Configure your Splunk environment to monitor DSP metrics
To use metrics to analyze the health of your DSP deployment with Splunk software, you must configure DSP to send metrics data to the Splunk platform using the HTTP Event Collector (HEC), and you must configure your Splunk environment to properly receive the metrics data.
Configure Splunk Enterprise or Splunk Cloud to receive DSP metrics
You must configure your Splunk environment to properly receive the metrics data from your DSP deployment. The default index for the DSP metrics data is _dsp_metrics
. It is best practice for Splunk Enterprise to use the default index, but depending on your needs and local configuration, you can define a custom index in the indexes.conf
file. If you are using Splunk Cloud, you must define a custom index.
See Create custom indexes for information about creating custom indexes in Splunk Enterprise. See Manage Splunk Cloud Platform indexes for information about creating indexes in Splunk Cloud.
If you define a custom index, you must edit the macros.conf
file in the Splunk App for DSP and update the definition for DSP metrics index in the following stanza.
[dsp_metrics_index] definition = index=_dsp_metrics iseval = 0
The index set in macros.conf
must match the target index you define in your Splunk Enterprise or Splunk Cloud configuration
To learn more about Splunk Enterprise configuration files, see:
Configure DSP to send metrics to the Splunk platform using HEC
You must configure DSP to send data to the Splunk platform using the HTTP Event Collector (HEC).
Prerequisites
- A Splunk instance with HEC enabled and a valid HEC token. Your HEC token must be configured to send data to the
_dsp_metrics
index. For information about how to enable HEC and create a HEC token, see Use the HTTP Event Collector.
Steps
Follow these steps to configure DSP to send data to the Splunk Platform using the HTTP Event Collector (HEC).
- Type the following in the working directory from a DSP node.
- Provide DSP with the HEC token to use to send DSP metrics data to the Splunk platform.
./set-config K8S_SPLUNK_ENTERPRISE_HEC_TOKEN <your token>
- Provide DSP with the HEC URL. For load balancing, you can specify multiple HEC URLs, separated by commas.
./set-config K8S_SPLUNK_ENTERPRISE_HEC_URL https://<your IP>:8088
- Set the metrics index to send data to. If you are using a custom metrics index, enter the name of your custom metrics index instead. You must change the application's knowledge objects if you are using a custom metrics index.
./set-config K8S_SPLUNK_ENTERPRISE_METRICS_INDEX _dsp_metrics
- (Optional - Skip this step if you've already given the DSP cluster a name during installation). Give the DSP cluster a name. This name will be shown in the dashboards in the Splunk App for DSP.
./set-config K8S_CLUSTER <cluster_name>
- (Optional) Set the scrape interval to define the frequency that metrics data is collected. The default is 30 seconds. You must specify the amount of time by using a number and the "s" time unit. For example, if you wanted the scrape interval to be 10 seconds, use
10s
in your command../set-config K8S_PROMETHEUS_SCRAPE_INTERVAL <number_of_seconds>
- Finally, enable the metrics to be sent.
./set-config K8S_SPLUNK_ENTERPRISE_METRICS_ENABLED true
- Provide DSP with the HEC token to use to send DSP metrics data to the Splunk platform.
- After setting the configurations, deploy your changes.
./deploy
- (Optional) Confirm that the deployment was successful by checking that a
prometheus-writer
pod is now running.kubectl -n monitoring get pods
- Wait for DSP to start sending data to your Splunk environment. This may take up to 10 minutes.
- To confirm that DSP is sending the metrics data to the Splunk platform, open the Search & Reporting app in your Splunk instance and search for your data. Use the following search criteria:
| mstats count(*) WHERE index="_dsp_metrics"
Install the Splunk App for DSP | Configure your Splunk environment to monitor DSP logs |
This documentation applies to the following versions of Splunk® Data Stream Processor: 1.2.0, 1.2.1-patch02, 1.2.1, 1.2.2-patch02, 1.2.4, 1.2.5, 1.3.0, 1.3.1
Feedback submitted, thanks!