = <num>
* Specifies that after every 'pollingFrequency' events are indexed,
the disk usage is checked.
* Default: 100000
pollingTimerFrequency = <num>
* Minimum time, in seconds, between two disk usage checks.
* Default: 10
Queue settings
[queue] maxSize = [<integer>|<integer>[KB|MB|GB]] * Specifies default capacity of a queue. * If specified as a lone integer (for example, maxSize=1000), maxSize indicates the maximum number of events allowed in the queue. * If specified as an integer followed by KB, MB, or GB (for example, maxSize=100MB), it indicates the maximum RAM allocated for queue. * Default: 500KB cntr_1_lookback_time = [<integer>[s|m]] * The lookback counters are used to track the size and count (number of elements in the queue) variation of the queues using an exponentially moving weighted average technique. Both size and count variation has 3 sets of counters each. The set of 3 counters is provided to be able to track short, medium and long term history of size/count variation. The user can customize the value of these counters or lookback time. * Specifies how far into history should the size/count variation be tracked for counter 1. * It must be an integer followed by [s|m] which stands for seconds and minutes respectively. * Default: 60s cntr_2_lookback_time = [<integer>[s|m]] * See above for explanation and usage of the lookback counter. * Specifies how far into history should the size/count variation be tracked for counter 2. * Default: 600s (10 minutes) cntr_3_lookback_time = [<integer>[s|m]] * See above for explanation and usage of the lookback counter.. * Specifies how far into history should the size/count variation be tracked for counter 3. * Default: 900s (15 minutes) sampling_interval = [<integer>[s|m]] * The lookback counters described above collects the size and count measurements for the queues. This specifies at what interval the measurement collection happens. Note that for a particular queue all the counters sampling interval is same. * It needs to be specified via an integer followed by [s|m] which stands for seconds and minutes respectively. * Default: 1s [queue=<queueName>] maxSize = [<integer>|<integer>[KB|MB|GB]] * Specifies the capacity of a queue. It overrides the default capacity specified in the [queue] stanza. * If specified as a lone integer (for example, maxSize=1000), maxSize indicates the maximum number of events allowed in the queue. * If specified as an integer followed by KB, MB, or GB (for example, maxSize=100MB), it indicates the maximum RAM allocated for queue. * Default: The default is inherited from the 'maxSize' value specified in the [queue] stanza cntr_1_lookback_time = [<integer>[s|m]] * Same explanation as mentioned in the [queue] stanza. * Specifies the lookback time for the specific queue for counter 1. * Default: The default value is inherited from the 'cntr_1_lookback_time' value that is specified in the [queue] stanza cntr_2_lookback_time = [<integer>[s|m]] * Specifies the lookback time for the specific queue for counter 2. * Default: The default value is inherited from the 'cntr_2_lookback_time' value that is specified in the [queue] stanza. cntr_3_lookback_time = [<integer>[s|m]] * Specifies the lookback time for the specific queue for counter 3. * Default: The default value is inherited from the 'cntr_3_lookback_time' value that is specified in the [queue] stanza. sampling_interval = [<integer>[s|m]] * Specifies the sampling interval for the specific queue. * Default: The default value is inherited from the 'sampling_interval' value specified in the [queue] stanza.
PubSub server settings for the http endpoint.
[pubsubsvr-http] disabled = <boolean> * If disabled, then http endpoint is not registered. Set this value to 'false' to expose PubSub server on http. * Default: true stateIntervalInSecs = <seconds> * The number of seconds before a connection is flushed due to inactivity. The connection is not closed, only messages for that connection are flushed. * Default: 300 (5 minutes)
General file input settings. ** NOT SUPPORTED **
# [fileInput] # outputQueue = <queue name> * REMOVED. Historically this allowed the user to set the target queue for the file-input (tailing) processor, but there was no valid reason to modify this. * This setting is now removed, and has no effect. * Tailing always uses the parsingQueue.
Settings controlling the behavior of 'splunk diag', the diagnostic tool
[diag]
# These settings provide defaults for invocations of the splunk diag
# command. Generally these can be further modified by command line flags to
# the diag command.
EXCLUDE-<class> = <glob expression>
* Specifies a glob / shell pattern to be excluded from diags generated on
this Splunk instance.
* Example: */etc/secret_app/local/*.conf
* Further excludes can be added at the splunk diag command line, but there
is no facility to disable configuration-based excludes at the command
line.
* There is one exclude by default, for the splunk.secret file.
# the following commands can be overridden entirely by their command-line
# equivalents.
components = <comma separated list>
* Specifies which components of the diag should be gathered.
* This allows the disabling and enabling, categorically, of entire portions
of diag functionality.
* All of these components are further subject to the exclude feature (see
above), and component-specific filters (see below).
* Currently, with no configuration, all components except "rest" are enabled
by default.
* Available components are:
* index_files : Files from the index that indicate their health
(Hosts|Sources|Sourcetypes.data and bucketManifests).
User data is not collected.
* index_listing : Directory listings of the index contents are
gathered, in order to see filenames, directory names,
sizes, timestamps and the like.
* etc : The entire contents of the $SPLUNK_HOME/etc
directory. In other words, the configuration files.
* log : The contents of $SPLUNK_HOME/var/log/...
* pool : If search head pooling is enabled, the contents of the
pool dir.
* dispatch : Search artifacts, without the actual results,
In other words var/run/splunk/dispatch, but not the
results or events files
* searchpeers : Directory listings of knowledge bundles replicated for
distributed search
In other words: $SPLUNK_HOME/var/run/searchpeers
* consensus : Consensus protocol files produced by search head clustering
In other words: $SPLUNK_HOME/var/run/splunk/_raft
* conf_replication_summary : Directory listing of configuration
replication summaries produced by search head clustering
In other words: $SPLUNK_HOME/var/run/splunk/snapshot
* rest : The contents of a variety of splunkd endpoints
Includes server status messages (system banners),
licenser banners, configured monitor inputs & tailing
file status (progress reading input files).
* On cluster managers, also gathers manager info, fixups,
current peer list, clustered index info, current
generation, & buckets in bad stats
* On cluster peers, also gathers local buckets & local
peer info, and the manager information remotely from
the configured manager.
* kvstore : Directory listings of the KV Store data directory
contents are gathered, in order to see filenames,
directory names, sizes, and timestamps.
* file_validate : Produce list of files that were in the install media
which have been changed. Generally this should be an
empty list.
* profiler : The profiler directory at $SPLUNK_HOME/var/run/profiler
* The special value "all" is also supported, enabling everything explicitly.
* Further controlling the components from the command line:
* The switch --collect replaces this list entirely.
* Example: --collect log,etc
This would set the components to log and etc only, regardless of
config
* The switch --enable adds a specific component to this list.
* Example: --enable pool
This would ensure that pool data is collected, regardless of
config
* The switch --disable removes a specific component from this list.
* Example: --disable pool
This would ensure that pool data is *NOT* collected, regardless of
config
* Default: To collect all components, except "rest".
# Data filters; these further refine what is collected
# most of the existing ones are designed to limit the size and collection
# time to pleasant values.
# NOTE: Most values here use underscores '_' while the command line uses
# hyphens '-'
all_dumps = <boolean>
* This setting currently is irrelevant on UNIX platforms.
* Affects the 'log' component of diag. (dumps are written to the log dir
on Windows)
* Can be overridden with the --all-dumps command line flag.
* Normally, Splunk diag gathers only three .DMP (crash dump) files on
Windows to limit diag size.
* If this is set to true, splunk diag collects *all* .DMP files from
the log directory.
* No default. (false equivalent).
index_files = [full|manifests]
* Selects a detail level for the 'index_files' component.
* Can be overridden with the --index-files command line flag.
* If set to 'manifests', limits the index file-content collection to just
.bucketManifest files which give some information about the general state of
buckets in an index.
* If set to 'full', adds the collection of Hosts.data, Sources.data, and
Sourcetypes.data which indicate the breakdown of count of items by those
categories per-bucket, and the timespans of those category entries
* 'full' can take quite some time on very large index sizes, especially
when slower remote storage is involved.
* Default: manifests
index_listing = [full|light]
* Selects a detail level for the 'index_listing' component.
* Can be overridden with the --index-listing command line flag.
* 'light' gets directory listings (ls, or dir) of the hot/warm and cold
container directory locations of the indexes, as well as listings of each
hot bucket.
* 'full' gets a recursive directory listing of all the contents of every
index location, which should mean all contents of all buckets.
* 'full' may take significant time as well with very large bucket counts,
especially on slower storage.
* Default: light
etc_filesize_limit = <non-negative integer in kilobytes>
* This filters the 'etc' component
* Can be overridden with the --etc-filesize-limit command line flag
* This value is specified in kilobytes.
* Example: 2000 - this would be approximately 2MB.
* Files in the $SPLUNK_HOME/etc directory which are larger than this limit
is not collected in the diag.
* Diag produces a message stating that a file has been skipped for size
to the console. (In practice, large files have been found to oftentimes
be a surprise to the administrator, and indicate problems).
* If desired, this filter may be entirely disabled by setting the value
to 0.
* Currently, as a special exception, the file
$SPLUNK_HOME?etc/system/replication/ops.json is permitted to be 10x the
size of this limit.
* Default: 10000 (10MB)
log_age = <non-negative integer in days>
* This filters the 'log' component
* Can be overridden with the --log-age command line flag
* This value is specified in days
* Example: 75 - this would be 75 days, or about 2.5 months.
* If desired, this filter may be entirely disabled by setting the value to 0.
* The idea of this default filter is that data older than this is rarely
helpful in troubleshooting cases in any event.
* Default: 60 (or approximately 2 months)
upload_proto_host_port = <protocol://host:port>|disabled
* URI base to use for uploading files/diags to Splunk support.
* If set to disabled (override in a local/server.conf file), effectively
disables diag upload functionality for this Splunk install.
* Modification may theoretically may permit operations with some forms of
proxies, but diag is not specifically designed for such, and support of proxy
configurations that do not currently work is considered an Enhancement
Request.
* The communication path with api.splunk.com is over a simple but not
documented protocol. If for some reason you wish to accept diag uploads into
your own systems, it probably is simpler to run diag and then upload via
your own means independently. However if you have business reasons that you
want this built-in, get in touch.
* Uploading to unencrypted http definitely not recommended.
* Default: https://api.splunk.com
SEARCHFILTERSIMPLE-<class> = regex
SEARCHFILTERLUHN-<class> = regex
* Redacts strings from ad-hoc searches logged in the audit.log and
remote_searches.log files.
* Substrings which match these regexes *inside* a search string in one of those
two files is replaced by sequences of the character X, as in XXXXXXXX.
* Substrings which match a SEARCHFILTERLUHN regex has the contained
numbers further tested against the luhn algorithm, used for data integrity
in mostly financial circles, such as credit card numbers. This permits more
accurate identification of that type of data, relying less heavily on regex
precision. See the Wikipedia article on the "Luhn algorithm" for additional
information.
* Search string filtering is entirely disabled if --no-filter-searchstrings is
used on the command line.
* NOTE: That matching regexes must take care to match only the bytes of the
term. Each match "consumes" a portion of the search string, so matches that
extend beyond the term (for example, to adjacent whitespace) could prevent
subsequent matches, and/or redact data needed for troubleshooting.
* Please use a name hinting at the purpose of the filter in the <class>
component of the setting name, and consider an additional explicative
comment, even for custom local settings. This might skip inquiries from
support.
Application License manager settings for configuring app license checking
[applicense] appLicenseHostPort = <IP:port> * Specifies the location of the IP address or DNS name and port of the app license server. appLicenseServerPath = <path> * Specifies the path portion of the URI of the app license server. caCertFile = <path> * Full path to a CA (Certificate Authority) certificate(s) PEM format file. * NOTE: Splunk plans to submit Splunk Enterprise for Common Criteria evaluation. Splunk does not support using the product in Common Criteria mode until it has been certified by NIAP. See the "Securing Splunk Enterprise" manual for information on the status of Common Criteria certification. * Default: $SPLUNK_HOME/etc/auth/applicenseCA.pem sslVersions = <versions_list> * Comma-separated list of SSL versions to support. * The special version "*" selects all supported versions. The version "tls" selects all versions tls1.0 or newer. * If a version is prefixed with "-" it is removed from the list. * SSLv2 is always disabled; "-ssl2" is accepted in the version list but does nothing. * When configured in FIPS mode, ssl3 is always disabled regardless of this configuration. * Default: The default can vary (See the 'sslVersions' setting in the $SPLUNK_HOME/etc/system/default/server.conf file for the current default) cipherSuite = <cipher suite string> * If set, uses the specified cipher string for the SSL connection. * Default: The default can vary (See the 'cipherSuite' setting in the $SPLUNK_HOME/etc/system/default/server.conf file for the current default) sslVerifyServerCert = <boolean> * If this is set to true, Splunk Enterprise verifies that the remote server (specified in 'url') being connected to is a valid one (authenticated). * Both the common name and the alternate name of the server are then checked for a match if they are specified in 'sslCommonNameToCheck' and 'sslAltNameToCheck'. * A certificate is considered verified if either is matched. * Default: true sslCommonNameToCheck = <commonName1>, <commonName2>, ... * If this value is set, and 'sslVerifyServerCert' is set to true, splunkd limits most outbound HTTPS connections to hosts which use a cert with one of the listed common names. * Default: Some common name checking . sslAltNameToCheck = <alternateName1>, <alternateName2>, ... * If this value is set, and 'sslVerifyServerCert' is set to true, splunkd is also willing to verify certificates which have a so-called "Subject Alternate Name" that matches any of the alternate names in this list. * Subject Alternate Names are effectively extended descriptive fields in SSL certs beyond the commonName. A common practice for HTTPS certs is to use these values to store additional valid hostnames or domains where the cert should be considered valid. * Accepts a comma-separated list of Subject Alternate Names to consider valid. * Items in this list are never validated against the SSL Common Name. * Default: Some alternate name checking disabled = <boolean> * Select true to disable this feature or false to enable this feature. App licensing is experimental, so it is disabled by default. * Default: true
License manager settings for configuring the license pool(s)
[license]
master_uri = [self|<uri>]
* The URI of the license master that a license slave connects to.
* If set to a URI, the instance attempts to connect to the license master at the URI you specify.
* A URI consists of the following: <scheme>://<hostname>:<port>
* For example, if you set "master_uri = https://example.con:8089", then the instance attempts
a connection to the instance at "http://example.com:8089" to get licensing information.
* No default.
active_group = Enterprise|Trial|Forwarder|Free
* If the instance is a license master, the license type will be set in 'active_group'.
* Default: <empty>
connection_timeout = <integer>
* Maximum time, in seconds, to wait before sending data to master times out
* This timeout applies only if 'master_uri' is set.
* Default: 30
send_timeout = <integer>
* Maximum time, in seconds, to wait before sending data to master times out
* This timeout applies only if 'master_uri' is set.
* Default: 30
receive_timeout = <integer>
* Maximum time, in seconds, to wait before receiving data from master times out
* This timeout applies only if 'master_uri' is set.
* Default: 30
squash_threshold = <positive integer>
* Advanced setting. Periodically the indexer must report to license manager
the data indexed broken down by source, sourcetype, host, and index. If
the number of distinct (source, sourcetype, host, index) tuples grows over
the 'squash_threshold', the (host, source) values are squashed and only a
breakdown by (sourcetype, index) is reported. This is to prevent explosions in
memory + license_usage.log lines.
* Set this only after consulting a Splunk Support engineer.
* This needs to be set on license slaves as well as license
master.
* Default: 2000
report_interval = <nonnegative integer>[s|m|h]
* Selects a time period for reporting in license usage to the license
master.
* This value is intended for very large deployments (hundreds of indexers)
where a large number of indexers may overwhelm the license server.
* The maximum permitted interval is 1 hour.
* The minimum permitted interval is 1 minute.
* Can be expressed as a positive number of seconds, minutes or hours.
* If no time unit is provided, seconds is assumed.
* Default: 1m
license_warnings_update_interval = <nonnegative integer>
* Specifies a time period, in seconds, for license master to update
license warnings in Splunk Web bulletin messages.
* License master checks at every second the last time it updated the
warnings, and updates if this time period has elapsed.
* Increase this value for very large deployments that contain very
large number of source types.
* The minimum permitted interval is 10.
* The maximum permiited interval is 3600, equivalent to 1 hour.
* If set to special value of 0, the license master automatically tunes this
setting to accommodate the size of the deployment.
* Default: 0
strict_pool_quota = <boolean>
* Toggles strict pool quota enforcement
* If set to true, members of pools receive warnings for a given day if
usage exceeds pool size regardless of whether overall stack quota was
exceeded
* If set to false, members of pool only receive warnings if both pool
usage exceeds pool size AND overall stack usage exceeds stack size
* Default: true
pool_suggestion = <string>
* Suggest a pool to the master for this slave.
* The master uses this suggestion if the master doesn't have an explicit
rule mapping the slave to a given pool (ie...no slave list for the
relevant license stack contains this slave explicitly)
* If the pool name doesn't match any existing pool, it is ignored, no
error is generated
* This setting is intended to give an alternative management option for
pool/slave mappings. When onboarding an indexer, it may be easier to
manage the mapping on the indexer itself via this setting rather than
having to update server.conf on master for every addition of new indexer
* NOTE: If you have multiple stacks and a slave maps to multiple pools, this
feature is limited in only allowing a suggestion of a single pool;
This is not a common scenario however.
* No default. (which means this feature is disabled)
[lmpool:auto_generated_pool_forwarder]
* This is the auto generated pool for the forwarder stack
description = <textual description of this license pool>
quota = MAX|<maximum amount allowed by this license>
* MAX indicates the total capacity of the license. You may have only 1 pool
with MAX size in a stack
* The quota can also be specified as a specific size eg. 20MB, 1GB etc
slaves = *|<slave list>
* An asterisk(*) indicates that any slave can connect to this pool
* You can also specify a comma separated slave guid list
stack_id = forwarder
* The stack to which this pool belongs.
[lmpool:auto_generated_pool_free]
* This is the auto generated pool for the free stack
* Field descriptions are the same as that for
the 'lmpool:auto_generated_pool_forwarder' setting.
[lmpool:auto_generated_pool_enterprise]
* This is the auto generated pool for the enterprise stack
* Field descriptions are the same as that for
the 'lmpool:auto_generated_pool_forwarder' setting.
[lmpool:auto_generated_pool_fixed-sourcetype_<sha256 hash of srctypes>]
* This is the auto generated pool for the enterprise fixed srctype stack
* Field descriptions are the same as that for
the 'lmpool:auto_generated_pool_forwarder' setting.
[lmpool:auto_generated_pool_download_trial]
* This is the auto generated pool for the download trial stack
* Field descriptions are the same as that for
the "lmpool:auto_generated_pool_forwarder"
[pooling]
state = [enabled|disabled]
* UNSUPPORTED: This setting is no longer supported
storage = <path to shared storage>
* UNSUPPORTED: This setting is no longer supported
app_update_triggers = true|false|silent
* UNSUPPORTED: This setting is no longer supported
lock.timeout = <time range string>
* UNSUPPORTED: This setting is no longer supported
lock.logging = <boolean>
* UNSUPPORTED: This setting is no longer supported
poll.interval.rebuild = <time range string>
* UNSUPPORTED: This setting is no longer supported
poll.interval.check = <time range string>
* UNSUPPORTED: This setting is no longer supported
poll.blacklist.<name> = <regex>
* UNSUPPORTED: This setting is no longer supported
AWS EC2 IMDS configuration
[imds] imds_version = [v1|v2] * Sets IMDS version for EC2 instances metadata endpoints. * This setting is AWS specific. * Certain features of the Splunk platform use AWS Instance Metadata Service (IMDS) when hosted on EC2. IMDS is accessible from the instance via a link-local address. It provides metadata about the instance. * v1 uses request/response method while v2 uses a session-oriented method to access IMDS. The version must match the setting used on your EC2 instance. * More information about IMDS can be found in the AWS documentation. * Default: v1
High availability clustering configuration
[clustering]
mode = [manager|peer|searchhead|disabled]
* Sets operational mode for this cluster node.
* Only one manager may exist per cluster.
* Note: "manager" and "peer" replace the prior 'mode' values of
"master" and "slave". The prior values are currently still supported,
but they will be removed from the product in a future release.
* Default: disabled
master_uri = [<uri> | clustermanager:stanzaName1, clustermanager:stanzaName2]
* DEPRECATED. Use the 'manager_uri' setting instead.
manager_uri = [<uri> | clustermanager:stanzaName1, clustermanager:stanzaName2]
* Only valid for 'mode=peer' or 'mode=searchhead'.
* The URI of the cluster manager that this peer or search head
should connect to.
* An example of <uri>: <scheme>://<hostname>:<port>
* Only for 'mode=searchhead' - If the search head is a part of multiple
clusters, the manager URIs can be specified by a comma separated list.
advertised_disk_capacity = <integer>
* Percentage to use when advertising disk capacity to the cluster manager.
This is useful for modifying weighted load balancing in indexer discovery.
* For example, if you set this attribute to 50 for an indexer with a
500GB disk, the indexer advertises its disk size as 250GB, not 500GB.
* Acceptable value range is 10 to 100.
* Default: 100
pass4SymmKey = <password>
* Secret shared among the nodes in the cluster to prevent any
arbitrary node from connecting to the cluster. If a peer or
search head is not configured with the same secret as the manager,
it is not able to communicate with the manager.
* If it is not set in the [clustering] stanza, the key
is looked in the [general] stanza
* Unencrypted passwords must not begin with "$1$", as this is used by
Splunk software to determine if the password is already encrypted.
* No default.
pass4SymmKey_minLength = <integer>
* The minimum length, in characters, that a 'pass4SymmKey' should be for a particular stanza.
* When you start the Splunk platform, if the 'pass4SymmKey' is shorter in length than
what you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* If you use the CLI to modify 'pass4SymmKey' to a value that is shorter than what
you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* Default: 12
service_interval = <zero or positive integer>
* Only valid when 'mode=manager'.
* Specifies, in seconds, how often the manager runs its service
loop. In its service loop, the manager checks the state of the
peers and the buckets in the cluster and also schedules
corrective action, if possible, for buckets that are not in
compliance with replication policies.
* A special default value of 0 indicates an auto mode where the service
interval for the next service call is determined by the time taken by
previous call. It also sets the minimum service interval to be 0.5 second.
* Service interval is bounded by the values 1 and
the 'max_auto_service_interval' setting.
If previous service call takes more than 'max_auto_service_interval'
seconds, next service interval is set to
'max_auto_service_interval' seconds.
* Default: 0
service_execution_threshold_ms = <zero or positive integer>
* Only valid when 'mode=manager'.
* Specifies, in milliseconds, the maximum period for one execution
of the manager's service loop.
* This setting is useful for large clusters with large numbers of
buckets, to prevent the service loop from blocking
other operations for significant amounts of time.
* Default: 1500
deferred_cluster_status_update = <boolean>
* Only valid when 'mode=manager'
* If set to true (default), SF/RF met (complete cluster state) checks are
performed lazily for optimal performance, only when CM is busy with
cluster maintenance operations (e.g peer addition, fix ups, data rebalance).
* If set to false, SF/RF met checks are performed relatively more aggressively
to improve accuracy, increasing CM overhead and slowing down cluster
maintenance operations (e.g peer addition, fix ups, data rebalance).
* Note: It is recommended to be set as true with high number of indexers
and buckets.
* Default: true
deferred_rest_api_update = <boolean>
* Only valid when 'mode=manager'
* If set to true (default), the manager responds to a REST API call from a source
peer immediately. It might defer part of the actions related to the call until
it completes already pending work.
* If set to false, the manager finishes all work for a received REST API call
and only then responds to the source peer. The response might be delayed
if the manager is busy with other work.
* Default: true
max_fixup_time_ms = <zero or positive integer>
* Only valid for 'mode=manager'.
* Specifies, in milliseconds, how long each fixup level runs before
short circuiting to continue to the next fixup level. This
introduces an upper-bound on each service level, and likewise
introduces an upper bound on the full service() call.
* This setting is useful for larger clusters that have lots of
buckets, where service() calls can consume a significant amount
of time blocking other operations.
* 0 denotes that there is no max fixup timer.
* Default: 1000
max_delayed_updates_time_ms = <zero or positive integer>
* Only valid for 'mode=manager'.
* Specifies, in milliseconds, how long cluster manager can continuously
serves the delayed jobs before quitting to run other jobs.
* This setting is useful for larger clusters that have a large number of
peer nodes and indexes, where customer manager could occasionally receive
thousands of REST APIs in a short period.
* Do not change this setting without first consulting with Splunk Support.
* 0 denotes that there is no limit to how long the delayed jobs thread
can run continuously.
* Default: 1000
primary_src_persist_secs = <zero or positive integer>
* Only valid for 'mode=manager'.
* For a warm bucket, this setting specifies the interval after the bucket's
latest time that a primary rebalance operation attempts to assign the primary
to the copy on the source peer node. Once the interval is exceeded,
the rebalance operation no longer considers the bucket's origin when
assigning its primary.
* For a hot bucket, a non-zero value causes the primary to always reside with the
source's hot bucket.
* Do not change this setting without first consulting with Splunk Support.
* If set to 0, the rebalance operation does not consider bucket origin
when assigning primaries, for both hot and warm buckets.
* Default: 604800 (1 week, 60 * 60 * 24 * 7 seconds)
cxn_timeout = <integer>
* Lowlevel timeout, in seconds, for establishing connection between
cluster nodes.
* Default: 60
send_timeout = <integer>
* Lowlevel timeout, in seconds, for sending data between cluster nodes.
* Default: 60
rcv_timeout = <integer>
* Lowlevel timeout, in seconds, for receiving data between cluster nodes.
* Default: 60
rep_cxn_timeout = <integer>
* Lowlevel timeout, in seconds, for establishing connection for replicating
data.
* Default: 5
rep_send_timeout = <integer>
* Only valid for 'mode=peer'
* Lowlevel timeout, in seconds, for sending replication slice data between
cluster nodes.
* This is a soft timeout. When this timeout is triggered on source peer,
it tries to determine if target is still alive. If it is still alive, it
reset the timeout for another 'rep_send_timeout interval' and continues. If
target has failed or cumulative timeout has exceeded the
'rep_max_send_timeout', replication fails.
* Default: 5
rep_rcv_timeout = <integer>
* Only valid for 'mode=peer'
* Lowlevel timeout, in seconds, for receiving acknowledgment data from peers.
* This is a soft timeout. When this timeout is triggered on source peer,
it tries to determine if target is still alive. If it is still alive,
it reset the timeout for another 'rep_send_timeout' interval and continues.
* If target has failed or cumulative timeout has exceeded
'rep_max_rcv_timeout', replication fails.
* Default: 10
search_files_retry_timeout = <integer>
* Timeout, in seconds, after which request for search files from a
peer is aborted.
* To make a bucket searchable, search specific files are copied from
another source peer with search files. If search files on source
peers are undergoing chances, it asks requesting peer to retry after
some time. If cumulative retry period exceeds the specified timeout,
the requesting peer aborts the request and requests search files from
another peer in the cluster that may have search files.
* Default: 600 (10 minutes)
re_add_on_bucket_request_error = <boolean>
* Valid only for 'mode=peer'.
* If set to true, peer re-adds itself to the cluster manager if
cluster manager returns an error on any bucket request. On re-add,
peer updates the manager with the latest state of all its buckets.
* If set to false, peer doesn't re-add itself to the cluster manager.
Instead, it updates the manager with those buckets that manager
returned an error.
* Default: false
decommission_search_jobs_wait_secs = <unsigned integer>
* Valid only for 'mode=peer'.
* Determines maximum time, in seconds, that a peer node waits for search
jobs to finish before it transitions to the down (or) 'GracefulShutdown'
state, in response to the 'splunk offline' (or)
'splunk offline --enforce-counts' command.
* Note: When using this setting, the 'decommission_search_jobs_wait_secs'
setting in the '[general]' stanza must remain set to its default value.
* You do not need to restart the cluster peer when making changes to
this setting. This setting reloads automatically.
* Default: 180 (3 minutes)
decommission_node_force_timeout = <seconds>
* Valid only for 'mode=peer' and during node offline operation
* The maximum time, in seconds, that a peer node waits for searchable copy
reallocation jobs to finish before it transitions to the down (or)
'GracefulShutdown' state.
* This period begins after the peer node receives a 'splunk offline' command
or its '/cluster/slave/control/control/decommission' REST endpoint
is accessed.
* This attribute is not applicable to the "--enforce-counts" version of the
"splunk offline" command
* Default: 300 seconds
decommission_force_finish_idle_time = <zero or positive integer>
* Valid only for 'mode=manager'.
* Time in minutes the manager waits before forcibly finishing the
decommissioning of a peer when there is no progress in the associated
fixup activity.
* A value of zero (0) means that the manager does not forcibly finish
decommissioning.
* Default: 0
rolling_restart = restart|shutdown|searchable|searchable_force
* Only valid for 'mode=manager'.
* Determines whether indexer peers restart or shutdown during a rolling
restart.
* If set to restart, each peer automatically restarts during a rolling
restart.
* If set to shutdown, each peer is stopped during a rolling restart,
and the customer must manually restart each peer.
* If set to searchable, the cluster attempts a best-effort to maintain
a searchable state during the rolling restart by reassigning primaries
from peers that are about to restart to other searchable peers, and
performing a health check to ensure that a searchable rolling restart is
possible.
* If set to 'searchable_force', the cluster performs a searchable
rolling restart, but overrides the health check and enforces
'decommission_force_timeout' and 'restart_inactivity_timeout'.
* If set to 'searchable' or 'searchable_force', scheduled searches
are deferred or run during the rolling restart based on the
'defer_scheduled_searchable_idx' setting in 'savedsearches.conf'.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: restart
searchable_rolling_peer_state_delay_interval = <zero or positive integer>
* Only valid for 'mode=manager'.
* Specifies an extra time interval, in seconds, during which the peer remains
in the ReassigningPrimaries state.
* Extending the amount of time the peer remains in the ReassigningPrimaries
state gives the peer more time to complete inflight searches and ingest
events.
* This also reduces the impact of incomplete searches and bucket corruption,
which can impede the searchable rolling restart process.
* Default: 60
searchable_rolling_site_down_policy = full|most|half
* Only valid for 'mode=manager' and is only used if 'multisite=true' and
'site_by_site=true'.
* Sets the policy for calculating the maximum number of peers in a site allowed
to shutdown at the same time during a searchable rolling restart.
* If set to 'full', the manager will allow an entire site to shutdown at once
if there are searchable copies of buckets available in at least 3 sites.
* If set to 'most', the manager will maintain a few peers in a site to act as
hot bucket streaming targets, and shutdown the other peers. At least one
peer is available as a streaming target, but there can be more depending on
the cluster's search factor. The 'most' policy attempts to speed up the
rolling restart more aggressively than 'half', at the expense of a longer
period fixing up replication and search factors afterwards.
* If set to 'half', the manager will maintain half the peers in a site to act
as hot bucket streaming targets, and shutdown the other peers.
* The maximum number of peers allowed down at one time is the smallest peer
count between 'percent_peers_to_restart' and
'searchable_rolling_site_down_policy'.
* Default: half
rolling_restart_condition = up|batch_adding|starting
* Only valid for 'mode=manager'.
* Determines the target peer status the manager waits for, when restarting
a peer during a rolling restart, before it restarts other peers.
* If set to 'up', the manager will wait for a restarting peer to reach the
'Up' status before restarting other peers. A peer reaches 'Up' status
when it has finished reporting all of its buckets to the manager. This
option will always respect 'percent_peers_to_restart'.
* If set to 'batch_adding', the manager will wait for a restarting peer to
reach the 'BatchAdding' status before restarting other peers. A peer
reaches 'BatchAdding' status when it is in the process of reporting its
buckets to the manager. This option will respect 'percent_peers_to_restart'
as long as the current restarting peer finishes adding before the next
restarting peer finishes shutting down, which is extremely likely.
* If set to 'starting', the manager will wait for a restarting peer to
reach the 'Starting' status before restarting other peers. A peer
reaches 'Starting' status when it first starts up and is in the process
of scanning its buckets on disk. This option is the fastest, but may
not always respect 'percent_peers_to_restart'.
* Default: batch_adding
site_by_site = <boolean>
* Only valid for 'mode=manager' and 'multisite=true'.
* If set to true, the manager restarts peers from one site at a time,
waiting for all peers from a site to restart before moving on to another
site, during a rolling restart.
* If set to false, the manager randomly selects peers to restart, from
across all sites, during a rolling restart.
* Default: true
decommission_force_timeout = <zero or positive integer>
* Only valid for 'rolling_restart=searchable_force'
* The amount of time, in seconds, the cluster manager waits for a
peer in primary decommission status to finish primary reassignment
and restart, during a searchable rolling restart with timeouts.
* Differs from 'decommission_force_finish_idle_time' in its default value
and its presence only during a searchable rolling restart with timeouts.
* If you set this parameter to 0, it is automatically reset
to default value.
* Maximum accepted value is 1800 (30 minutes).
* Default: 180 (3 minutes)
restart_inactivity_timeout = <zero or positive integer>
* Only valid for 'rolling_restart=searchable_force'
* The amount of time, in seconds, that the manager waits for a peer to
restart and rejoin the cluster before it considers the restart a failure
and proceeds to restart other peers.
* More specifically, the amount of time that the manager waits for a peer in
the 'Down' status to transition to 'BatchAdding' or 'Up' status.
* A value of zero (0) means that the manager waits indefinitely for a peer
to restart.
* Default: 600 (10 minutes)
rebalance_pipeline_batch_size = <integer>
* Valid only for 'mode=manager'.
* Valid only for 'searchable_rebalance=true'.
* The maximum number of buckets for a batch entering the excess bucket removal
phase of the rebalance pipeline.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: 60
rebalance_primary_failover_timeout = <zero or positive integer>
* Valid only for 'mode=manager'.
* Valid only for 'searchable_rebalance=true'.
* The maximum length of time, in seconds, that the manager waits for primacy to
be reassigned from the batch of excess buckets to other buckets.
* Default: 75
rebalance_newgen_propagation_timeout = <zero or positive integer>
* Valid only for 'mode=manager'.
* Valid only for 'searchable_rebalance=true'.
* The amount of time, in seconds, that the manager waits for the search heads to
get the newly committed generation after the discarded buckets' primacy has
been reassigned.
* Default: 60 (1 minute)
rebalance_search_completion_timeout = <integer>
* Valid only for 'mode=manager'.
* Valid only for 'searchable_rebalance=true'.
* The amount of time, in seconds, that the manager waits for older generation
searches on indexers to complete before removing any excess buckets.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: 180 (3 minute)
searchable_rebalance = <boolean>
* Valid only for 'mode=manager'.
* Controls whether searches can continue uninterrupted during data rebalancing.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: false
rep_max_send_timeout = <integer>
* Only valid for 'mode=peer'
* Maximum send timeout, in seconds, for sending replication slice
data between cluster nodes.
* On rep_send_timeout source peer determines if total send timeout has
exceeded 'rep_max_send_timeout'. If so, replication fails.
* If cumulative 'rep_send_timeout' exceeds 'rep_max_send_timeout',
replication fails.
* For a standalone indexer, changes to this setting are dynamically reloadable
and do not require a restart.
* For indexer clusters, changes to this setting trigger a rolling restart
of peer nodes.
* Default: 180 (3 minutes)
rep_max_rcv_timeout = <integer>
* Only valid for 'mode=peer'
* Maximum cumulative receive timeout, in seconds, for receiving
acknowledgment data from peers.
* On 'rep_rcv_timeout' source peer determines if total
receive timeout has exceeded 'rep_max_rcv_timeout'.
If so, replication fails.
* For a standalone indexer, changes to this setting are dynamically reloadable
and do not require a restart.
* For indexer clusters, changes to this setting trigger a rolling restart
of peer nodes.
* Default: 180 (3 minutes)
multisite = <boolean>
* Turns on the multisite feature for this manager.
* Make sure you set site parameters on the peers when you turn this to true.
* Default: false
replication_factor = <positive integer>
* Only valid for 'mode=manager'.
* Determines how many copies of rawdata are created in the cluster.
* Use 'site_replication_factor' instead of this in case 'multisite'
is turned on.
* Must be greater than 0.
* Default: 3
site_replication_factor = <comma-separated string>
* Only valid for 'mode=manager' and is only used if 'multisite=true'.
* This specifies the per-site replication policy for any given
bucket represented as a comma-separated list of per-site entries.
* Currently specified globally and applies to buckets in all
indexes.
* Each entry is of the form <site-id>:<positive integer> which
represents the number of copies to make in the specified site
* Valid site-ids include two mandatory keywords and optionally
specific site-ids from site1 to site63
* The mandatory keywords are:
- origin: Every bucket has a origin site which is the site of
the peer that originally created this bucket. The notion of
'origin' makes it possible to specify a policy that spans across
multiple sites without having to enumerate it per-site.
- total: The total number of copies needed for each bucket.
* When a site is the origin, it could potentially match both the
origin and a specific site term. In that case, the max of the
two is used as the count for that site.
* The total must be greater than or equal to sum of all the other
counts (including origin).
* The difference between total and the sum of all the other counts
is distributed across the remaining sites.
* Example 1: site_replication_factor = origin:2, total:3
Given a cluster of 3 sites, all indexing data, every site has 2
copies of every bucket ingested in that site and one rawdata
copy is put in one of the other 2 sites.
* Example 2: site_replication_factor = origin:2, site3:1, total:3
Given a cluster of 3 sites, 2 of them indexing data, every
bucket has 2 copies in the origin site and one copy in site3. So
site3 has one rawdata copy of buckets ingested in both site1 and
site2 and those two sites have 2 copies of their own buckets.
* Default: origin:2, total:3
search_factor = <positive integer>
* Only valid for 'mode=manager'.
* Determines how many buckets have index structures pre-built.
* Must be less than or equal to the 'replication_factor' setting and
greater than 0.
* Default: 2
site_search_factor = <comma-separated string>
* Only valid for 'mode=manager' and is only used if 'multisite=true'.
* This specifies the per-site policy for searchable copies for any
given bucket represented as a comma-separated list of per-site
entries.
* This is similar to the 'site_replication_factor' setting.
Please see that entry for more information on the syntax.
* Default: origin:1, total:2
available_sites = <comma-separated string>
* Only valid for 'mode=manager' and is only used if 'multisite=true'.
* This is a comma-separated list of all the sites in the cluster.
* If 'multisite=true' then 'available_sites' must be
explicitly set.
* Default: an empty string
forwarder_site_failover = <comma-separated string>
* Only valid for 'mode=manager' and is only used if 'multisite=true'.
* This is a comma-separated list of pair of sites, "site1:site2",
in the cluster.
* If 'multisite' is turned on 'forwarder_site_failover' must be
explicitly set.
* Default: an empty string
site_mappings = <comma-separated string>
* Only valid for 'mode=manager'.
* When you decommission a site, you must update this attribute so that the
origin bucket copies on the decommissioned site are mapped to a remaining
active site. This attribute maps decommissioned sites to active sites.
The bucket copies for which a decommissioned site is the origin site
are then replicated to the active site specified by the mapping.
* Used only if multisite is true and sites have been decommissioned.
* Each comma-separated entry is of the form
<decommissioned_site_id>:<active_site_id>
or default_mapping:<default_site_id>.
<decommissioned_site_id> is a decommissioned site and <active_site_id> is
an existing site,specified in the 'available_sites' setting.
For example, if available_sites=site1,site2,site3,site4 and you
decommission site2, you can map site2 to a remaining site such as site4,
like this: site2:site4 .
* If a site used in a mapping is later decommissioned, its previous mappings
must be remapped to an available site. For instance, if you have the
mapping site1:site2 but site2 is later decommissioned, you can remap
both site1 and site2 to an active site3 using the following replacement
mappings - site1:site3,site2:site3.
* Optional entry with syntax default_mapping:<default_site_id> represents the
default mapping, for cases where an explicit mapping site is not specified.
For example: default_mapping:site3 maps any decommissioned site to site3,
if they are not otherwise explicitly mapped to a site.
There can only be one such entry.
* Example 1: site_mappings = site1:site3,default_mapping:site4.
The cluster must include site3 and site4 in available_sites, and site1
must be decommissioned.
The origin bucket copies for decommissioned site1 is mapped to site3.
Bucket copies for any other decommissioned sites is mapped to site4.
* Example 2: site_mappings = site2:site3
The cluster must include site3 in available_sites, and site2 must be
decommissioned. The origin bucket copies for decommissioned site2 is
mapped to site3. This cluster has no default.
* Example 3: site_mappings = default_mapping:site5
The above cluster must include site5 in available_sites.
The origin bucket copies for any decommissioned sites is mapped onto
site5.
* Default: an empty string
constrain_singlesite_buckets = <boolean>
* Only valid for 'mode=manager' and is only used if multisite is true.
* Specifies whether the cluster keeps single-site buckets within one site
in multisite clustering.
* When this setting is "true", buckets in a single site cluster do not
replicate outside of their site. The buckets follow 'replication_factor'
'search factor' policies rather than 'site_replication_factor'
'site_search_factor' policies. This is to mimic the behavior of
single-site clustering.
* When this setting is "false", buckets in non-multisite clusters can
replicate across sites, and must meet the specified
'site_replication_factor' and 'site_search_factor' policies.
* Default: true
heartbeat_timeout = <positive integer>
* Only valid for 'mode=manager'.
* Specifies, in seconds, when the manager considers a peer down. After a
peer is down, the manager initiates fixup steps to replicate
buckets from the dead peer to its peers.
* Default: 60
access_logging_for_heartbeats = <boolean>
* Only valid for 'mode=manager'.
* Enables/disables logging to the splunkd_access.log file for peer
heartbeats.
* NOTE: you do not have to restart manager to set this config parameter.
Simply run the cli command on manager:
% splunk edit cluster-config -access_logging_for_heartbeats <<boolean>>
* Default: false (logging disabled)
restart_timeout = <positive integer>
* Only valid for 'mode=manager'.
* This is the amount of time, in seconds, the manager waits for a peer
to come back when the peer is restarted (to avoid the overhead of
trying to fixup the buckets that were on the peer).
* More specifically, the amount of time that the manager waits for a
peer in the 'Restarting' status to transition to the 'Down' status.
* Note that this only works with the offline command or if the peer
is restarted vi the UI.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: 60
streaming_replication_wait_secs = <positive integer>
* Only valid for 'mode=manager'.
* The amount of time, in seconds, that a peer node waits to restart after
receiving a restart request from the manager. During this period, the node
remains in the eRestartRequested state. This time allows the ongoing
replications on the peer to complete.
* Default: 60
quiet_period = <positive integer>
* Only valid for 'mode=manager'.
* This setting determines the amount of time, in seconds, that the manager is
quiet upon start-up.
* However, if peers are still registering themselves with the manager after
the initial quiet_period has elapsed, the manager continues to remain
quiet until all peers finish registering, up to a total quiet time not to
exceed 3x the specified 'quiet_period', including the initial quiet time.
* During the quiet time, the manager does not initiate any actions. At the end of
this period, the manager builds its view of the cluster based on the
registered information. It then starts normal operations.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: 60
reporting_delay_period = <positive integer>
* Only valid for 'mode=manager'.
* The acceptable amount of delay, in seconds, for reporting both unmet
search and unmet replication factors for newly created buckets.
* This setting helps provide more reliable cluster status reporting
by limiting updates to the specified granularity.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: 30
generation_poll_interval = <positive integer>
* How often, in seconds, the search head polls the manager for
generation information.
* This setting is valid only if 'mode=manager' or 'mode=searchhead'.
* This setting reloads automatically and does not require a restart.
* Default: 5
generation_max_staleness = <interval><unit>
* Search heads will ignore search generation information from a cluster
manager if it looks incomplete, either because it doesn't contain any
search peers, or because it consists of a subset of the search head's known
peers and at least one of the missing peers is currently joining
the cluster.
* This setting specifies how old a search head's own information can be before
the search head forcefully accepts new search generation information from a
cluster manager.
* When search performance is compromised by cluster manager restarts, increase
'generation_max_staleness' to the time it takes for all indexers to join the
cluster.
* This setting is valid only when 'mode=searchhead'.
* Default: 60s
max_peer_build_load = <integer>
* This is the maximum number of concurrent tasks to make buckets
searchable that can be assigned to a peer.
* Default: 2
max_peer_rep_load = <integer>
* This is the maximum number of concurrent non-streaming
replications that a peer can take part in as a target.
* Default: 5
max_peer_sum_rep_load = <integer>
* This is the maximum number of concurrent summary replications
that a peer can take part in as either a target or source.
* Default: 5
max_nonhot_rep_kBps = <integer>
* This is the maximum throughput (kB(Bytes)/s) for warm/cold/summary
replications on a specific source peer. Similar to forwarder's 'maxKBps'
setting in the limits.conf file.
* This setting throttles total bandwidth consumption for all
outgoing non-hot replication connections from a given source peer.
It does not throttle at the per-replication-connection, per-target
level.
* This setting can be updated without restart on the source peers
by using the command "splunk edit cluster-config" or by making the
corresponding REST call.
* If set to 0, signifies unlimited throughput.
* Default: 0
max_replication_errors = <integer>
* Only valid for 'mode=peer'.
* This is the maximum number of consecutive replication errors
(currently only for hot bucket replication) from a source peer
to a specific target peer. Until this limit is reached, the
source continues to roll hot buckets on streaming failures to
this target. After the limit is reached, the source no
longer rolls hot buckets if streaming to this specific target
fails. This is reset if at least one successful (hot bucket)
replication occurs to this target from this source.
* The special value of 0 turns off this safeguard; so the source
always rolls hot buckets on streaming error to any target.
* This setting is dynamically reloadable and does not require restart
of cluster peer.
* Default: 3
searchable_targets = <boolean>
* Only valid for 'mode=manager'.
* Tells the manager to make some replication targets searchable
even while the replication is going on. This only affects
hot bucket replication for now.
* Default: true
searchable_target_sync_timeout = <integer>
* Only valid for 'mode=peer'.
* If a hot bucket replication connection is inactive for this time,
in seconds, a searchable target flushes out any pending search
related in-memory files.
* Regular syncing - when the data is flowing through
regularly and the connection is not inactive - happens at a
faster rate (default of 5 secs controlled by
streamingTargetTsidxSyncPeriodMsec in indexes.conf).
* The special value of 0 turns off this timeout behavior.
* Default: 60
target_wait_time = <positive integer>
* Only valid for 'mode=manager'.
* Specifies the time, in seconds, that the manager waits for the
target of a replication to register itself before it services
the bucket again and potentially schedules another fixup.
* This setting is dynamically reloadable and does not require restart
of cluster manager.
* Default: 150 (2 minutes 30 seconds)
summary_wait_time = <positive integer>
* Only valid when 'mode=manager' and 'summary_replication=true'.
* Specifies the time, in seconds, that the manager waits before
scheduling fixups for a newly 'done' summary that transitioned
from 'hot_done'. This allows for other copies of the 'hot_done'
summary to also make their transition into 'done', avoiding
unnecessary replications.
* Default: 660 (11 minutes)
commit_retry_time = <positive integer>
* Only valid for 'mode=manager'.
* Specifies the interval, in seconds, after which, if the last
generation commit failed, the manager forces a retry. A retry is usually
automatically kicked off after the appropriate events. This is just
a backup to make sure that the manager does retry no matter what.
* Default: 300 (5 minutes)
percent_peers_to_restart = <integer between 0-100>
* Suggested percentage of maximum peers to restart for rolling-restart.
* Actual percentage may vary due to lack of granularity for smaller peer
sets.
* Regardless of setting, a minimum of 1 peer is restarted per round.
percent_peers_to_reload = <integer between 0-100>
* Only valid for mode=master
* Suggested percentage of maximum peers to reload for bundle push.
* Actual percentage may vary due to lack of granularity for smaller peer
sets.
* If set to 0, a minimum of 1 peer reloads the bundle per round.
max_peers_to_download_bundle = <positive integer>
* Only valid for 'mode=manager'
* Maximum no. of peers to simultaneously download the configuration bundle
from the manager, in response to the 'splunk apply cluster-bundle' command.
* When a peer finishes the download, the next waiting peer, if any, begins
its download.
* If set to 0, all peers try to download at once.
* Default: 5
precompress_cluster_bundle = <boolean>
* Only valid for 'mode=manager'.
* Whether or not the manager compresses the configuration bundle files before
it pushes them to peers.
* When set to "true", the manager compresses the configuration bundle, which
helps reduce network bandwidth consumption during the bundle push.
* Set this option to 'true' only when SSL compression is off. Otherwise, the
files will be compressed twice, which wastes CPU resources and does not save
network bandwidth. To turn off SSL compression, set
'allowSslCompression = false' in server.conf on the manager.
* Compressed bundles are denoted by the suffix ".bundle.gz". Uncompressed
bundles use the suffix ".bundle".
* Default: true
auto_rebalance_primaries = <boolean>
* Only valid for 'mode=manager'.
* Specifies if the manager should automatically rebalance bucket
primaries on certain triggers. Currently the only defined
trigger is when a peer registers with the manager. When a peer
registers, the manager redistributes the bucket primaries so the
cluster can make use of any copies in the incoming peer.
* Default: true
rebalance_primaries_execution_limit = <non-negative integer>
* DEPRECATED. Use the 'rebalance_primaries_execution_limit_ms' setting instead.
rebalance_primaries_execution_limit_ms = <non-negative integer>
* Only valid for 'mode=manager'.
* Specifies, in milliseconds, the maximum period for one execution
of the rebalance primary operation.
* This setting is useful for large clusters with large numbers of
buckets, to prevent the primary rebalance operation from blocking
other operations for significant amounts of time.
* The default value of 0 signifies auto mode. In auto mode, the cluster
manager uses the value of the 'service_interval' setting to determine the
maximum time for the operation.
* Default: 0
commit_generation_execution_limit_ms = <non-negative integer>
* Only valid for 'mode=manager'.
* Specifies, in milliseconds, the maximum period for one execution
of the committing pending generation.
* This setting is useful for large clusters with large numbers of
buckets, to prevent the commit-geenration operation from blocking
other operations for significant amounts of time.
* The default value of 0 signifies auto mode. In auto mode, the cluster
manager uses the value of the 'service_interval' setting to determine the
maximum time for the operation.
* If 'service_interval' is auto, the range of this value will be within the
range of 10ms and 25ms.
* Default: 0
idle_connections_pool_size = <integer>
* Only valid for 'mode=manager'.
* Specifies how many idle http(s) connections that should be kept
alive to reuse.
* Reusing connections improves the time it takes to send messages to peers
in the cluster.
* -1 corresponds to "auto", letting the manager determine the
number of connections to keep around based on the number of peers in the
cluster.
* Default: -1
use_batch_mask_changes = <boolean>
* Only valid for 'mode=manager'
* Specifies if the manager should process bucket mask changes in
batch or individually one by one.
* Set to 'false' when there are version 6.1 peers in the cluster for
backwards compatibility.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: true
service_jobs_msec = <positive integer>
* Only valid for 'mode=manager'.
* Max time, in milliseconds, that the cluster manager spends in servicing
finished jobs for each service call. Increase this if the 'metrics.log'
file has very high 'current_size' values.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: 100 (0.1 seconds)
summary_replication = true|false|disabled
* Valid for both 'mode=manager' and 'mode=peer'.
* Cluster Manager:
If set to true, summary replication is enabled.
If set to false, summary replication is disabled, but can be enabled
at runtime.
If set to disabled, summary replication is disabled. Summary replication
cannot be enabled at runtime.
* Peers:
If set to true or false, there is no effect. The indexer follows
whatever setting is on the Cluster Manager.
If set to disabled, summary replication is disabled. The indexer does
no scanning of summaries (increased performance during peers joing
the cluster for large clusters).
* Default: false (for both Cluster Manager and Peers)
rebalance_threshold = <number between 0.10 and 1.00>
* Only valid for 'mode=manager'.
* During rebalancing buckets amongst the cluster, this threshold is
used as a percentage to determine when the cluster is balanced.
* 1.00 is 100% indexers fully balanced.
* Default: 0.90
max_auto_service_interval = <positive integer>
* Only valid for 'mode=manager'.
* Only valid when 'service_interval' is in auto mode.
For example service_interval=0.
* Indicates the maximum value, in seconds, that service interval is
bounded by when the 'service_interval' is in auto mode. If the
previous service call took more than 'max_auto_service_interval'
seconds, the next service call runs after 'max_auto_service_interval'
seconds.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: 1
buckets_to_summarize = <primaries|primaries_and_hot|all>
* Only valid for 'mode=manager'.
* Determines which buckets are sent to '| summarize' searches (searches that
build report acceleration and data models).
* Set to "primaries" to apply only to primary buckets.
* Set to "primaries_and_hot" to also apply it to all hot searchable
buckets.
* Set to "all" to apply the search to all buckets.
* If "summary_replication' is enabled, then 'buckets_to_summarize' defaults
to "primaries_and_hot".
* Do not change this setting without first consulting with Splunk Support.
* Default: primaries
maintenance_mode = <boolean>
* Only valid for 'mode=manager'.
* To preserve the maintenance mode setting in case of manager
restart, the manager automatically updates this setting in the
etc/system/local/server.conf file whenever the user enables or disables
maintenance mode using CLI or REST.
* NOTE: Do not manually update this setting. Instead use CLI or REST
to enable or disable maintenance mode.
backup_and_restore_primaries_in_maintenance = <boolean>
* Only valid for 'mode=manager'.
* Determines whether the manager performs a backup/restore of bucket
primary masks during maintenance mode or rolling-restart of cluster peers.
* If set to true, restoration of primaries occurs automatically when the peers
rejoin the cluster after a scheduled restart or upgrade.
* Default: false
max_primary_backups_per_service = <zero or positive integer>
* Only valid for 'mode=manager'.
* For use with the 'backup_and_restore_primaries_in_maintenance' setting.
* Determines the number of peers for which the manager backs up primary
masks for each service call.
* The special value of 0 causes the manager to back up the primary masks for
all peers in a single service call.
* Default: 10
allow_default_empty_p4symmkey = <boolean>
* Only valid for 'mode=manager'.
* Affects behavior of manager during start-up, if 'pass4SymmKey'resolves
to the null string or the default password ("changeme").
* If set to true, the manager posts a warning but still launches.
* If set to false, the manager posts a warning and stops.
* Default: true
register_replication_address = <IP address or fully qualified machine/domain name>
* Only valid for 'mode=peer'.
* This is the address on which a peer is available for accepting
replication data. This is useful in the cases where a peer host machine
has multiple interfaces and only one of them can be reached by another
splunkd instance
register_forwarder_address = <IP address or fully qualified machine/domain name>
* Only valid for 'mode=peer'.
* This is the address on which a peer is available for accepting
data from forwarder.This is useful in the cases where a splunk host
machine has multiple interfaces and only one of them can be reached by
another splunkd instance.
register_search_address = <IP address, or fully qualified machine/domain name>
* Only valid for 'mode=peer'
* This is the address that advertises the peer to search heads.
This is useful in the cases where a splunk host machine has multiple
interfaces and only one of them can be reached by another splunkd
instance.
executor_workers = <positive integer>
* Only valid if 'mode=manager' or 'mode=peer'.
* Number of threads that can be used by the clustering thread pool.
* A value of 0 defaults to 1.
* Default: 10
* This setting reloads automatically and does not require a restart.
local_executor_workers = <positive integer>
* DEPRECATED.
manual_detention = on|on_ports_enabled|off
* Only valid for 'mode=peer'.
* Puts this peer node in manual detention.
* Default: off
allowed_hbmiss_count = <non-zero positive integer>
* Only valid for 'mode=peer'.
* Sets the count of number of heartbeat failures before the peer node
disconnects from the manager.
* Default: 3
buckets_per_addpeer = <non-negative integer>
* Only valid for 'mode=peer'.
* Controls the number of buckets for each add peer request.
* When a peer is added or re-added to the cluster, it sends the manager
information for each of its buckets. Depending on the number of buckets,
this could take a while. For example, a million buckets could require
more than a minute of the manager's processing time. To prevent the manager
from being occupied by this single task too long, you can use this setting to
split large numbers of buckets into several"batch-add-peer" requests.
* If it is invalid or non-existant, the peer uses the default setting instead.
* If it is set to 0, the peer sends only one request with all buckets
instead of batches.
* You do not need to restart the cluster peer when making changes to
this setting. This setting reloads automatically.
* Default: 1000
heartbeat_period = <non-zero positive integer>
* Only valid for 'mode=peer'.
* Controls the frequency the peer attempts to send heartbeats.
bucketsize_mismatch_strategy = smallest | largest
* Only valid for 'mode=manager'.
* This setting determines how the manager decides which target peer's bucket copy
is retained on the cluster when the source peer is not present at the time
that a hot bucket is rolled, and there is a bucket size mismatch between
the target peers
* When "largest" is selected, the largest copy of the bucket on any target
peer gets propagated to the other peers through fixups, overwriting all other
copies.
* When "smallest" is selected, the smallest copy of the bucket on any target
peer gets propagated to the other peers through fixups, overwriting all other
copies.
* Do not alter this value without contacting Splunk Support.
* Default: largest
remote_storage_upload_timeout = <non-zero positive integer>
* Only valid for 'mode=peer'.
* For a remote storage enabled index, this attribute specifies the interval
in seconds, after which target peers assume responsibility for
uploading a bucket to the remote storage, if they do not hear from
the source peer.
* This setting is dynamically reloadable and does not require restart
of cluster peer.
* Default: 60 (1 minute)
report_remote_storage_bucket_upload_to_targets = <boolean>
* Only valid for 'mode=peer'.
* For a remote storage enabled index, this attribute specifies whether
the source peer reports the successful bucket upload to target peers.
This notification is used by target peers to cancel their upload timers
and synchronize their bucket state with the uploaded bucket on remote
storage.
* Do not change the value from the default unless instructed by
Splunk Support.
* You do not need to restart the cluster manager when making changes to
this setting. This setting reloads automatically.
* Default: false
remote_storage_retention_period = <non-zero positive integer>
* Only valid for 'mode=manager'.
* The interval, in seconds, after which the manager checks buckets
in remote storage enabled indexes against the retention policy.
It then triggers freeze operations on the cluster peers as necessary.
* This setting also determines the time that the manager waits
following a restart before checking retention policy.
* For details on retention policies, examine the
'maxGlobalDataSizeMB' and 'frozenTimePeriodInSecs' settings.
* This setting is dynamically reloadable and does not require restart
of cluster manager.
* Default: 900 (15 minutes)
recreate_bucket_attempts_from_remote_storage = <positive integer>
* Only valid for 'mode=manager'.
* Controls the number of attempts the manager makes to recreate the
bucket of a remote storage enabled index on a random peer node
in these scenarios:
* Manager detects that the bucket is not present on any peers.
* A peer informs the manager about the bucket as part of the
re-creation of an index.
See recreate_index_attempts_from_remote_storage attribute.
* Re-creation of the bucket involves the following steps:
1. Manager provides a random peer with the bucket ID of the bucket that
needs to be recreated.
2. Peer fetches the metadata of the bucket corresponding to this
bucket ID from the remote storage.
3. Peer creates a bucket with the fetched metadata locally and informs
the manager that a new bucket has been added.
4. Manager initiates fix-ups to add the bucket on the necessary number
of additional peers to match the replication and search factors.
* If set to 0, disables the re-creation of the bucket.
* Default: 10
recreate_bucket_max_per_service = <positive integer>
* Only valid for 'mode=manager'.
* Only applies when using remote storage enabled indexes.
* Controls the maximum number of buckets that the cluster can recreate
during a service interval.
* Do not change the value from the default unless instructed by
Splunk Support.
* If set to 0, recreating buckets will go at full speed.
* Default: 20000
recreate_bucket_fetch_manifest_batch_size = <positive integer>
* Only valid for 'mode=manager'.
* Controls the maximum number of bucket IDs for which a peer
attempts to initiate a parallel fetch of manifests at a time
in the process of recreating buckets that have been
requested by the manager.
* The manager sends this setting to all the peers that are
involved in the process of recreating the buckets.
* Default: 50
recreate_index_attempts_from_remote_storage = <positive integer>
* Only valid for 'mode=manager'.
* Controls the number of attempts the manager makes to recreate
a remote storage enabled index on a random peer node when the manager
is informed about the index by a peer.
* Re-creation of an index involves the following steps:
1. Manager pushes a bundle either when it is ready for service or
when requested by the user.
2. Manager waits for the bundle to be applied successfully on the
peer nodes.
3. Manager requests that a random peer node provide it with the list
of newly added remote storage enabled indexes.
4. Manager distributes a subset of indexes from this list to
random peer nodes.
5. Each of those peer nodes fetches the list of bucket IDs for the
requested index from the remote storage and provides it
to the manager.
6. The manager uses the list of bucket IDs to recreate the buckets.
See recreate_bucket_attempts_from_remote_storage.
* If set to 0, disables the re-creation of the index.
* Default: 10
recreate_index_fetch_bucket_batch_size = <positive integer>
* Only valid for 'mode=manager'.
* Controls the maximum number of bucket IDs that the manager
requests a random peer node to fetch from remote storage as part of
a single transaction for a remote storage enabled index.
The manager uses the bucket IDs for re-creation of the index.
See the 'recreate_index_attempts_from_remote_storage' setting.
* Default: 2000
use_batch_remote_rep_changes = <boolean> or <positive integer>
* Only valid for 'mode=manager'.
* Specifies whether the manager processes bucket copy changes (to meet
replication_factor and search_factor) in batch or individually.
* Also controls the maximum number of bucket replications that are processed in
one replication batch.
* This is applicable to buckets belonging to
remote storage enabled indexes only.
* Do not change this setting without consulting with Splunk Support.
* This setting is dynamically reloadable and does not require restart
of cluster manager.
* If 'false' is specified, batching of buckets would be turned off
* If 'true' is specified, batching of buckets would be turned on, the maximum
number of buckets processed per batch would be the system default (1000)
* If 0 is specified, batching of buckets would be turned off
* If <any non zero positive integer> is specified, batching of buckets
would be turned on, and the maximum number of buckets processed per batch
would be the value of the integer specified
* Default: 1000
max_peer_batch_rep_load = <positive integer>
* Only valid for 'mode=manager'.
* This setting is applicable to buckets belonging to
remote storage enabled indexes only.
* Only valid when 'use_batch_remote_rep_changes=true'
* This setting specifies the maximum number of concurrent batch replications
that a peer node can take part in, as a source.
* Default: 5
enable_primary_fixup_during_maintenance = <boolean>
* Only valid for 'mode=manager'.
* Specifies whether the manager performs primary fixups during
maintenance mode. This gets overridden by searchable rolling restart.
* This setting is dynamically reloadable and does not require restart
of cluster manager.
* Default: true
freeze_during_maintenance = <boolean>
* Only valid for 'mode=manager'.
* Specifies whether the manager will tell peers to freeze buckets during
maintenance mode.
* This setting is dynamically reloadable and does not require restart
of cluster manager.
* Default: false
assign_primaries_to_all_sites = <boolean>
* Only valid for 'mode=manager' and 'multisite=true'
* Controls how the manager assigns bucket primary copies on a
multisite cluster.
* If set to 'true', the manager assigns a primary copy to each site
defined in 'available_sites', as well as site0.
* If set to 'false':
* The manager assigns a primary copy only to sites with a search head.
* Sites without search heads do not get primary copies.
* When a new site with a search head joins the cluster, or an existing
site attains its first search head, the cluster manager gradually
adds all buckets in the cluster to its fixup list to ensure that the
site will be populated with primaries.
* If a site loses its search heads, no action is taken to remove
existing primaries from the site.
* Setting this parameter to 'false' can significantly reduce the work of primary
assignments, especially if search heads are only on site0 and
search affinity is disabled.
* Default: false
log_bucket_during_addpeer = <boolean>
* Only valid for 'mode=manager'
* Controls the log level for bucket information during add-peer activities.
* If set to 'true', the manager logs bucket information to INFO level under
CMMaster componenet during add-peer.
* If set to 'false', the manager logs bucket information to DEBUG level under
CMMaster component during add-peer.
* Set to 'false' for large clusters with large numbers of buckets.
* Default: false
max_concurrent_peers_joining = <nonzero integer>
* Only valid for 'mode=manager'.
* Limits the number of peers that are allowed to join the cluster at one time.
* The peer reports its buckets to the cluster manager upon first establishing a
connection with the manager, and it finishes joining the cluster when all of
its buckets have been reported.
* Once this limit is hit, any remaining peers check at one second intervals
for an available slot to join the cluster.
* By limiting the number of peers that can join simultaneously, this setting
can facilitate faster restart for some peers, thus more quickly restoring
partial ingest to the cluster.
* Default: 10
enable_parallel_add_peer = <bool>
* Only valid for 'mode=manager'.
* Enables the cluster manager to accept and process multiple 'add peer' requests
in parallel.
* The upper limit of concurrent 'add peer' requests that the manager can handle is
limited by the 'max_concurrent_peers_joining setting'.
* When this feature is enabled, the largest recommended value for
'max_concurrent_peers_joining' is half the number of CPU cores of
the indexer. For example, if the indexer has 24 CPU cores, the largest
recommended value for 'max_concurrent_peers_joining' is 12.
* This setting is useful for clusters with large numbers of buckets
and large numbers of indexers. It also improves the responsiveness
of the cluster manager, helping to prevent unnecessary timeouts.
* Default: true
buckets_status_notification_batch_size = <positive integer>
* Only valid for 'mode=peer'.
* Controls the number of existing buckets IDs that the peer
reports to the manager every notify_scan_period seconds.
The manager then initiates fix-ups for these buckets.
* CAUTION: Do not modify this setting without guidance from
Splunk personnel.
* Default: 1000
notify_scan_period = <non-zero positive integer>
* Only valid for 'mode=peer'.
* Controls the frequency, in seconds, that the indexer handles
the following options:
1. summary_update_batch_size
2. summary_registration_batch_size
* CAUTION: Do not modify this setting without guidance from
Splunk personnel.
* Default: 10
notify_scan_min_period = <non-zero positive integer>
* Only valid for 'mode=peer'.
* Controls the highest frequency, in milliseconds, that the indexer
scans summary folders
for summary updates/registrations. The notify_scan_period temporarily
becomes notify_scan_min_period when there are more summary
updates/registration events to be processed but has been limited due to
either summary_update_batch_size or summary_registration_batch_size.
* CAUTION: Do not modify this setting without guidance from Splunk
personnel.
* Default: 10
notify_buckets_period = <non-zero positive integer>
* Only valid for 'mode=peer'.
* Controls the frequency, in milliseconds, that the indexer handles
buckets_status_notification_batch_size
* CAUTION: Do not modify this setting without guidance from
Splunk personnel.
* Default: 10
summary_update_batch_size = <non-zero positive integer>
* Only valid for 'mode=peer'.
* Controls the number of summary updates the indexer sends per batch to
the manager every notify_scan_period.
* CAUTION: Do not modify this setting without guidance from
Splunk personnel.
* Default: 10
summary_registration_batch_size = <non-zero positive integer>
* Only valid for 'mode=peer'.
* Controls the number of summaries that get asynchronously registered
on the indexer and sent as a batch to the manager every
notify_scan_period.
* Caution: Do not modify this setting without guidance from Splunk personnel.
* Default: 1000
enableS2SHeartbeat = <boolean>
* Only valid for 'mode=peer'.
* Splunk software monitors each replication connection for
presence of a heartbeat, and if the heartbeat is not seen for
's2sHeartbeatTimeout' seconds, it closes the connection.
* Default: true
s2sHeartbeatTimeout = <seconds>
* This specifies the global timeout value, in seconds, for monitoring
heartbeats on replication connections.
* Splunk software closes a replication connection if heartbeat is not seen
for 's2sHeartbeatTimeout' seconds.
* Replication source sends heartbeats every 30 seconds.
* Default: 600 (10 minutes)
throwOnBucketBuildReadError = <boolean>
* Valid only for 'mode=peer'.
* If set to true, index clustering peer throws an exception if it
encounters a journal read error while building the bucket for a new
searchable copy. It also throws all the search & other files generated
so far in this particular bucket build.
* If set to false, index clustering peer just logs the error and preserves
all the search & other files generated so far & finalizes them as it
cannot proceed further with this bucket.
* Default: false
cluster_label = <string>
* This specifies the label of the indexer cluster
warm_bucket_replication_pre_upload = <boolean>
* Valid only for 'mode=peer'.
* This setting applies to remote storage enabled indexes only.
* If set to true, the target peers replicate all warm bucket contents when necessary for
bucket-fixing if the source peer has not yet uploaded the bucket to remote storage.
* If set to false, the target peers never replicate warm bucket contents.
* In either case the target peers replicate metadata only, once the source peer uploads
the bucket to remote storage.
* Default: false
bucketsize_upload_preference = largest | smallest
* Valid only for 'mode=peer'.
* This setting applies to remote storage enabled indexes only.
* This setting determines the criteria a target peer uses when deciding whether to
overwrite a bucket copy uploaded to remote storage by another target peer. Target
peers never overwrite copies uploaded by a source peer.
* When "largest" is selected, the largest copy of the bucket on any target
peer gets uploaded.
* When "smallest" is selected, the smallest copy of the bucket on any target
peer gets uploaded.
* Note, this and "bucketsize_mismatch_strategy" should follow same scheme.
* Do not alter this value without contacting Splunk Support.
* Default: largest
upload_rectifier_timeout_secs = <unsigned integer>
* Valid only for 'mode=peer'.
* This setting applies to remote storage enabled indexes only.
* When a peer uploads a bucket copy to remote storage, it checks, after a ,
timeout based on the value of this setting, to determine whether another
peer overwrote the copy.
* Depending on the value of "bucketsize_upload_preference" it will determine
if the bucket needs to be re-uploaded.
* This setting controls the timeout that the peer waits before checking.
* Default: 2
[clustermanager:<stanza>]
* Only valid for 'mode=searchhead' when the search head is a part of
multiple clusters.
master_uri = <uri>
* DEPRECATED. Use the 'manager_uri' setting instead.
manager_uri = <uri>
* Only valid for 'mode=searchhead' when present in this stanza.
* URI of the cluster manager that this search head should connect to.
pass4SymmKey = <password>
* Secret shared among the nodes in the cluster to prevent any
arbitrary node from connecting to the cluster. If a search head
is not configured with the same secret as the manager,
it not be able to communicate with the manager.
* If it is not present here, the key in the clustering stanza is used.
If it is not present in the clustering stanza, the value in the general
stanza is used.
* Unencrypted passwords must not begin with "$1$", as this is used by
Splunk software to determine if the password is already encrypted.
* No default.
pass4SymmKey_minLength = <integer>
* The minimum length, in characters, that a 'pass4SymmKey' should be for a particular stanza.
* When you start the Splunk platform, if the 'pass4SymmKey' is shorter in length than
what you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* If you use the CLI to modify 'pass4SymmKey' to a value that is shorter than what
you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* Default: 12
site = <site-id>
* Specifies the site this search head belongs to for this particular manager
when multisite is enabled (see below).
* Valid values for site-id include site0 to site63.
* The special value "site0" disables site affinity for a search head in a
multisite cluster. It is only valid for a search head.
multisite = <boolean>
* Turns on the multisite feature for this manager_uri for the search head.
* Make sure the manager has the multisite feature turned on.
* Make sure you specify the site in case this is set to true. If no
configuration is found in the [clustermanager] stanza, the search head defaults
to any value for 'site' that might be defined in the [general]
stanza.
* Default: false
[replication_port://<port>]
# Configure Splunk to listen on a given TCP port for replicated data from
# another cluster member.
# If 'mode=peer' is set in the [clustering] stanza at least one
# 'replication_port' must be configured and not disabled.
disabled = <boolean>
* Set to true to disable this replication port stanza.
* Default: false
listenOnIPv6 = no|yes|only
* Toggle whether this listening port listens on IPv4, IPv6, or both.
* If not present, the setting in the [general] stanza is used.
acceptFrom = <network_acl> ...
* Lists a set of networks or addresses from which to accept connections.
* Separate multiple rules with commas or spaces.
* Each rule can be in one of the following formats:
1. A single IPv4 or IPv6 address (examples: "10.1.2.3", "fe80::4a3")
2. A Classless Inter-Domain Routing (CIDR) block of addresses
(examples: "10/8", "192.168.1/24", "fe80:1234/32")
3. A DNS name, possibly with a "*" used as a wildcard
(examples: "myhost.example.com", "*.splunk.com")
4. "*", which matches anything
* You can also prefix an entry with '!' to cause the rule to reject the
connection. The input applies rules in order, and uses the first one that
matches.
For example, "!10.1/16, *" allows connections from everywhere except
the 10.1.*.* network.
* Default: "*" (accept from anywhere)
[replication_port-ssl://<port>]
* This configuration is same as the [replication_port] stanza above,
but uses SSL.
disabled = <boolean>
* Set to true to disable this replication port stanza.
* Default: false
listenOnIPv6 = no|yes|only
* Toggle whether this listening port listens on IPv4, IPv6, or both.
* If not present, the setting in the [general] stanza is used.
acceptFrom = <network_acl> ...
* This setting is the same as the setting in the [replication_port] stanza.
serverCert = <path>
* Full path to file containing private key and server certificate.
* The <path> must refer to a PEM format file.
* No default.
sslPassword = <password>
* Server certificate password, if any.
* No default.
password = <password>
* DEPRECATED; use 'sslPassword' instead.
rootCA = <path>
* DEPRECATED; use '[sslConfig]/sslRootCAPath' instead.
* Full path to the root CA (Certificate Authority) certificate store.
* The <path> must refer to a PEM format file containing one or more root CA
certificates concatenated together.
* No default.
cipherSuite = <cipher suite string>
* If set, uses the specified cipher string for the SSL connection.
* Must specify 'dhFile' to enable any Diffie-Hellman ciphers.
* Default: The default can vary (See the cipherSuite setting in
the $SPLUNK_HOME/etc/system/default/server.conf file for the current default)
sslVersions = <versions_list>
* Comma-separated list of SSL versions to support.
* The versions available are "ssl3", "tls1.0", "tls1.1", and "tls1.2".
* The special version "*" selects all supported versions. The version "tls"
selects all versions tls1.0 or newer.
* If a version is prefixed with "-" it is removed from the list.
* SSLv2 is always disabled; "-ssl2" is accepted in the version list but
does nothing.
* When configured in FIPS mode, ssl3 is always disabled regardless
of this configuration.
* Default: The default can vary (See the sslVersions setting in
the $SPLUNK_HOME/etc/system/default/server.conf file for the current default)
ecdhCurves = <comma separated list of ec curves>
* ECDH curves to use for ECDH key negotiation.
* The curves should be specified in the order of preference.
* The client sends these curves as a part of Client Hello.
* The server supports only the curves specified in the list.
* Splunk software only supports named curves specified
by their SHORT names.
* The list of valid named curves by their short/long names can be obtained
by executing this command:
$SPLUNK_HOME/bin/splunk cmd openssl ecparam -list_curves
* e.g. ecdhCurves = prime256v1,secp384r1,secp521r1
* Default: The default can vary (See the 'ecdhCurves' setting in
the $SPLUNK_HOME/etc/system/default/server.conf file for the current default)
dhFile = <path>
* PEM format Diffie-Hellman parameter file name.
* DH group size should be no less than 2048bits.
* This file is required in order to enable any Diffie-Hellman ciphers.
* No default.
dhfile = <path>
* DEPRECATED; use 'dhFile' instead.
supportSSLV3Only = <boolean>
* DEPRECATED. SSLv2 is now always disabled. The exact set of SSL versions
allowed is now configurable by using the 'sslVersions' setting above.
useSSLCompression = <boolean>
* If true, enables SSL compression.
* Default: true
compressed = <boolean>
* DEPRECATED. Use 'useSSLCompression' instead.
* Used only if 'useSSLCompression' is not set.
requireClientCert = <boolean>
* Requires that any peer that connects to replication port has a certificate
that can be validated by certificate authority specified in rootCA.
* Default: false
allowSslRenegotiation = <boolean>
* In the SSL protocol, a client may request renegotiation of the connection
settings from time to time.
* Setting this to false causes the server to reject all renegotiation
attempts, breaking the connection. This limits the amount of CPU a
single TCP connection can use, but it can cause connectivity problems
especially for long-lived connections.
* Default: true
sslCommonNameToCheck = <commonName1>, <commonName2>, ...
* Optional.
* Check the common name of the client's certificate against this list of names.
* requireClientCert must be set to "true" for this setting to work.
* No default.
sslAltNameToCheck = <alternateName1>, <alternateName2>, ...
* Optional.
* Check the alternate name of the client's certificate against this list
of names.
* If there is no match, assume that Splunk is not authenticated against this
server.
* requireClientCert must be set to true for this setting to work.
* No default.
Introspection settings
[introspection:generator:disk_objects]
* For 'introspection_generator_addon', packaged with Splunk; provides the
data ("i-data") consumed, and reported on, by 'introspection_viewer_app'
(due to ship with a future release).
* This stanza controls the collection of i-data about: indexes; bucket
superdirectories (homePath, coldPath, ...); volumes; search dispatch
artifacts.
* On forwarders the collection of index, volumes and dispatch disk objects
is disabled.
acquireExtra_i_data = true | false
* If true, extra Disk Objects i-data is emitted; you can gain more insight
into your site, but at the cost of greater resource consumption both
directly (the collection itself) and indirectly (increased disk and
bandwidth utilization, to store the produced i-data).
* Please consult documentation for list of regularly emitted Disk Objects
i-data, and extra Disk Objects i-data, appropriate to your release.
* Default: false
collectionPeriodInSecs = <positive integer>
* Controls frequency of Disk Objects i-data collection; higher frequency
(hence, smaller period) gives a more accurate picture, but at the cost of
greater resource consumption both directly (the collection itself) and
indirectly (increased disk and bandwidth utilization, to store the
produced i-data).
* Default: 600 (10 minutes)
[introspection:generator:disk_objects__indexes]
* This stanza controls the collection of i-data about indexes.
* Inherits the values of 'acquireExtra_i_data' and 'collectionPeriodInSecs'
attributes from the 'introspection:generator:disk_objects' stanza, but
may be enabled/disabled independently of it.
* This stanza should only be used to force collection of i-data about
indexes on dedicated forwarders.
* Default: Data collection is disabled on universal forwarders and
enabled on all other installations.
[introspection:generator:disk_objects__volumes]
* This stanza controls the collection of i-data about volumes.
* Inherits the values of 'acquireExtra_i_data' and 'collectionPeriodInSecs'
attributes from the 'introspection:generator:disk_objects' stanza, but
may be enabled/disabled independently of it.
* This stanza should only be used to force collection of i-data about
volumes on dedicated forwarders.
* Default: Data collection is disabled on universal forwarders and
enabled on all other installations.
[introspection:generator:disk_objects__dispatch]
* This stanza controls the collection of i-data about search dispatch
artifacts.
* Inherits the values of 'acquireExtra_i_data' and 'collectionPeriodInSecs'
attributes from the 'introspection:generator:disk_objects' stanza, but
may be enabled/disabled independently of it.
* This stanza should only be used to force collection of i-data about
search dispatch artifacts on dedicated forwarders.
* Default: Data collection is disabled on universal forwarders and
enabled on all other installations.
[introspection:generator:disk_objects__fishbucket]
* This stanza controls the collection of i-data about:
$SPLUNK_DB/fishbucket, where per-input status of file-based
inputs is persisted.
* Inherits the values of 'acquireExtra_i_data' and 'collectionPeriodInSecs'
attributes from the 'introspection:generator:disk_objects' stanza, but may
be enabled/disabled independently of it.
[introspection:generator:disk_objects__bundle_replication]
* This stanza controls the collection of i-data about:
bundle replication metrics of distributed search
* Inherits the values of 'acquireExtra_i_data' and 'collectionPeriodInSecs'
attributes from the 'introspection:generator:disk_objects' stanza, but may
be enabled/disabled independently of it.
[introspection:generator:disk_objects__partitions]
* This stanza controls the collection of i-data about: disk partition space
utilization.
* Inherits the values of 'acquireExtra_i_data' and 'collectionPeriodInSecs'
attributes from the 'introspection:generator:disk_objects' stanza, but may
be enabled/disabled independently of it.
[introspection:generator:disk_objects__summaries]
* Introspection data about summary disk space usage. Summary disk usage
includes both data model and report summaries. The usage is collected
for each summaryId, locally at each indexer.
disabled = true | false
* If not specified, inherits the value from
[introspection:generator:disk_objects] stanza.
collectionPeriodInSecs = <positive integer>
* Controls frequency, in seconds, of Disk Objects - summaries
collection; higher frequency (hence, smaller period) gives a more accurate
picture, but at the cost of greater resource consumption directly
(the summaries collection itself);
it is not recommended for a period less than 15 minutes.
* If you enable summary collection, the first collection happens 5 minutes
after the Splunk instance is started. For every subsequent collection, this
setting is honored.
* If 'collectionPeriodInSecs' is smaller than 5 * 60, it resets to
30 minutes internally.
* Set to (N*300) seconds. Any remainder is ignored.
* Default: 1800 (30 minutes)
[introspection:generator:resource_usage]
* For 'introspection_generator_addon', packaged with Splunk; provides the
data ("i-data") consumed, and reported on, by 'introspection_viewer_app'
(due to ship with a future release).
* "Resource Usage" here refers to: CPU usage; scheduler overhead; main
(physical) memory; virtual memory; pager overhead; swap; I/O; process
creation (a.k.a. forking); file descriptors; TCP sockets; receive/transmit
networking bandwidth.
* Resource Usage i-data is collected at both hostwide and per-process
levels; the latter, only for processes associated with this SPLUNK_HOME.
* Per-process i-data for Splunk search processes include additional,
search-specific, information.
acquireExtra_i_data = true | false
* If set to true, extra Resource Usage i-data is emitted; you can gain
more insight into your site, but at the cost of greater resource
consumption both directly (the collection itself) and indirectly
(increased disk and bandwidth utilization, to store the produced i-data).
* Please consult documentation for list of regularly emitted Resource Usage
i-data, and extra Resource Usage i-data, appropriate to your release.
* Default: false
collectionPeriodInSecs = <positive integer>
* Controls frequency of Resource Usage i-data collection; higher frequency
(hence, smaller period) gives a more accurate picture, but at the cost of
greater resource consumption both directly (the collection itself) and
indirectly (increased disk and bandwidth utilization, to store the
produced i-data).
* Default: 600 (10 minutes) on Universal Forwarders, and 10 (1/6th of a minute)
on non-Universal Forwarders
[introspection:generator:resource_usage__iostats]
* This stanza controls the collection of i-data about: IO Statistics data
* "IO Statistics" here refers to: read/write requests; read/write sizes;
io service time; cpu usage during service
* IO Statistics i-data is sampled over the collectionPeriodInSecs
* Does not inherit the value of the 'collectionPeriodInSecs' attribute from the
'introspection:generator:resource_usage' stanza, and may be enabled/disabled
independently of it.
collectionPeriodInSecs = <positive integer>
* Controls interval of IO Statistics i-data collection; higher intervals
gives a more accurate picture, but at the cost of greater resource consumption
both directly (the collection itself) and indirectly (increased disk and
bandwidth utilization, to store the produced i-data).
* Default: 60 (1 minute)
[introspection:generator:kvstore]
* For 'introspection_generator_addon', packaged with Splunk Enterprise.
* "KV Store" here refers to: statistics information about KV Store process.
serverStatsCollectionPeriodInSecs = <positive integer>
* Controls frequency, in seconds, of KV Store server status collection
* Default: 27
operationStatsCollectionPeriodInSecs = <positive integer>
* Controls frequency, in seconds, of KV Store operation statistics collection (currentOp).
* Default: 60 seconds
collectionStatsCollectionPeriodInSecs = <positive integer>
* Controls frequency, in seconds, of KV Store db statistics collection.
* Default: 600 (10 minutes)
profilingStatsCollectionPeriodInSecs = <positive integer>
* Controls frequency, in seconds, of KV Store profiling data collection.
* Default: 5 seconds
rsStatsCollectionPeriodInSecs = <positive integer>
* Controls frequency, in seconds, of KV Store replica set stats collection
* Default: 60 seconds
[introspection:distributed-indexes]
* This stanza controls the collection of information for distributed indexes.
disabled = <boolean>
* Whether or not collection of introspection information on distributed
indexes is disabled.
* If set to "false", information on distributed indexes is collected.
* This provides additional insight into index usage at the cost of greater
resource consumption.
* Default: true
collectionPeriodInSecs = <positive integer>
* The frequency, in seconds, of distributed index data collection.
Shorter intervals provide more accurate results, at the cost of
greater resource consumption.
* Must be set between 300 (5 minutes) and 86400 (24 hours).
* Default: 3600 (60 minutes)
collectLocalIndexes = <boolean>
* This setting determines whether the search head retrieves index metadata,
such as current size and event count.
* In single-instance configurations, where the instance serves as both search head and indexer,
set the value to "true", so that the local index metadata is retrieved.
* In distributed search deployments, with separate search heads and indexers, set the
value to "false" to retrieve metadata only from indexes on the indexers.
* Default: false
Settings used to control commands started by Splunk
[commands:user_configurable] prefix = <path> * All non-internal commands started by splunkd are prefixed with this string, allowing for "jailed" command execution. * Should be only one word. In other words, commands are supported, but commands and arguments are not. * Applies to commands such as: search scripts, scripted inputs, SSL certificate generation scripts. (Any commands that are user-configurable). * Does not apply to trusted/non-configurable command executions, such as: splunk search, splunk-optimize, gunzip. * $SPLUNK_HOME is expanded. * No default. [app_backup] backup_path = <path> * Full path to the directory that contains configuration backups created by Splunk Enterprise. * For search head clusters, this directory resides on the deployer. * Default: $SPLUNK_HOME/var/backup
search head clustering configuration
[shclustering]
disabled = <boolean>
* Disables or enables search head clustering on this instance.
* When enabled, the captain needs to be selected via a
bootstrap mechanism. Once bootstrapped, further captain
selections are made via a dynamic election mechanism.
* When enabled, you must also specify the cluster member's own server
address / management URI for identification purpose. This can be
done in 2 ways: by specifying the 'mgmt_uri' setting individually on
each member or by specfing pairs of 'GUID, mgmt-uri' strings in the
servers_list attribute.
* Default: true
mgmt_uri = [ mgmt-URI ]
* The management URI is used to identify the cluster member's own address to
itself.
* Either 'mgmt_uri' or 'servers_list' is necessary.
* The 'mgmt_uri' setting is simpler to author but is unique for each member.
* The 'servers_list' setting is more involved, but can be copied as a
config string to all members in the cluster.
servers_list = [ <(GUID, mgmt-uri);>+ ]
* A semicolon separated list of instance GUIDs and management URIs.
* Each member uses its GUID to identify its own management URI.
adhoc_searchhead = <boolean>
* This setting configures a member as an ad-hoc search head; i.e., the member
does not run any scheduled jobs.
* Use the setting 'captain_is_adhoc_searchhead' to reduce compute load on the
captain.
* Default: false
no_artifact_replications = <boolean>
* Prevent this Search Head Cluster member to be selected as a target for
replications.
* This is an advanced setting, and not to be changed without proper
understanding of the implications.
* Default: false
precompress_artifacts = <boolean>
* Determines whether this search head cluster member compresses the
search artifacts before replicating them to other members.
* When set to "true", the search head compresses the artifacts
before replicating them to all other members.
This helps reduce network bandwidth consumption during artifact replications.
* Set this option to 'true' only when SSL compression is off on
each search head cluster member. To turn off SSL compression, set
'allowSslCompression = false' in the [sslconfig] stanza in server.conf
of each member.
* Default: true
captain_is_adhoc_searchhead = <boolean>
* This setting prohibits the captain from running scheduled jobs.
* The captain is dedicated to controlling the activities of the cluster,
but can also run adhoc search jobs from clients.
* Default: false
preferred_captain = <boolean>
* The cluster tries to assign captaincy to a member with
'preferred_captain=true'.
* Note that it is not always possible to assign captaincy to a member with
preferred_captain=true - for example, if none of the preferred members is
reachable over the network. In that case, captaincy might remain on a
member with preferred_captain=false.
* Default: true
prevent_out_of_sync_captain = <boolean>
* This setting prevents a node that could not sync config changes to current
captain from becoming the cluster captain.
* This setting takes precedence over the preferred_captain setting. For example,
if there are one or more preferred captain nodes but the nodes cannot
sync config changes with the current captain, then the current captain
retains captaincy even if it is not a preferred captain.
* This must be set to the same value on all members.
* Default: true
replication_factor = <positive integer>
* Determines how many copies of search artifacts are created in the cluster.
* This must be set to the same value on all members.
* Default: 3
pass4SymmKey = <password>
* Secret shared among the members in the search head cluster to prevent any
arbitrary instance from connecting to the cluster.
* All members must use the same value.
* If set in the [shclustering] stanza, it takes precedence over any setting
in the [general] stanza.
* Unencrypted passwords must not begin with "$1$", as this is used by
Splunk software to determine if the password is already encrypted.
* Default: The 'changeme' from the [general] stanza in the default the
server.conf file.
pass4SymmKey_minLength = <integer>
* The minimum length, in characters, that a 'pass4SymmKey' should be for a particular stanza.
* When you start the Splunk platform, if the 'pass4SymmKey' is shorter in length than
what you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* If you use the CLI to modify 'pass4SymmKey' to a value that is shorter than what
you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* Default: 12
async_replicate_on_proxy = <boolean>
* If the jobs/${sid}/results REST endpoint had to be proxied to a different
member due to missing local replica, this attribute automatically
schedules an async replication to that member when set to true.
* Default: true
master_dump_service_periods = <integer>
* If SHPMaster info is switched on in log.cfg, then captain statistics
are dumped in splunkd.log after the specified number of service periods.
Purely a debugging aid.
* Default: 500
long_running_jobs_poll_period = <integer>
* Long running delegated jobs are polled by the captain every
"long_running_jobs_poll_period" seconds to ascertain whether they are
still running, in order to account for potential node/member failure.
* Default: 600 (10 minutes)
scheduling_heuristic = <string>
* This setting configures the job distribution heuristic on the captain.
* There are currently two supported strategies: 'round_robin' or
'scheduler_load_based'.
* Default: 'scheduler_load_based'
id = <GUID>
* Unique identifier for this cluster as a whole, shared across all cluster
members.
* Default: Splunk software arranges for a unique value to be generated and
shared across all members.
cxn_timeout = <integer>
* Low-level timeout, in seconds, for establishing connection between
cluster members.
* Default: 60
send_timeout = <integer>
* Low-level timeout, in seconds, for sending data between search head
cluster members.
* Default: 60
rcv_timeout = <integer>
* Low-level timeout, in seconds, for receiving data between search head
cluster members.
* Default: 60
cxn_timeout_raft = <integer>
* Low-level timeout, in seconds, for establishing connection between search
head cluster members for the raft protocol.
* Default: 2
send_timeout_raft = <integer>
* Low-level timeout, in seconds, for sending data between search head
cluster members for the raft protocol.
* Default: 5
rcv_timeout_raft = <integer>
* Low-level timeout, in seconds, for receiving data between search head
cluster members for the raft protocol.
* Default: 5
rep_cxn_timeout = <integer>
* Low-level timeout, in seconds, for establishing connection for replicating
data.
* Default: 5
rep_send_timeout = <integer>
* Low-level timeout, in seconds, for sending replication slice data
between cluster members.
* This is a soft timeout. When this timeout is triggered on source peer,
it tries to determine if target is still alive. If it is still alive,
it reset the timeout for another rep_send_timeout interval and continues.
If target has failed or cumulative timeout has exceeded
rep_max_send_timeout, replication fails.
* Default: 5
rep_rcv_timeout = <integer>
* Low-level timeout, in seconds, for receiving acknowledgement data from
members.
* This is a soft timeout. When this timeout is triggered on source member,
it tries to determine if target is still alive. If it is still alive,
it reset the timeout for another rep_send_timeout interval and continues.
If target has failed or cumulative timeout has exceeded
the 'rep_max_rcv_timeout' setting, replication fails.
* Default: 10
rep_max_send_timeout = <integer>
* Maximum send timeout, in seconds, for sending replication slice data
between cluster members.
* On 'rep_send_timeout' source peer determines if total send timeout has
exceeded rep_max_send_timeout. If so, replication fails.
* If cumulative rep_send_timeout exceeds 'rep_max_send_timeout', replication
fails.
* Default: 600 (10 minutes)
rep_max_rcv_timeout = <integer>
* Maximum cumulative receive timeout, in seconds, for receiving acknowledgement
data from members.
* On 'rep_rcv_timeout' source member determines if total receive timeout has
exceeded 'rep_max_rcv_timeout'. If so, replication fails.
* Default: 600 (10 minutes)
log_heartbeat_append_entries = <boolean>
* If true, Splunk software logs the the low-level heartbeats between members in
splunkd_access.log file. These heartbeats are used to maintain the authority
of the captain authority over other members.
* Default: false
election_timeout_ms = <positive_integer>
* The amount of time, in milliseconds, that a member waits before
trying to become the captain.
* Note that modifying this value can alter the heartbeat period (See
election_timeout_2_hb_ratio for further details)
* A very low value of election_timeout_ms can lead to unnecessary captain
elections.
* Default: 60000 (1 minute)
election_timeout_2_hb_ratio = <positive_integer>
* The ratio between the election timeout, set in election_timeout_ms, and
the raft heartbeat period.
* Raft heartbeat period = election_timeout_ms / election_timeout_2_hb_ratio
* A typical ratio between 5 - 20 is desirable. Default is 12 to keep the
raft heartbeat period at 5s, i.e election_timeout_ms(60000ms) / 12
* This ratio determines the number of heartbeat attempts that would fail
before a member starts to timeout and tries to become the captain.
heartbeat_timeout = <positive integer>
* The amount of time, in seconds, that the captain considers a member down.
After a member is down, the captain initiates fixup steps to replicate
artifacts from the dead member to its peers.
* This heartbeat exchanges data between the captain and members, which helps in
maintaining the in-memory centralized state for all the cluster members.
* Note that this heartbeat is different from the Raft heartbeat described
in the 'election_timeout_2_hb_ratio' setting.
* Default: 60 (1 minute)
raft_rpc_backoff_time_ms = <positive integer>
* Provides a delay, in milliseconds, should a raft RPC request fail.
* This avoids rapid connection requests being made to unreachable peers.
* This setting should not normally be changed from the default.
* Default: 5000 (5 seconds)
access_logging_for_heartbeats = <boolean>
* Only valid on captain
* Enables/disables logging to the splunkd_access.log file for member heartbeats
* NOTE: you do not have to restart captain to set this config parameter.
Simply run the cli command on master:
% splunk edit shcluster-config -access_logging_for_heartbeats <<boolean>>
* Default: false (logging disabled)
restart_timeout = <positive integer>
* This is the amount of time the captain waits for a member to come
back when the instance is restarted (to avoid the overhead of
trying to fixup the artifacts that were on the peer).
quiet_period = <positive integer>
* Determines the amount of time, in seconds, for which a newly
elected captain waits for members to join. During this period the
captain does not initiate any fixups but instead waits for the
members to register themselves. Job scheduling and conf
replication still happen as usual during this time. At the end
of this time period, the captain builds its view of the cluster
based on the registered peers and starts normal
processing.
* Default: 60
max_peer_rep_load = <integer>
* This is the maximum number of concurrent replications that a
member can take part in as a target.
* Default: 5
target_wait_time = <positive integer>
* Specifies the time, in seconds, that the captain waits for the target
of a replication to register itself before it services the artifact again
and potentially schedules another fixup.
* Default: 150
manual_detention = on|off
* This property toggles manual detention on member.
* When a node is in manual detention, it does not accept new search jobs,
including both scheduled and ad-hoc searches. It also does not receive
replicated search artifacts from other nodes.
* Default: off
percent_peers_to_restart = <integer>
* The percentage of members to restart at one time during rolling restarts.
* Actual percentage may vary due to lack of granularity for smaller peer
sets regardless of setting, a minimum of 1 peer is restarted per
round.
* Valid values are between 0 and 100.
* CAUTION: Do not set this attribute to a value greater than 20%.
Otherwise, issues can arise during the captain election process.
rolling_restart_with_captaincy_exchange = <boolean>
* If this boolean is turned on, captain tries to exchange captaincy
with another node during rolling restart.
* If set to false, captain restarts and captaincy transfers to some
other node.
* Default: true
rolling_restart = restart|searchable|searchable_force
* Determines the rolling restart mode for a search head cluster.
* If set to restart, a rolling restart runs in classic mode.
* If set to searchable, a rolling restart runs in searchable (minimal
search disruption) mode.
* If set to searchable_force, the search head cluster performs a
searchable rolling restart, but overrides the health check
* Note: You do not have to restart any search head members to set this
parameter.
Run this CLI command from any member:
% splunk edit shcluster-config -rolling_restart
restart|searchable|searchable_force
* Default: restart (runs in classic rolling-restart mode)
decommission_search_jobs_wait_secs = <unsigned integer>
* The amount of time, in seconds, that a search head cluster member waits for
existing searches to complete before restarting.
* Applies only when rolling restart is triggered in searchable or
searchable_force mode
(i.e.'rolling_restart' is set to "searchable" or "searchable_force").
* Note: You do not have to restart search head members to set this parameter.
Run this CLI command from any member:
% splunk edit shcluster-config -decommission_search_jobs_wait_secs
<positive integer>
* Note: If this setting is used 'decommission_search_jobs_wait_secs'
defined in the '[general]' stanza should be left unchanged at it default value.
* Default: 180
register_replication_address = <IP address ormachine/domain name>
* This setting is the address on which a member is available for
accepting replication data. This is useful in the cases where a member
host machine has multiple interfaces and only one of them can be reached
by another splunkd instance.
* Can be an IP address, or fully qualified machine/domain name.
executor_workers = <positive integer>
* Number of threads that can be used by the search head clustering
threadpool.
* A value of 0 is interpreted as 1.
* Default: 10
heartbeat_period = <non-zero positive integer>
* Controls the frequency, in seconds, with which the member attempts
to send heartbeats to the captain.
* This heartbeat exchanges data between the captain and members, which
helps in maintaining the in-memory centralized state for all the
cluster members.
* Note that this heartbeat period is different from the Raft heartbeat
period in the election_timeout_2_hb_ratio setting.
* Default: 5
enableS2SHeartbeat = <boolean>
* Splunk software monitors each replication connection for presence of
a heartbeat.
* If the heartbeat is not seen for s2sHeartbeatTimeout seconds, it closes
the connection.
* Default: true
s2sHeartbeatTimeout = <integer>
* This specifies the global timeout, in seconds, value for monitoring
heartbeats on replication connections.
* Splunk software closes a replication connection if a heartbeat is not seen
for 's2sHeartbeatTimeout' seconds.
* Replication source sends a heartbeat every 30 seconds.
* Default: 600 (10 minutes)
captain_uri = [ static-captain-URI ]
* The management URI of static captain is used to identify the cluster
captain for a static captain.
election = <boolean>
* This is used to classify a cluster as static or dynamic (RAFT based).
* If set to "false", a static captain, which is used for DR situation.
* If set to "true", a dynamic captain election enabled through RAFT protocol.
mode = <member>
* Accepted values are captain and member, mode is used to identify
the function of a node in static search head cluster. Setting mode
as captain assumes it to function as both captain and a member.
#proxying related
sid_proxying = <boolean>
* Enable or disable search artifact proxying.
* Changing this affects the proxying of search results, and jobs feed
is not cluster-aware.
* Only for internal/expert use.
* Default: true
ss_proxying = <boolean>
* Enable or disable saved search proxying to captain.
* Changing this affects the behavior of Searches and Reports page
in Splunk Web.
* Only for internal/expert use.
* Default: true
ra_proxying = <boolean>
* Enable or disable saved report acceleration summaries proxying to captain.
* Changing this affects the behavior of report acceleration summaries
page.
* Only for internal/expert use.
* Default: true
alert_proxying = <boolean>
* Enable or disable alerts proxying to captain.
* Changing this impacts the behavior of alerts, and essentially make them
not cluster-aware.
* Only for internal/expert use.
* Default: true
csv_journal_rows_per_hb = <integer>
* Controls how many rows of CSV from the delta-journal are sent per hb
* Used for both alerts and suppressions
* Do not alter this value without contacting Splunk Support.
* Default: 10000
conf_replication_period = <integer>
* Controls how often, in seconds, a cluster member replicates
configuration changes.
* A value of 0 disables automatic replication of configuration changes.
* Default: 5
conf_replication_max_pull_count = <integer>
* Controls the maximum number of configuration changes a member
replicates from the captain at one time.
* A value of 0 disables any size limits.
* Default: 1000
conf_replication_max_push_count = <integer>
* Controls the maximum number of configuration changes a member
replicates to the captain at one time.
* A value of 0 disables any size limits.
* Default: 100
conf_replication_max_json_value_size = [<integer>|<integer>[KB|MB|GB]]
* Controls the maximum size of a JSON string element at any nested
level while parsing a configuration change from JSON representation.
* If a knowledge object created on a member has some string element
that exceeds this limit, the knowledge object is not replicated
to the rest of the search head cluster, and a warning that mentions
conf_replication_max_json_value_size is written to splunkd.log.
* If you do not specify a unit for the value, the unit defaults to bytes.
* The lower limit of this setting is 512KB.
* When increasing this setting beyond the default, you must take into
account the available system memory.
* Default: 15MB
conf_replication_include.<conf_file_name> = <boolean>
* Controls whether Splunk replicates changes to a particular type of *.conf
file, along with any associated permissions in *.meta files.
* Default: false
conf_replication_summary.whitelist.<name> = <whitelist_pattern>
* Files to be included in configuration replication summaries.
conf_replication_summary.blacklist.<name> = <blacklist_pattern>
* Files to be excluded from configuration replication summaries.
conf_replication_summary.concerning_file_size = <integer>
* Any individual file within a configuration replication summary that is
larger than this value (in MB) triggers a splunkd.log warning message.
* Default: 50
conf_replication_summary.period = <timespan>
* Controls how often configuration replication summaries are created.
* Default: 1m (1 minute)
conf_replication_purge.eligibile_count = <integer>
* Controls how many configuration changes must be present before any become
eligible for purging.
* In other words: controls the minimum number of configuration changes
Splunk software remembers for replication purposes.
* Default: 20000
conf_replication_purge.eligibile_age = <timespan>
* Controls how old a configuration change must be before it is eligible for
purging.
* Default: 1d (1 day).
conf_replication_purge.period = <timespan>
* Controls how often configuration changes are purged.
* Default: 1h (1 hour)
conf_replication_find_baseline.use_bloomfilter_only = <boolean>
* Controls whether or not a search head cluster only uses bloom filters to
determine a baseline, when it replicates configurations.
* Set to true to only use bloom filters in baseline determination during
configuration replication.
* Set to false to first attempt a standard method, where the search head
cluster captain interacts with members to determine the baseline, before
falling back to using bloom filters.
* Default: false
conf_deploy_repository = <path>
* Full path to directory containing configurations to deploy to cluster
members.
conf_deploy_staging = <path>
* Full path to directory where preprocessed configurations may be written
before being deployed cluster members.
conf_deploy_concerning_file_size = <integer>
* Any individual file within <conf_deploy_repository> that is larger than
this value (in MB) triggers a splunkd.log warning message.
* Default: 50
conf_deploy_precompress_bundles = <boolean>
* Determines whether or not the deployer compresses the configuration bundle
files before pushing them to search heads, which reduces network
bandwidth consumption.
* Set this option to "true" only when SSL compression is off. Otherwise, the
files will be compressed twice, which wastes CPU resources and does not save
network bandwidth. To turn off SSL compression, set
"allowSslCompression = false" in server.conf on the deployer.
* Default: true
conf_deploy_fetch_url = <URL>
* Specifies the location of the deployer from which members fetch the
configuration bundle.
* This value must be set to a <URL> in order for the configuration bundle to
be fetched.
* No default.
conf_deploy_fetch_mode = auto|replace|none
* Controls configuration bundle fetching behavior when the member starts up.
* When set to "replace", a member checks for a new configuration bundle on
every startup.
* When set to "none", a member does not fetch the configuration bundle on
startup.
* Regarding "auto":
* If no configuration bundle has yet been fetched, "auto" is equivalent
to "replace".
* If the configuration bundle has already been fetched, "auto" is
equivalent to "none".
* Default: replace
artifact_status_fields = <field> ...
* Give a comma separated fields to pick up values from status.csv and
info.csv for each search artifact.
* These fields are shown in the CLI/REST endpoint splunk list
shcluster-member-artifacts
* Default: user, app, label
encrypt_fields = <field> ...
* These are the fields that need to be re-encrypted when a Search Head
Cluster does its own first time run on syncing all members with a new s
plunk.secret key
* Give a comma separated fields as a triple elements
<conf-file>:<stanza-prefix>:<key elem>
* For matching all stanzas from a conf, leave the stanza-prefix
empty. For example: "server: :pass4SymmKey" matches all stanzas
with pass4SymmKey as key in server.conf
* Default: storage/passwords, secret key for clustering/shclustering,
server ssl config
enable_jobs_data_lite = <boolean>
*This is for memory reduction on the captain for Search head clustering,
leads to lower memory in captain while slaves send the artifacts
status.csv as a string.
* Default: true
shcluster_label = <string>
* This specifies the label of the search head cluster.
retry_autosummarize_or_data_model_acceleration_jobs = <boolean>
* Controls whether the captain tries a second time to delegate an
auto-summarized or data model acceleration job, if the first attempt to
delegate the job fails.
* Default: true
deployerPushThreads = <positive integer>|auto
* The maximum number of threads to use when performing a deployer bundle push
to target members.
* If set to "auto", the deployer auto-tunes the number of threads it uses
for a deployer bundle push. There will be one thread per target member.
* Default: 1
[replication_port://<port>]
############################################################################
# Configures the member to listen on a given TCP port for replicated data
# from another cluster member.
# At least one replication_port must be configured and not disabled.
############################################################################
disabled = <boolean>
* Set to true to disable this replication port stanza.
* Default: false
listenOnIPv6 = no|yes|only
* Toggle whether this listening port listens on IPv4, IPv6, or both.
* If not present, the setting in the [general] stanza is used.
acceptFrom = <network_acl> ...
* Lists a set of networks or addresses from which to accept connections.
* Separate multiple rules with commas or spaces.
* Each rule can be in one of the following formats:
1. A single IPv4 or IPv6 address (examples: "10.1.2.3", "fe80::4a3")
2. A Classless Inter-Domain Routing (CIDR) block of addresses
(examples: "10/8", "192.168.1/24", "fe80:1234/32")
3. A DNS name, possibly with a "*" used as a wildcard
(examples: "myhost.example.com", "*.splunk.com")
4. "*", which matches anything
* You can also prefix an entry with '!' to cause the rule to reject the
connection. The input applies rules in order, and uses the first one that
matches.
For example, "!10.1/16, *" allows connections from everywhere except
the 10.1.*.* network.
* Default: "*" (accept from anywhere)
[replication_port-ssl://<port>]
* This configuration is the same as the replication_port stanza, but uses SSL.
disabled = <boolean>
* Set to true to disable this replication port stanza.
* Default: false
listenOnIPv6 = no|yes|only
* Toggle whether this listening port listens on IPv4, IPv6, or both.
* Default: The setting in the [general] stanza
acceptFrom = <network_acl> ...
* This setting is the same as the setting in the [replication_port] stanza.
serverCert = <path>
* Full path to file containing private key and server certificate.
* The <path> must refer to a PEM format file.
* No default.
sslPassword = <password>
* Server certificate password, if any.
* No default.
password = <password>
* DEPRECATED; use 'sslPassword' instead.
* Used only if 'sslPassword' is not set.
rootCA = <path>
* DEPRECATED; use '[sslConfig]/sslRootCAPath' instead.
* Used only if '[sslConfig]/sslRootCAPath' is not set.
* Full path to the root CA (Certificate Authority) certificate store.
* The <path> must refer to a PEM format file containing one or more root CA
certificates concatenated together.
* No default.
cipherSuite = <cipher suite string>
* If set, uses the specified cipher string for the SSL connection.
* If not set, uses the default cipher string.
* provided by OpenSSL. This is used to ensure that the server does not
accept connections using weak encryption protocols.
supportSSLV3Only = <boolean>
* DEPRECATED. SSLv2 is now always disabled. The exact set of SSL versions
allowed is now configurable via the "sslVersions" setting above.
useSSLCompression = <boolean>
* If true, enables SSL compression.
* Default: false
compressed = <boolean>
* DEPRECATED; use 'useSSLCompression' instead.
* Used only if 'useSSLCompression' is not set.
requireClientCert = <boolean>
* Requires that any peer that connects to replication port has a certificate
that can be validated by certificate authority specified in rootCA.
* Default: false
allowSslRenegotiation = <boolean>
* In the SSL protocol, a client may request renegotiation of the connection
settings from time to time.
* Setting this to false causes the server to reject all renegotiation
attempts, breaking the connection. This limits the amount of CPU a
single TCP connection can use, but it can cause connectivity problems
especially for long-lived connections.
* Default: true
KV Store configuration
[kvstore]
disabled = <boolean>
* Set to true to disable the KV Store process on the current server. To
completely disable KV Store in a deployment with search head clustering or
search head pooling, you must also disable KV Store on each individual
server.
* Default: false
port = <port>
* Port to connect to the KV Store server.
* Default: 8191
replicaset = <replset>
* Replicaset name.
* Default: splunkrs
distributedLookupTimeout = <seconds>
* This setting has been removed, as it is no longer needed.
shutdownTimeout = <integer>
* Time, in seconds, to wait for a clean shutdown of the KV Store. If this time
is reached after signaling for a shutdown, KV Store is forcibly terminated
* Default: 100
initAttempts = <integer>
* The maximum number of attempts to initialize the KV Store when starting
splunkd.
* Default: 300
replication_host = <host>
* The host name to access the KV Store.
* This setting has no effect on a single Splunk instance.
* When using search head clustering, if the "replication_host" value is not
set in the [kvstore] stanza, the host you specify for
"mgmt_uri" in the [shclustering] stanza is used for KV
Store connection strings and replication.
* In search head pooling, this host value is a requirement for using KV
Store.
* This is the address on which a kvstore is available for accepting
remotely.
verbose = <boolean>
* Set to true to enable verbose logging.
* Default: false
verboseLevel = <nonnegative integer>
* When verbose logging is enabled specify verbose level for logging
from 0 to 5, where 5 is the most verbose.
* Default: 2
dbPath = <path>
* Path where KV Store data is stored.
* Changing this directory after initial startup does not move existing data.
The contents of the directory should be manually moved to the new
location.
* Default: $SPLUNK_DB/kvstore
storageEngine = mmapv1 | wiredTiger
* The storage engine that KV Store uses to manage its data.
* "mmapv1" will be deprecated for KV Store. Where possible, use the "wiredTiger" option,
which is more performant.
* When you upgrade the Splunk platform, you get the option to upgrade the KV Store
storage engine. If you choose not to, you can upgrade the engine later using the
'splunk migrate kvstore-storage-engine' CLI command.
* Default: mmapv1
storageEngineMigration = <boolean>
* Whether or not you can migrate the KV Store storage engine on this instance.
* Migrating the storage engine means changing the engine from 'mmap' to
the newer 'wiredTiger'.
* If you set this to "true", the instance lets you migrate the engine,
depending on the following scenarios:
* If this instance is standalone, you can migrate the engine during
an upgrade by enabling this setting with a "true" value as part of
the upgrade process and answering affirmatively when that process
prompts you, or after an upgrade by using the
'splunk migrate kvstore-storage-engine' CLI command.
* If the instance is a search head cluster member, this setting is ignored.
Perform a migration using the '/services/shcluster/captain/kvmigrate/start'
REST endpoint.
* A value of "false" prevents the storage engine from being migrated on this instance.
* Default: false
oplogSize = <integer>
* The size of the replication operation log, in megabytes, for environments
with search head clustering or search head pooling.
In a standalone environment, 20% of this size is used.
* After the KV Store has created the oplog for the first time, changing this
setting does NOT affect the size of the oplog. A full backup and restart
of the KV Store is required.
* Do not change this setting without first consulting with Splunk Support.
* Default: 1000 (1GB)
replicationWriteTimeout = <integer>
* The time to wait, in seconds, for replication to complete while saving KV
store operations. When the value is 0, the process never times out.
* Used for replication environments (search head clustering or search
head pooling).
* Default: 1800 (30 minutes)
clientConnectionTimeout = <positive integer>
* The time, in seconds, to wait while attempting a connection to the KV Store
before the attempt times out.
* Default: 10
clientSocketTimeout = <positive integer>
* The time, in seconds, to wait while attempting to send or receive on a
socket before the attempt times out.
* Default: 300 (5 minutes)
clientConnectionPoolSize = <positive integer>
* The maximum number of active client connections to the KV Store.
* When the number of active connections exceeds this value, KV Store will
reject new connection attempts until at least one active connection closes.
* Do not change this setting without first consulting with Splunk Support.
* Default: 500
caCertFile = <path>
* DEPRECATED; use '[sslConfig]/sslRootCAPath' instead.
* Used only if 'sslRootCAPath' is not set.
* Full path to a CA (Certificate Authority) certificate(s) PEM format file.
* If specified, it is used in KV Store SSL connections and
authentication.
* Only used when Common Criteria is enabled (SPLUNK_COMMON_CRITERIA=1)
or FIPS is enabled (i.e. SPLUNK_FIPS=1).
* NOTE: Splunk plans to submit Splunk Enterprise for Common Criteria
evaluation. Splunk does not support using the product in Common
Criteria mode until it has been certified by NIAP. See the "Securing
Splunk Enterprise" manual for information on the status of Common
Criteria certification.
* Default: $SPLUNK_HOME/etc/auth/cacert.pem
caCertPath = <filepath>
* DEPRECATED; use '[sslConfig]/sslRootCAPath' instead.
serverCert = <filepath>
* A certificate file signed by the signing authority specified above by
caCertPath.
* In search head clustering or search head pooling, the certificates at
different members must share the same 'subject'.
* The Distinguished Name (DN) found in the certificate's subject, must
specify a non-empty value for at least one of the following attributes:
Organization (O), the Organizational Unit (OU) or the
Domain Component (DC).
* Only used when Common Criteria is enabled (SPLUNK_COMMON_CRITERIA=1)
or FIPS is enabled (i.e. SPLUNK_FIPS=1).
* NOTE: Splunk plans to submit Splunk Enterprise for Common Criteria
evaluation. Splunk does not support using the product in Common
Criteria mode until it has been certified by NIAP. See the "Securing
Splunk Enterprise" manual for information on the status of Common
Criteria certification.
sslKeysPath = <filepath>
* DEPRECATED; use 'serverCert' instead.
* Used only when 'serverCert' is empty.
sslPassword = <password>
* Password of the private key in the file specified by 'serverCert' above.
* Must be specified if FIPS is enabled (i.e. SPLUNK_FIPS=1), otherwise, KV
Store is not available.
* Only used when Common Criteria is enabled (SPLUNK_COMMON_CRITERIA=1)
or FIPS is enabled (i.e. SPLUNK_FIPS=1).
* NOTE: Splunk plans to submit Splunk Enterprise for Common Criteria
evaluation. Splunk does not support using the product in Common
Criteria mode until it has been certified by NIAP. See the "Securing
Splunk Enterprise" manual for information on the status of Common
Criteria certification.
* No default.
sslKeysPassword = <password>
* DEPRECATED; use 'sslPassword' instead.
* Used only when 'sslPassword' is empty.
sslCRLPath = <filepath>
* Certificate Revocation List file.
* Only used when Common Criteria is enabled (SPLUNK_COMMON_CRITERIA=1)
or FIPS is enabled (i.e. SPLUNK_FIPS=1).
* NOTE: Splunk plans to submit Splunk Enterprise for Common Criteria
evaluation. Splunk does not support using the product in Common
Criteria mode until it has been certified by NIAP. See the "Securing
Splunk Enterprise" manual for information on the status of Common
Criteria certification.
* Optional.
* Default: empty string (no Revocation List)
modificationsReadIntervalMillisec = <integer>
* How often, in milliseconds, to check for modifications to
KV Store collections in order to replicate changes for distributed
searches.
* Default: 1000 (1 second)
modificationsMaxReadSec = <integer>
* Maximum time interval KVStore can spend while checking for modifications
before it produces collection dumps for distributed searches.
* Default: 30
initialSyncMaxFetcherRestarts = <positive integer>
* Specifies the maximum number of query restarts an oplog fetcher can perform
before failing the ongoing Initial Sync attempt.
* Increasing this value might help in dynamic deployments with very large
KV Store databases where Initial Sync might take a long time.
* NOTE: This setting should be changed only if you have been asked to set it by
a Splunk Support engineer. It might increase KV Store cluster failover time.
* Default: 0
delayShutdownOnBackupRestoreInProgress = <boolean>
* Whether or not splunkd should delay a shutdown if a KV Store backup or restore
operation is in progress.
* If set to "true", splunkd waits until either the running backup/restore operation
completes, or 'splunkd_stop_timeout' seconds have elapsed since it received
the shutdown request.
* NOTE: Setting this to "true" might delay splunkd shutdown for several minutes,
depending on the amount of data that KV Store uses and the value of
'splunkd_stop_timeout'.
* Default: false
percRAMForCache = <positive integer>
* The percentage of total system memory that KV store can use.
* Value can range from 5 to 50, inclusive.
* If less than 1 GB of system memory is present, only 256 MB of cache will be used.
* If you have less than 256 MB of system memory, you cannot use KVStore with wiredTiger.
* Changing this value can affect performance on KV store,
Splunk Enterprise apps that use KV store, and KV store lookups.
For more information, search the Splunk documentation for "KV store troubleshooting tools".
* If you are not using the WiredTiger storage engine, Splunk Enterprise ignores this setting.
* Default: 15
Indexer Discovery configuration
[indexer_discovery] pass4SymmKey = <password> * Security key shared between manager node and forwarders. * If specified here, the same value must also be specified on all forwarders connecting to this manager. * Unencrypted passwords must not begin with "$1$", as this is used by Splunk software to determine if the password is already encrypted. pass4SymmKey_minLength = <integer> * The minimum length, in characters, that a 'pass4SymmKey' should be for a particular stanza. * When you start the Splunk platform, if the 'pass4SymmKey' is shorter in length than what you specify with this setting, the platform warns you and advises that you change the pass4SymKey. * If you use the CLI to modify 'pass4SymmKey' to a value that is shorter than what you specify with this setting, the platform warns you and advises that you change the pass4SymKey. * Default: 12 polling_rate = <integer> * A value between 1 to 10. This value affects the forwarder polling frequency to achieve the desired polling rate. The number of connected forwarders is also taken into consideration. * The formula used to determine effective polling interval, in Milliseconds, is: (number_of_forwarders/polling_rate + 30 seconds) * 1000 * Default: 10 indexerWeightByDiskCapacity = <boolean> * If set to true, it instructs the forwarders to use weighted load balancing. In weighted load balancing, load balancing is based on the total disk capacity of the target indexers, with the forwarder streaming more data to indexers with larger disks. * The traffic sent to each indexer is based on the ratio of: indexer_disk_capacity/total_disk_capacity_of_indexers_combined * Default: false
Cascading Replication Configuration
[cascading_replication]
pass4SymmKey = <password>
* Security key shared between indexers participating in cascading replication.
* The same value must be specified on all indexers participating in cascading replication.
* Unencrypted passwords must not begin with "$1$", as this is used by
Splunk software to determine if the password is already encrypted.
* Empty passwords will not be accepted.
* Default: None
pass4SymmKey_minLength = <integer>
* The minimum length, in characters, that a 'pass4SymmKey' should be for a particular stanza.
* When you start the Splunk platform, if the 'pass4SymmKey' is shorter in length than
what you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* If you use the CLI to modify 'pass4SymmKey' to a value that is shorter than what
you specify with this setting, the platform warns you and advises that you
change the pass4SymKey.
* Default: 12
max_replication_threads = <integer>
* Maximum threads used for replicating metadata and payload to search peers.
* If set to "auto", the peer auto-tunes the number of threads it uses for
cascading replication.
* If the peer has 3 or fewer CPUs, it allocates 2 threads.
* If the peer has 4-7 CPUs, it allocates up to '# of CPUs - 2' threads.
* If the peer has 8-15 CPUs, it allocates up to '# of CPUs - 3' threads.
* If the peer has 16 or more CPUs, it allocates up to
'# of CPUs - 4' threads.
* Maximum accepted value for this setting is 16.
* Default: auto
max_replication_jobs = <integer>
* Maximum jobs used for replicating metadata and payload to search peers.
* Default: 5
cascade_replication_plan_reap_interval = <interval>
* The interval at which the cascade replication plans are reaped.
* The interval can be specified as a string for minutes, seconds, hours, days.
For example: 60s, 1m, 1h, 1d etc.
* Maximum accepted value is 5h
* Default: 1h
cascade_replication_plan_age = <interval>
* The age of the cascade replication plan when it gets reaped.
* The interval can be specified as a string for minutes, seconds, hours, or days.
For example: 60s, 1m, 1h, 1d etc.
* Maximum accepted value is 24h
* Default: 8h
cascade_replication_plan_fanout = auto|<positive integer>
* Number of receivers that each sender replicates to at a time.
* If set to auto, Splunk automatically calculates an optimal fanout, based on
the maximum number of replication threads, as determined by the
'max_replication_threads' setting under [cascading_replication] in server.conf.
* If set to an integer, the integer must be no greater than the number of cluster
peers, or, in the case of multisite clustering, no greater than the least number
of peers on any one site.
* Default: auto
cascade_replication_plan_topology = size_balanced
* Topology used for building a cascading plan.
* When set to size_balanced, receivers are evenly distributed among senders.
Senders on the same layer have same or similar number of receivers.
* Default: size_balanced
cascade_replication_plan_select_policy = random
* Policy for deciding which receivers the senders pick.
* When set to random, receivers are randomly picked.
* Default: random
Node level authentication
[node_auth] signatureVersion = <comma-separated list> * A list of authentication protocol versions that nodes of a Splunk deployment use to authenticate to other nodes. * Each version of node authentication protocol implements an algorithm that specifies cryptographic parameters to generate authentication data. * Nodes may only communicate using the same authentication protocol version. * For example, if you set "signatureVersion = v1,v2" on one node, that node sends and accepts authentication data using versions "v1" and "v2" of the protocol, and you must also set "signatureVersion" to one of "v1", "v2", or "v1,v2" on other nodes for those nodes to mutually authenticate. * For higher levels of security, set 'signatureVersion' to "v2". * Default: v1,v2
Cache Manager Configuration
[cachemanager]
max_concurrent_downloads = <unsigned integer>
* The maximum number of buckets that can be downloaded simultaneously from
external storage
* Default: 8
max_concurrent_uploads = <unsigned integer>
* The maximum number of buckets that can be uploaded simultaneously to external
storage.
* Default: 8
eviction_policy = <string>
* The name of the eviction policy to use.
* Current options: lru, clock, random, lrlt, noevict, lruk
* Do not change the value from the default unless instructed by
Splunk Support.
* Default: lru
enable_eviction_priorities = <boolean>
* When requesting buckets, search peers can give hints to the Cache Manager
about the relative importance of buckets.
* When enabled, the Cache Manager takes the hints into consideration; when
disabled, hints are ignored.
* Default: true
eviction_padding = <positive integer>
* Specifies the additional space, in megabytes, beyond 'minFreeSpace' that the
cache manager uses as the threshold to start evicting data.
* If free space on a partition falls below
('minFreeSpace' + 'eviction_padding'), then the cache manager tries to evict
data from remote storage enabled indexes.
* Default: 5120 (~5GB)
max_cache_size = <positive integer>
* Specifies the maximum space, in megabytes, per partition, that the cache can
occupy on disk. If this value is exceeded, the cache manager starts
evicting buckets.
* A value of 0 means this setting is not used to control cache eviction.
Eviction will instead be based on the sum of 'minFreeSpace' and 'eviction_padding'
settings, which limits the size of the partition that the cache resides on.
* Default: 0
persist_pending_upload_from_external = <bool>
* Currently not supported. This setting is related to a feature that is
still under development.
* Specifies whether the information of the buckets that have been uploaded
to remote storage can be serialized to disk or not.
* When set to true, this information is serialized to disk and
the bucket is deemed to be on remote storage.
* Otherwise, the bucket is deemed to be not on remote storage and
bucket is then uploaded to remote storage.
* Default: true
persistent_id_set_remove_min_sync_secs = <unsigned integer>
* Currently not supported. This setting is related to a feature that is
still under development.
* Cache manager persists the set of objects that are
no longer pending upload to the remote storage based
on when the previous set of changes were persisted to disk.
* This setting controls the interval from the last persist time that
cache manager waits to persist the current set of changes to disk.
* Default: 5
enable_open_on_stale_object = <bool>
* Currently not supported. This setting is related to a feature that is
still under development.
* Specifies whether the buckets with stale files can be opened for search.
* When set to true, these buckets can be opened for search.
Otherwise, searches are not allowed to open these buckets.
* Default: true
local_delete_summary_metadata_ttl = <unsigned integer>
* Currently not supported. This setting is related to a feature that is
still under development.
* The local copy of a bucket needs to be synced with the copy in remote
storage only when the bucket switches primaries.
* However in certain experimental modes of operation the delete journals
in the remote storage could be mutated without an update to the local copy.
* Similarly, accelerated summaries in remote storage could be updated without
an update to the local copy.
* This setting is meant for use in such modes. The Cache manager will make
a best effort to invalidate the local delete journals and summary
metadata files periodically.
* The period will be controlled by this ttl. A value of 0 will disable
this behavior
* Default: 0
hotlist_recency_secs = <unsigned integer>
* When a bucket is older than this value, it becomes eligible for eviction.
Buckets younger than this value are evicted only if there are no older
buckets eligible for eviction.
* For the purpose of determining recency, the age of a bucket is calculated by
subtracting the time of the bucket's most recent event data from the current time.
* For example, if the current time (expressed in UTC epoch time) is 1567891234 and
the bucket is named db_1567809123_1557891234_10_8A21BEE9-60D4-436B-AA6D-21B68F631A8B,
thus indicating that the time of the most recent event in the bucket is 1567809123,
then the bucket's age, in seconds, is 82111 (~23 hours).
* Ensure that the cache is of sufficient size to handle the value of this setting.
Otherwise, cache eviction cannot function optimally. In other words, do not
configure this setting to a size that will cause the cache to retain a quantity of
buckets that approach or exceed the size of the cache based on this setting
alone.
* Also, consider the amount of data you're ingesting and the needs of
the types of searches you run. As a best practice, start with a fairly low value
for this setting and adjust over time.
* For example, if the cache size is 100 GB and you typically add 10 GB of new buckets to
the indexer in a 24 hour period, setting this to 172800 (48 hours) would mean that
the cache manager will try to keep those 20 GB of recent buckets in the cache all the time.
* This setting can be overridden on a per-index basis in indexes.conf.
* Default: 86400 (24 hours)
hotlist_bloom_filter_recency_hours = <unsigned integer>
* When a bucket's non-journal and non-tsidx files (such as bloomfilter files)
are older than this value, those files become eligible for eviction. Bloomfilter
and associated files younger than this value are evicted only if there are
no older files eligible for eviction.
* The recency of a bloomfilter file is based on its bucket's recency and is calculated
in the same manner described for hotlist_recency_secs.
* This setting works in concert with hotlist_recency_secs which is designed to be
configured for a shorter age. If hotlist_recency_secs leads to the eviction of a
bucket, the bloomfilter and associated files will continue to remain in the cache
until they reach the age configured by this setting. Thus, the bucket will remain
in cache, but without its journal and tsidx files.
* This setting can be overridden on a per-index basis in indexes.conf.
* Default: 360 (15 days)
evict_on_stable = <boolean>
* When the source peer completes upload of a bucket to remote storage, it notifies the
target peers so that they can evict any local copies of the bucket.
* When set to true, each target peer evicts its local copy, if any, upon such notification.
* When set to false, each target peer continues to store its local copy, if any, until its
cache manager eventually evicts the bucket according to its cache eviction policy.
* Default: false
max_file_exists_retry_count = <unsigned integer>
* The cache manager retries its check on whether the file exists on
remote storage when the check fails due to network errors until
the retry count exceeds this setting.
* Default: 5
access_logging = <boolean>
* Enables/disables logging to the splunkd_access.log file for cachemanager requests.
* Default: false
cache_usage_collection_interval_minutes = <positive integer>
* Currently not supported. This setting is related to a feature that is
still under development.
* Interval at which cache usage information is reported to metrics.log.
* The cache usage logging reports cache usage, in bytes, broken down by
cache type (bid, dma, ra, metrics), index and by time range. The time
bins are defined by the setting 'cache_usage_collection_time_bins'.
* A value of 0 will disable this feature.
* Do not use a value less than 10 (minutes). Doing so can
affect performance.
* Hot buckets are not managed by the cache manager and not reflected
in the log messages.
* Default: 10
cache_usage_collection_time_bins = <positive integer list>
* Currently not supported. This setting is related to a feature that is
still under development.
* This setting is used when 'cache_usage_collection_interval_minutes' is
non-zero. See the 'cache_usage_collection_interval_minutes' section for
more information.
* This comma-separated list of integers, representing days, are boundaries
to the time ranges to which the cache usage is broken down and reported.
There is an implicit bin, 0, that represents all data more recent than the
first non-zero value. The highest value specified will represent all data
older than that value.
* For example, using the default "1, 3, 7, 14, 30, 60, 90", cache usage will
collect the size of buckets whose latest-time (endEpoch) into the following
bins: 0 (future-1d), 1 (1d-3d), 3 (3d-7d), 7 (7d-15d), 15 (15d-30d),
30 (30d-60d), 60 (60d-90d), 90 (90d and older).
* Default: 1, 3, 7, 15, 30, 60, 90
cache_usage_collection_per_index = <boolean>
* Currently not supported. This setting is related to a feature that is
still under development.
* Enables the reporting cache usage information by index.
* This setting is used when 'cache_usage_collection_interval_minutes' is
non-zero. See the 'cache_usage_collection_interval_minutes' section for
more information.
* Default: false
batch_registration = <boolean>
* This setting enables/disables batch registration of buckets upon startup of indexer.
* If this setting is disabled, then when an indexer starts up, its cache manager registers
each index bucket individually. This can slow the startup process. If an indexer is
experiencing long startup durations, enable this setting to register buckets in batches.
* The size of each batch of buckets is set with 'batch_registration_size'.
* Default: true
batch_registration_size = <unsigned integer>
* This setting specifies the size of each batch of buckets that are
registered.
* This setting is used when 'batch_registration' is enabled.
* Use the default value unless instructed otherwise by Splunk Support.
* Default: 5000
cache_upload_backoff_sleep_secs = <unsigned_integer>
* This setting specifies the interval, in seconds, that the cache manager waits to
retry an upload to the remote store after encountering a 4xx HTTP error.
* A value of 0 causes the cache manager to continue retrying the upload without
performing a backoff.
* Default: 60
max_known_remote_absent_summaries = <unsigned_integer>
* This setting specifies the maximum number of frozen (absent) summaries that the
cache manager maintains in a list.
* The list of frozen summaries helps the cache manager to avoid making calls to the
remote store that could result in an HTTP 404 "not found" error. By increasing the
limit, you decrease the likelihood of such calls, while potentially using more
memory in the process.
* When this value is reached, the cache manager deletes the oldest frozen
summaries from the list.
* Default: 200000 (200K)
Raft Statemachine configuration
[raft_statemachine]
disabled = <boolean>
* Set to true to disable the raft statemachine.
* This feature require search head clustering to be enabled.
* Any consensus replication among search heads use this feature.
* Default: true
replicate_search_peers = <boolean>
* Add/remove search-server request is applied on all members
of a search head cluster, when this value to set to true.
* Require a healthy search head cluster with a captain.
[watchdog]
disabled = <boolean>
* Disables thread monitoring functionality.
* Any thread that has been blocked for more than 'responseTimeout' milliseconds
is logged to $SPLUNK_HOME/var/log/watchdog/watchdog.log
* Default: false.
responseTimeout = <decimal>
* Maximum time, in seconds, that a thread can take to respond before the
watchdog logs a 'thread blocked' incident.
* The minimum value for 'responseTimeout' is 0.1.
* If you set 'responseTimeout' to lower than 0.1, the setting uses the minimum
value instead.
* Default: 8
actions = <actions_list>
* A comma-separated list of actions that execute sequentially when a blocked
thread is encountered.
* Currently, the only available actions are 'pstacks', 'script' and 'bulletin'.
* 'pstacks' enables call stack generation for a blocked thread.
* Call stack generation gives the user immediate information on the potential
bottleneck or deadlock.
* The watchdog saves each call stack in a separate file in
$SPLUNK_HOME/var/log/watchdog with the following file name format:
wd_stack_<pid>_<thread_name>_%Y_%m_%d_%H_%M_%S.%f_<uid>.log.
* 'script' executes specified script.
* 'bulletin' shows a message on the web interface.
* NOTE: This setting should be used only during troubleshooting, and if you have
been asked to set it by a Splunk Support engineer. It might degrade
performance by increasing CPU and disk usage.
* Default: empty list (no action executed)
actionsInterval = <decimal>
* The timeout, in seconds, that the watchdog uses while tracing a blocked
thread. The watchdog executes each action every 'actionsInterval' seconds.
* The minimum value for 'actionsInterval' is 0.01.
* If you set 'actionsInterval' to lower than 0.01, the setting uses the minimum
value instead.
* NOTE: Very small timeout may have impact performance by increasing CPU usage.
Splunk may be also slowed down by frequently executed action.
* Default: 1
pstacksEndpoint = <boolean>
* Enables pstacks endpoint at /services/server/pstacks
* Endpoint allows ad-hoc pstacks generation of all running threads.
* This setting is ignored if 'watchdog' is not enabled.
* NOTE: This setting should be used only during troubleshooting and only if you
have been explicitly asked to set it by a Splunk Support engineer.
* Default: true
usePreloadedPstacks = <boolean>
* Use preloaded wrapper to enable pstacks.
* NOTE: This setting should be changed only during troubleshooting and only if you
have been explicitly asked to disable it by a Splunk Support engineer.
* Default: true
[watchdog:timeouts]
reaperThread = <decimal>
* Maximum time, in seconds, that a reaper thread can take to respond before the
watchdog logs a 'thread blocked' incident.
* The minimum value for 'reaperThread' is 0.1.
* If you set 'reaperThread' to lower than 0.1, the setting uses the minimum
value instead.
* This value is used only for threads dedicated to clean up dispatch directories
and search artifacts.
* Default: 30
[watchdogaction:pstacks]
* Setting under this stanza are ignored if 'pstacks' is not enabled in the
'actions' list.
* NOTE: Change these settings only during troubleshooting, and if you have
been asked to set it by a Splunk Support engineer. It can affect performance
by increasing CPU and disk usage.
dumpAllThreads = <boolean>
* Determines whether or not the watchdog saves stacks of all monitored threads
when it encounters a blocked thread.
* If you set 'dumpAllThreads' to true, the watchdog generates call stacks for
all threads, regardless of thread state.
* Default: true
stacksBufferSizeOrder = <unsigned integer>
* Controls the maximum number of call stacks an internal queue can hold.
* The watchdog uses the internal queue to temporarily store a call stack between
the time the watchdog generates the call stack and the time it saves the call
stack to a file.
* Increase the value of this setting if you see gaps in stack files due to high
frequency of call stack generation. This might occur when, for example, you
set 'stacksBufferSizeOrder' to a very low value, or if the number of threads
is high.
* This number must be in the range 1 to 16.
* The watchdog uses this value to calculate the real size of the buffer, whose
value must be a power of 2. For example, if 'stackBufferSizeOrder' is 4, the
size of the buffer is 4 ^ 2, or 16.
* CAUTION: Setting to too low a value can cause dropped call stacks, and too
high a value can cause increased memory consumption.
* Default: 14
maxStacksPerBlock = <unsigned integer>
* Maximum number of stacks that the watchdog can generate for a blocked thread.
* If you set 'dumpAllThreads' to true, the watchdog generates call stacks for
all threads.
* If the blocked thread starts responding again, the count of stacks that the
watchdog has generated resets to zero.
* If another thread blockage occurs, the watchdog begins generating stacks
again, up to 'maxStacksPerBlock' stacks.
* When set to 0, an unlimited number of stacks will be generated.
list.
* Default: 60
batchStacksThreshold = <unsigned integer>|auto
* The timeout, in milliseconds, after which the watchdog generates a new call stack file.
* This setting controls the batching up of call stacks when saving them to files, and can
decrease the number of files the watchdog creates.
* When set to 0, batching is disabled.
* When set to 'auto', Splunk Enterprise determines the best frequency to create new
call stack files.
* Default: auto
[watchdogaction:script]
* Setting under this stanza are ignored if 'script' is not enabled in the
'actions' list.
* NOTE: Change these settings only during troubleshooting, and if you have
been asked to set it by a Splunk Support engineer. It can affect performance
by increasing CPU and disk usage.
path = <string>
* The path to the script to execute when the watchdog triggers the action.
* If you do not set 'path', the watchdog ignores the action.
* No default.
useShell = <boolean>
* If set to true, the script runs from the OS shell
("/bin/sh -c" on UNIX, "cmd.exe /c" on Windows)
* If set to false, the program will be run directly without attempting to
expand shell metacharacters.
* Default: false
forceStop = <boolean>
* Whether or not the watchdog forcefully stops an active watchdog action script
when a blocked thread starts to respond.
* Use this setting when, for example, the watchdog script has internal logic
that controls its lifetime and must run without interruption.
* Default: false
forceStopOnShutdown = <boolean>
* If you set this setting to "true", the watchdog forcefully stops active
watchdog scripts upon receipt of a shutdown request.
* Default: true
Parallel Reduce Configuration
[parallelreduce] pass4SymmKey = <password> * DEPRECATED. The setting is no longer required. pass4SymmKey_minLength = <integer> * The minimum length, in characters, that a 'pass4SymmKey' should be for a particular stanza. * When you start the Splunk platform, if the 'pass4SymmKey' is shorter in length than what you specify with this setting, the platform warns you and advises that you change the pass4SymKey. * If you use the CLI to modify 'pass4SymmKey' to a value that is shorter than what you specify with this setting, the platform warns you and advises that you change the pass4SymKey. * Default: 12 @ @ @ @ @ @ # @@INCLUDED AS WITH_CLOUD ## Do not remove scsTokenScriptPath = <string> * The path to the platform extension that retrieves SCS access tokens from Hashicorp Vault. * Default: /usr/local/bin/get_scs_tokens.sh @ @[bucket_catalog_service] @ @uri = <uri> @* Points to the tenant bucket catalog service. @* Required. @* Currently, only HTTP is supported by the service. @* Example: <scheme>://<hostname>:<port>/<tenantId>/<bucket_catalog_path> @ @[cache_manager_service] @ @uri = <uri> @* Points to the cache manager service. @* Required. @ @ping_enabled = <boolean> @* Currently not supported. This setting is related to a feature that is @ still under development. @* Enables "ping" keep-alive transactions to the Cache Manager Service. @* Default: true @ @timeout.ping = <unsigned integer> @* Currently not supported. This setting is related to a feature that is @ still under development. @* Sets the ping timeout, in milliseconds, to use when interacting with the @ Cache Manager Service. @* Default: 30000 @ @timeout.connect = <unsigned integer> @* Currently not supported. This setting is related to a feature that is @ still under development. @* Sets the connection timeout, in milliseconds, to use when connecting to the @ Cache Manager Service. @* Default: 5000 @ @timeout.read = <unsigned integer> @* Currently not supported. This setting is related to a feature that is @ still under development. @* Sets the read timeout, in milliseconds, to use when interacting with the @ Cache Manager Service. @* Default: 60000 @ @timeout.write = <unsigned integer> @* Currently not supported. This setting is related to a feature that is @ still under development. @* Sets the write timeout, in milliseconds, to use when interacting with the @ Cache Manager Service. @* Default: 60000
Remote Storage of Search Artifacts Configuration
[search_artifact_remote_storage]
disabled = <boolean>
* Currently not supported. This setting is related to a feature that is
still under development.
* Optional.
* Specifies whether or not search artifacts should be stored remotely.
* Splunkd does not clean up artifacts from remote storage. Set up cleanup
separately with the remote storage provider.
* Default: true
path = <path on server>
* The path attribute points to the remote storage location where
artifacts reside.
* The format for this attribute is: <scheme>://<remote-location-specifier>
* The "scheme" identifies a supported external storage system type.
* The "remote-location-specifier" is an external system-specific string for
identifying a location inside the storage system.
* These external systems are supported:
* Object stores that support AWS's S3 protocol. These use the scheme "s3".
For example, "path=s3://mybucket/some/path".
* This is a required setting. If you do not set the path, the search artifact
remote storage feature is disabled.
* No default.
upload_archive_format = [none|tar.lz4]
* Creates a tarball so that the entire artifact can be stored as a single object
on the remote storage.
* This can reduce time to upload and artifact when the remote storage has a high
seek penalty and the search artifact contains more than 100 individual files
* Default : none
S3 specific settings
remote.s3.header.<http-method-name>.<header-field-name> = <String>
* Optional.
* Enable server-specific features, such as reduced redundancy, encryption,
and so on, by passing extra HTTP headers with the REST requests.
* The <http-method-name> can be any valid HTTP method. For example, GET,
PUT, or ALL, for setting the header field for all HTTP methods.
* Example: remote.s3.header.PUT.x-amz-storage-class = REDUCED_REDUNDANCY
remote.s3.access_key = <String>
* Optional.
* Specifies the access key to use when authenticating with the remote storage
system supporting the S3 API.
* If not specified, the indexer looks for these environment variables:
AWS_ACCESS_KEY_ID or AWS_ACCESS_KEY (in that order).
* If the environment variables are not set and the indexer is running on EC2,
the indexer attempts to use the access key from the IAM role.
* No default.
remote.s3.secret_key = <String>
* Optional.
* Specifies the secret key to use when authenticating with the remote storage
system supporting the S3 API.
* If not specified, the indexer looks for these environment variables:
AWS_SECRET_ACCESS_KEY or AWS_SECRET_KEY (in that order).
* If the environment variables are not set and the indexer is running on EC2,
the indexer attempts to use the secret key from the IAM role.
* No default.
remote.s3.list_objects_version = v1|v2
* The AWS S3 Get Bucket (List Objects) Version to use.
* See AWS S3 documentation "GET Bucket (List Objects) Version 2" for details.
* Default: v1
remote.s3.signature_version = v2|v4
* Optional.
* The signature version to use when authenticating with the remote storage
system supporting the S3 API.
* For 'sse-kms' server-side encryption scheme, you must use
signature_version=v4.
* Default: v4
remote.s3.auth_region = <String>
* Optional
* The authentication region to use for signing requests when interacting with
the remote storage system supporting the S3 API.
* Used with v4 signatures only.
* If unset and the endpoint (either automatically constructed or explicitly
set with remote.s3.endpoint setting) uses an AWS URL
(for example, https://s3-us-west-1.amazonaws.com), the instance attempts
to extract the value from the endpoint URL (for example, "us-west-1"). See
the description for the remote.s3.endpoint setting.
* If unset and an authentication region cannot be determined, the request
will be signed with an empty region value.
* No default.
remote.s3.use_delimiter = true | false
* Optional.
* Specifies whether a delimiter (currently "guidSplunk") should be
used to list the objects that are present on the remote storage.
* A delimiter groups objects that have the same delimiter value
so that the listing process can be more efficient as it
does not need to report similar objects.
* Default: true
remote.s3.supports_versioning = true | false
* Optional.
* Specifies whether the remote storage supports versioning.
* Versioning is a means of keeping multiple variants of an object
in the same bucket on the remote storage.
* This setting determines how splunkd removes data from remote storage.
If set to true, splunkd will delete all versions of objects at
time of data removal. Otherwise, if set to false, splunkd will use a simple DELETE
(See https://docs.aws.amazon.com/AmazonS3/latest/dev/DeletingObjectVersions.html).
* Default: true
remote.s3.endpoint = <URL>
* Optional.
* The URL of the remote storage system supporting the S3 API.
* The scheme, http or https, can be used to enable or disable SSL connectivity
with the endpoint.
* If not specified and the indexer is running on EC2, the endpoint is
constructed automatically based on the EC2 region of the instance where the
indexer is running, as follows: https://s3-<region>.amazonaws.com
* Example: https://s3-us-west-2.amazonaws.com
remote.s3.multipart_download.part_size = <unsigned integer>
* Optional.
* Sets the download size of parts during a multipart download.
* This setting uses HTTP/1.1 Range Requests (RFC 7233) to improve throughput
overall and for retransmission of failed transfers.
* A value of 0 disables downloading in multiple parts, i.e., files are always
downloaded as a single (large) part.
* Do not change this value unless that value has been proven to improve
throughput.
* Minimum value: 5242880 (5 MB)
* Default: 134217728 (128 MB)
remote.s3.multipart_upload.part_size = <unsigned integer>
* Optional.
* Sets the upload size of parts during a multipart upload.
* Minimum value: 5242880 (5 MB)
* Default: 134217728 (128 MB)
remote.s3.multipart_max_connections = <unsigned integer>
* Specifies the maximum number of HTTP connections to have in progress for
either multipart download or upload.
* A value of 0 means unlimited.
* Default: 8
remote.s3.retry_policy = max_count
* Sets the retry policy to use for remote file operations.
* Optional.
* A retry policy specifies whether and how to retry file operations that fail
for those failures that might be intermittent.
* Retry policies:
+ "max_count": Imposes a maximum number of times a file operation is
retried upon intermittent failure both for individual parts of a multipart
download or upload and for files as a whole.
* Default: max_count
remote.s3.max_count.max_retries_per_part = <unsigned integer>
* When the remote.s3.retry_policy setting is max_count, sets the maximum number
of times a file operation is retried upon intermittent failure.
* Optional.
* The count is maintained separately for each file part in a multipart download
or upload.
* Default: 9
remote.s3.max_count.max_retries_in_total = <unsigned integer>
* Optional.
* When the remote.s3.retry_policy setting is max_count, sets the maximum number
of times a file operation is retried upon intermittent failure.
* The count is maintained for each file as a whole.
* Default: 128
remote.s3.timeout.connect = <unsigned integer>
* Optional
* Set the connection timeout, in milliseconds, to use when interacting with
S3 for this volume.
* Default: 5000 (5 seconds)
remote.s3.timeout.read = <unsigned integer>
* Optional
* Set the read timeout, in milliseconds, to use when interacting with S3
for this volume.
* Default: 60000 (60 seconds)
remote.s3.timeout.write = <unsigned integer>
* Optional
* Set the write timeout, in milliseconds, to use when interacting with S3
for this volume.
* Default: 60000 (60 seconds)
remote.s3.sslVerifyServerCert = <boolean>
* Optional.
* If this is set to true, Splunk verifies certificate presented by S3
server and checks that the common name/alternate name matches the
ones specified in 'remote.s3.sslCommonNameToCheck'
and 'remote.s3.sslAltNameToCheck'.
* Default: false
remote.s3.sslVersions = <versions_list>
* Optional.
* Comma-separated list of SSL versions to connect to 'remote.s3.endpoint'.
* The versions available are "ssl3", "tls1.0", "tls1.1", and "tls1.2".
* The special version "*" selects all supported versions. The version "tls"
selects all versions tls1.0 or newer.
* If a version is prefixed with "-" it is removed from the list.
* SSLv2 is always disabled; "-ssl2" is accepted in the version list
but does nothing.
* When configured in FIPS mode, ssl3 is always disabled regardless
of this configuration.
* Default: tls1.2
remote.s3.sslCommonNameToCheck = <commonName1>, <commonName2>, ..
* If this value is set, and 'remote.s3.sslVerifyServerCert' is set to
true, splunkd checks the common name of the certificate presented by
the remote server (specified in 'remote.s3.endpoint') against this
list of common names.
* No default.
remote.s3.sslAltNameToCheck = <alternateName1>, <alternateName2>, ..
* If this value is set, and 'remote.s3.sslVerifyServerCert' is set to true,
splunkd checks the alternate name(s) of the certificate presented by
the remote server (specified in 'remote.s3.endpoint') against this list
of subject alternate names.
* No default.
remote.s3.sslRootCAPath = <path>
* Optional
* Full path to the Certificate Authority (CA) certificate PEM format file
containing one or more certificates concatenated together. S3 certificate
is validated against the CAs present in this file.
* Default: [sslConfig/caCertFile] in the server.conf file
remote.s3.cipherSuite = <cipher suite string>
* Optional.
* If set, uses the specified cipher string for the SSL connection.
* If not set, uses the default cipher string.
* Must specify 'dhFile' to enable any Diffie-Hellman ciphers.
* Default: TLSv1+HIGH:TLSv1.2+HIGH:@STRENGTH
remote.s3.ecdhCurves = <comma separated list of ec curves>
* Optional
* ECDH curves to use for ECDH key negotiation.
* The curves should be specified in the order of preference.
* The client sends these curves as a part of Client Hello.
* Splunk software only supports named curves specified
by their SHORT names.
* The list of valid named curves by their short/long names can be obtained
by executing this command:
$SPLUNK_HOME/bin/splunk cmd openssl ecparam -list_curves
* e.g. ecdhCurves = prime256v1,secp384r1,secp521r1
* No default.
remote.s3.dhFile = <path>
* Optional
* PEM format Diffie-Hellman parameter file name.
* DH group size should be no less than 2048bits.
* This file is required in order to enable any Diffie-Hellman ciphers.
* No default.
remote.s3.encryption = sse-s3 | sse-kms | sse-c | none
* Optional
* Specifies the scheme to use for Server-side Encryption (SSE) for
data-at-rest.
* sse-s3: Check http://docs.aws.amazon.com/AmazonS3/latest/dev/UsingServerSideEncryption.html
* sse-kms: Check http://docs.aws.amazon.com/AmazonS3/latest/dev/UsingKMSEncryption.html
* sse-c: Check http://docs.aws.amazon.com/AmazonS3/latest/dev/ServerSideEncryptionCustomerKeys.html
* none: no Server-side encryption enabled. Data is stored unencrypted on
the remote storage.
* Default: none
remote.s3.encryption.sse-c.key_type = kms
* Optional
* Determines the mechanism Splunk uses to generate the key for sending
over to S3 for SSE-C.
* The only valid value is 'kms', indicating AWS KMS service.
* One must specify required KMS settings: e.g. remote.s3.kms.key_id
for Splunk to start up while using SSE-C.
* Default: kms
remote.s3.encryption.sse-c.key_refresh_interval = <unsigned integer>
* Optional
* Specifies the period, in seconds, at which a new key is generated and used
for encrypting any new data being uploaded to S3.
* Default: 86400 (24 hours)
remote.s3.kms.key_id = <string>
* Required if remote.s3.encryption = sse-c | sse-kms
* Specifies the identifier for Customer Master Key (CMK) on KMS. It can be the
unique key ID or the Amazon Resource Name (ARN) of the CMK or the alias
name or ARN of an alias that refers to the CMK.
* Examples:
Unique key ID: 1234abcd-12ab-34cd-56ef-1234567890ab
CMK ARN: arn:aws:kms:us-east-2:111122223333:key/1234abcd-12ab-34cd-56ef-1234567890ab
Alias name: alias/ExampleAlias
Alias ARN: arn:aws:kms:us-east-2:111122223333:alias/ExampleAlias
* No default.
remote.s3.kms.access_key = <string>
* Optional.
* Similar to 'remote.s3.access_key'.
* If not specified, KMS access uses 'remote.s3.access_key'.
* No default.
remote.s3.kms.secret_key = <string>
* Optional.
* Similar to 'remote.s3.secret_key'.
* If not specified, KMS access uses 'remote.s3.secret_key'.
* No default.
remote.s3.kms.auth_region = <string>
* Required if 'remote.s3.auth_region' is not set and Splunk can not
automatically extract this information.
* Similar to 'remote.s3.auth_region'.
* If not specified, KMS access uses 'remote.s3.auth_region'.
* No default.
remote.s3.kms.max_concurrent_requests = <unsigned integer>
* Optional.
* Limits maximum concurrent requests to KMS from this Splunk instance.
* NOTE: Can severely affect search performance if set to very low value.
* Default: 10
remote.s3.kms.<ssl_settings> = <...>
* Optional.
* Check the descriptions of the SSL settings for remote.s3.<ssl_settings>
above. e.g. remote.s3.sslVerifyServerCert.
* Valid ssl_settings are sslVerifyServerCert, sslVersions, sslRootCAPath,
sslAltNameToCheck, sslCommonNameToCheck, cipherSuite, ecdhCurves and dhFile.
* All of these are optional and fall back to same defaults as
the 'remote.s3.<ssl_settings>'.
[hot_bucket_streaming]
slices_list_executor_workers = <unsigned integer>
* Currently not supported. This setting is related to a feature that is
still under development.
* Number of workers that do list operations to discover slices during bucket recovery.
* Must be greater than 0.
* Default: 4
slices_download_executor_workers = <unsigned integer>
* Currently not supported. This setting is related to a feature that is
still under development.
* Number of workers that download slices during bucket recovery.
* Must be greater than 0.
* Default: 10
slices_build_executor_workers = <unsigned integer>
* Currently not supported. This setting is related to a feature that is
still under development.
* Maximum number of parallel bucket rebuilds during bucket recovery.
* Must be greater than 0.
* Default: 4
slices_removal_executor_workers = <unsigned integer>
* Currently not supported. This setting is related to a feature that is
still under development.
* Number of workers that remove slices after a bucket rolls to warm or is rebuilt.
* Must be greater than 0.
* Default: 2
slices_upload_executor_workers = <unsigned integer>
* Currently not supported. This setting is related to a feature that is
still under development.
* Number of workers that upload slices from hot buckets.
* Must be greater than 0.
* Default: 10
slices_upload_executor_capacity = <unsigned integer>
* Currently not supported. This setting is related to a feature that is
still under development.
* Maximum number of queued slices to be uploaded. This affects the same thread
pool that uses the 'slices_upload_executor_workers' setting.
* A value of 0 means that the queue capacity is unlimited.
* Default: 10
slices_upload_send_interval = <interval><unit>
* Currently not supported. This setting is related to a feature that is
still under development.
* Periodic send interval, in seconds, for the slices to be uploaded.
* Examples: 10s, 1m
* Must not be greater than 300s or 5m
* Default: 5s
slices_upload_size_threshold = <unsigned integer>[B|KB|MB]
* Currently not supported. This setting is related to a feature that is
still under development.
* Slice size threshold.
* When this threshold is reached, slice coalescing ends and the accumulated slice is uploaded.
* Must be a positive number followed by a size suffix.
* Valid suffixes: b: bytes, kb: kilobytes, mb: megabytes
* Suffixes are case insensitive.
* Must not be greater than 10MB
* Default: 1MB
[federated_search]
# This section contains settings for the data federation feature.
disabled = <boolean>
* Set this flag to 'false' to enable the data federation functionality on this instance.
* Default: true
[distributed_leases]
sslVerifyServerCert = <boolean>
* If set to true, the instance authenticates the remote server endpoint that
it is attempting to connect to.
* Default: false
disabled = <boolean>
* Determines whether or not the distributed lease manager is enabled.
* Default: true
server.conf.example
# Version 8.2.12 # # This file contains an example server.conf. Use this file to configure SSL # and HTTP server options. # # To use one or more of these configurations, copy the configuration block # into server.conf in $SPLUNK_HOME/etc/system/local/. You must restart # Splunk to enable configurations. # # To learn more about configuration files (including precedence) please see # the documentation located at # http://docs.splunk.com/Documentation/Splunk/latest/Admin/Aboutconfigurationfiles # Allow users 8 hours before they time out [general] sessionTimeout=8h pass4SymmKey = changeme # Listen on IPv6 in addition to IPv4... listenOnIPv6 = yes # ...but make all outgoing TCP connections on IPv4 exclusively connectUsingIpVersion = 4-only # Turn on SSL: [sslConfig] enableSplunkdSSL = true useClientSSLCompression = true serverCert = $SPLUNK_HOME/etc/auth/server.pem sslPassword = password sslRootCAPath = $SPLUNK_HOME/etc/auth/cacert.pem certCreateScript = genMyServerCert.sh [proxyConfig] http_proxy = http://proxy:80 https_proxy = http://proxy:80 proxy_rules = * no_proxy = localhost, 127.0.0.1, ::1 ######## SSO Example ######## # This example trusts all logins from the splunk web server and localhost # Note that a proxy to the splunk web server should exist to enforce # authentication [general] trustedIP = 127.0.0.1 ####### Cascading Replication Example ###### [cascading_replication] pass4SymmKey = someSecret max_replication_threads = auto max_replication_jobs = 5 cascade_replication_plan_reap_interval = 1h cascade_replication_plan_age = 8h cascade_replication_plan_fanout = auto cascade_replication_plan_topology = size_balanced cascade_replication_plan_select_policy = random ############################################################################ # Set this node to be a cluster manager. ############################################################################ [clustering] mode = manager replication_factor = 3 pass4SymmKey = someSecret search_factor = 2 ############################################################################ # Set this node to be a slave to cluster manager "SplunkManager01" on port # 8089. ############################################################################ [clustering] mode = slave manager_uri = https://SplunkManager01.example.com:8089 pass4SymmKey = someSecret ############################################################################ # Set this node to be a searchhead to cluster manager "SplunkManager01" on # port 8089. ############################################################################ [clustering] mode = searchhead manager_uri = https://SplunkManager01.example.com:8089 pass4SymmKey = someSecret ############################################################################ # Set this node to be a searchhead to multiple cluster managers - # "SplunkManager01" with pass4SymmKey set to 'someSecret and "SplunkManager02" # with no pass4SymmKey set here. ############################################################################ [clustering] mode = searchhead manager_uri = clustermanager:east, clustermanager:west [clustermanager:east] manager_uri = https://SplunkManager01.example.com:8089 pass4SymmKey=someSecret [clustermanager:west] manager_uri = https://SplunkManager02.example.com:8089 ############################################################################ # Configuration file change audit # To enable the feature, set 'disabled=false'. # Set 'mode=auto' to include all available features. ############################################################################### [config_change_audit] disabled = false mode = auto ############################################################################ # Open an additional non-SSL HTTP REST port, bound to the localhost # interface (and therefore not accessible from outside the machine) Local # REST clients like the CLI can use this to avoid SSL overhead when not # sending data across the network. ############################################################################ [httpServerListener:127.0.0.1:8090] ssl = false
Last modified on 22 August, 2023
| segmenters.conf | serverclass.conf |
This documentation applies to the following versions of Splunk® Enterprise: 8.2.12
Download manual
Feedback submitted, thanks!