Rich telemetry is key to operational observability, but having too many metrics makes finding the right one a hassle.
Metrics Explorer makes that easier. In a Splunk dashboard, it helps you:
- find available metrics,
- filter metrics to what you want,
- see the dimensions and sources for each metric, and
- build a
mstatssearch.
Once metrics and dimensions are selected, the dashboard's output is a base search that can be used directly in Search or refined with a tool like the Outcold Solutions AI Agent.
The app collects no data and stores nothing. Every panel is a search over metrics you already index.
- Splunk Enterprise or Splunk Cloud Platform 10.0 or later.
- Read access to the metrics indexes you want to explore.
Install from Splunkbase, or download metricsexplorer-<version>.tgz from the
GitHub releases and install it with
Apps > Manage Apps > Install app from file. The "Source code" archives GitHub attaches to each
release are not installable apps.
- Install the app and open Metrics Explorer.
- Enter a metric name (wildcards allowed), an index (wildcards allowed) and a time range, then select Submit.
- Click a row in Metric names to show that metric's Topology.
- Click a metric, source, host, index or dimension cell in Topology to add it to Basic criteria. The charts, the SPL panel and Open in Search follow the criteria.
- Open the Recent samples tab for the raw samples behind a metric.
The app's objects are readable by every role and editable by admin, sc_admin and power.
Contact Scott Raiford on the Splunk user Slack, or open a GitHub issue.
This software is released under the MIT license.
See LICENSE for details.
All trademarks are property of their respective owners.
This project began as an internal tool used by Outcold Solutions to identify useful metrics in new data sets. It's now released to the Splunk community as open-source software to help others in getting useful insight from their metrics.
Contributions are welcome, as described in CONTRIBUTING.