ports and endpoints

These tables list the ports which must be open to inbound traffic and internet endpoints which must be accessible to use . Use these tables to design the firewall rules for your deployment.

Endpoints for all deployments

This table shows a list of the internet endpoints that a deployment uses. This list is not exhaustive.

Endpoint	Outbound port	Protocol	Required?	Description
splunkbase.splunk.com	443	HTTPS	Required	Required for app installation and app upgrades.
Splunk Cloud	9997, 8089	TCP	Conditional	If your deployment uses a Splunk Cloud deployment instead of the embedded Splunk Enterprise instance, must be able to reach your Splunk Cloud deployment.
grpc.prod1-cloudgateway.spl.mobi	443	HTTPS	Conditional	If you use Splunk Mobile to access on mobile devices, your deployment must be able to reach grpc.prod1-cloudgateway.spl.mobi
https://e1345286.api.splkmobile.com/1.0/e1345286	443	HTTPS	Conditional	telemetry
*.pool.ntp.org	123	UDP	Required	Used by the chronyd service for system clock synchronization.
CentOS and RHEL mirrors	443	HTTPS	Required	Required to run YUM updates for operating system components and installed software packages. If your organization prefers, you can use a satellite server instead. See the Red Hat Knowledgebase article https://access.redhat.com/solutions/29269.
github.com	443	HTTPS	Required	Used to access the community playbook repository.
Other source control system	Dependent on the source control system.	Dependent on the source control system.	Conditional	Access is required if your deployment uses an alternative repository for playbooks.
Google Maps embed API	443	HTTPS	Conditional	Used by the MaxMind app to add visualizations for IP address geolocation results.
pypi.org	443	HTTPS	Required	Used by some apps to update or install their PIP dependencies.
App specific endpoints	Dependent on the app.	Dependent on the app.	Conditional	Apps might need to reach specific endpoints in order to provide their functions. Consult the app's documentation for details.

Ports for a standalone deployment

On a single instance deployment of where all services are contained on the same host, open these inbound ports in addition to allowing the Endpoints for all deployments.

Port	Required?	Description
TCP 22	Required	SSH port. Used for administering the operating system.
TCP 80	Required	Port for requests sent over HTTP. redirects all HTTP requests to HTTPS.
TCP 443	Optional	HTTPS port for the web interface and REST API. This port is for HTTPS traffic if you use the --port-forward option or answer "yes" to the port forwarding question when you install . This port must be exposed to access services without specifying the custom HTTPS port in the URL. You specify the HTTPS port during installation. It must be a port greater than 1023. Users who use the soar-prepare-system script during installation and who do not specify a custom HTTPS port will have 8443 set for them. Starting with Splunk SOAR (On-premises) release 5.3.5, upgrades from privileged deployments will set firewalld rules to forward TCP 443 to TCP 8443. In an AMI-based deployment, the custom HTTPS port is set to 9999.
TCP 8443	Required, Configurable	HTTPS ports for the web interface and REST API. This port must be exposed to access services. You specify a custom HTTPS port during installation. It must be a port greater than 1023. During upgrades from privileged deployments, Splunk SOAR (On-premises) will set firewalld rules to forward TCP 443 to TCP 8443. AMI-based deployments do not need to open this port.
TCP 8888	Required	Used for local communications between the WebSocket server and other components. Port must be available, but does not have to be open for external ingress or egress.
TCP 9999	Required for AMI-based deployments	HTTPS ports for the web interface and REST API. This port must be exposed to access services. For all AMI-based deployments, port 9999 is set as the custom HTTPS port.

Ports for externalized services

If you opt to deploy services such as Splunk Enterprise or Splunk Cloud, PostgreSQL, or a file share separately from your deployment, you need to make sure that can reach those services on your network.

In a clustered deployment, all services are external to , and an added load balancer. See Example: cluster for a diagram of a cluster.

Required ports for embedded Splunk Enterprise

Open these inbound ports on each node for embedded Splunk cluster configuration.

Port	Purpose
TCP 5121	Splunk Enterprise server HTTP Event Collector (HEC) service. Can be blocked on the Shared Services server if using an alternate Splunk Enterprise server.
TCP 5122	Splunk Enterprise server REST port. Can be blocked on the Shared Services server if using an alternate Splunk Enterprise server.

Required ports for non-embedded Splunk Enterprise

If you are using the non-embedded version of Splunk Enterprise, open these inbound ports on each node.

Port	Purpose
TCP 8088	Used as the HTTP Event Collecter (HEC) and provides searching capabilities.
TCP 8089	Used for the REST endpoint to send information to the Splunk instances.
TCP 9996-9997	Used for the universal forwarder to either forward or direct the indexers.

PostgreSQL database

A single instance deployment of uses a local instance of a PostgreSQL database. If you choose to use an external PostgreSQL database instead, you must make sure that can reach the database on your network. Open these inbound ports on each node for PostgreSQL connections.

In a clustered deployment, each node must be able to reach the PostgreSQL database. See About clusters in Install and Upgrade .

Port	Description
TCP 5432	PostgreSQL service. This port is also used by warm standby configurations for PostgreSQL streaming replication.
TCP 6432	Used by PgBouncer to interact with the PostgreSQL database.

File Shares

A single instance deployment of uses the local file system to store files for the vault. You can choose to expand storage capacity by using an external file share.

You can use any file system that meets your organization's security and performance requirements for your external file shares. You need to configure any required mounts and permissions. See Supported file systems and required directories.

These following tables uses NFS and GlusterFS as an example for file shares. In a clustered deployment, these inbound ports must be opened on each node, and in the case of GlusterFS, on each member of the GlusterFS server cluster.

Port	Description
TCP 445	CIFS protocol.
UDP 111	RPC portmapper service for GlusterFS and NFS.
TCP 111	RPC portmapper service for GlusterFS and NFS.
TCP 2049	GlusterFS and NFS for NFS exports. Used by the nfsd process.
TCP 38465	NFS mount protocol.
TCP 38466	NFS mount protocol.
TCP 38468	NFS Lock Manager, NLM.
TCP 38469	NFS ACL support.
TCP 24007	glusterd management port.
TCP 24008	glusterd management port.
TCP 49152+	For GlusterFS brick mounts. The total number of ports required to be open depends on the total number of bricks exported on the server. In most cases, 10 bricks is sufficient. You might need to open additional ports later if you add additional bricks.

Warm Standby

A single instance deployment of can have a warm standby configured to take over in the event of a failure of the primary deployment . See Warm standby feature overview in Administer Splunk SOAR (On-premises). Opening these inbound ports on each node is a prerequisite to enable warm standby.

Port	Description
TCP 22	SSH port. Used to send commands and to rsync data to the warm standby.
TCP 5432	PostgreSQL service. This port is used by warm standby configurations for PostgreSQL streaming replication.

Ports for connecting mobile devices to using Splunk Connected Experience apps

Open these inbound ports to enable registration of mobile apps, such as Splunk Mobile for iOS or Splunk Mobile for Android. In a clustered deployment, these ports must be opened on each node.

When the Enable Mobile App toggle is in the ON position, launches a new daemon, ProxyD. ProxyD connects to the Splunk Cloud Gateway automatically at grpc.prod1-cloudgateway.spl.mobi, on port 443 using the gRPC protocol.

uses the gRPC protocol to communicate to mobile apps through the Splunk Cloud Gateway.

For more information on Splunk Cloud Gateway, its encryption, and the data that is sent and received, see About the Splunk Cloud Gateway security process in Install and Administer Splunk Cloud Gateway.

Port	Description
TCP 15505	Port 15505 is used by ProxyD to listen for inter-process communication from other daemons on the same instance.
TCP 443	Port 443 is the inbound port to 's REST endpoints from ProxyD. REST requests from connected mobile devices received from Splunk Cloud Gateway are sent to and received from other daemons by ProxyD on port 443.

For other ports you might need to open, see Prerequisites and Requirements in the Install and Administer Splunk Cloud Gateway manual.

Ports for clustered deployments of

can be deployed as a cluster of nodes connected to a server or set of servers providing a PostgreSQL database, file shares, a Splunk platform deployment, and a load balancer. A cluster can be deployed on-premises or in Amazon Web Services. See About clusters in Install and Upgrade .

This table lists the inbound ports required by nodes for inter-node communication and access to services on the cluster.

Port	Description
TCP 22	SSH port. Used for administering the operating system of the cluster node. Also used by SSHD for GlusterFS.
TCP 80	Port for requests sent over HTTP. redirects all HTTP requests to HTTPS.
TCP 443	HTTPS port for the web interface and REST API. This port is for HTTPS traffic if you use the --port-forward option or answer "yes" to the port forwarding question when you install . This port must be exposed to access services without specifying the custom HTTPS port in the URL. You specify the HTTPS port during installation. It must be a port greater than 1023. Users who use the soar-prepare-system script during installation and who do not specify a custom HTTPS port will have 8443 set for them. Starting with Splunk SOAR (On-premises) release 5.3.5, upgrades from privileged deployments will set firewalld rules to forward TCP 443 to TCP 8443. In an AMI-based deployment, the custom HTTPS port is set to 9999.
TCP 8443	HTTPS ports for the web interface and REST API. This port must be exposed to access services. You specify a custom HTTPS port during installation. It must be a port greater than 1023. During upgrades from privileged deployments, Splunk SOAR (On-premises) will set firewalld rules to forward TCP 443 to TCP 8443. AMI-based deployments do not need to open this port.
TCP 9999	HTTPS ports for the web interface and REST API. This port must be exposed to access services. For all AMI-based deployments, port 9999 is set as the custom HTTPS port.
TCP 4369	RabbitMQ port mapper. All cluster nodes must be able to communicate with each other on this port.
TCP 5100 - TCP 5120	Daemon inter-process communication ports.
TCP 5671	RabbitMQ service. All cluster nodes must be able to communicate with each other on this port.
TCP 8300	Consul RPC services. All cluster nodes must be able to communicate with each other on this port.
TCP 8301	Consul internode communication. All cluster nodes must be able to communicate with each other on this port.
TCP 8302	Consul internode communication. All cluster nodes must be able to communicate with each other on this port.
TCP 8888	Used for local communications between the WebSocket server and other components. Port must be available, but does not have to be open for external ingress or egress.
TCP 15672	RabbitMQ admin UI and HTTP API service. The RabbitMQ admin UI is disabled by default. Unless you want to use the admin UI, you can block this port. If you choose to activate the RabbitMQ HTTP API and web UI, all cluster nodes must be able to communicate with each other on this port.
TCP 25672	RabbitMQ internode communications. All cluster nodes must be able to communicate with each other on this port.

For information on RabbitMQ ports, see "Networking" on the RabbitMQ documentation. For more information on Consul's required ports, see "Ports" in the Consul documentation on the HashiCorp website.

Example: Default firewalld settings for an unprivileged cluster

Here is an example of the default settings for firewalld when is deployed as an unprivileged cluster. Splunk Connected experiences apps access is not enabled in this example.

[phantom@phantom ~]$ sudo firewall-cmd --list-all
public (active)
  target: default
  icmp-block-inversion: no
  interfaces: ens160
  sources:
  services: dhcpv6-client http https ssh
  ports: 9999/tcp 27100-27200/tcp 5121/tcp 5122/tcp 8300/tcp 8301/tcp 8302/tcp 4369/tcp 5671/tcp 25672/tcp 15672/tcp 443/tcp
  protocols:
  masquerade: no
  forward-ports: port=443:proto=tcp:toport=9999:toaddr=
  source-ports:
  icmp-blocks:
  rich rules:

Example: cluster

A cluster consists of a load balancer, three or more nodes, a PostgreSQL database, file shares, and either a Splunk Enterprise or Splunk Cloud deployment.

This diagram shows an example of a cluster, with the connections marked.