Sourcegraph's metrics include a single high-level metric alert_count which indicates the number of level=critical and level=warning alerts each service has fired over time for each Sourcegraph service. This is the same metric presented on the Overview Grafana dashboard:
Description: The number of alerts each service has fired and their severity level. The severity levels are defined as follows:
critical: something is definitively wrong with Sourcegraph.- Examples: Database inaccessible, running out of disk space, running out of memory.
- Suggested action: Page a site administrator to investigate.
warning: something could be wrong with Sourcegraph.- Examples: High latency, high search timeouts.
- Suggested action: Email a site administrator to investigate and monitor when convenient.
Values:
- Although the values of
alert_countare floating-point numbers, only their whole numbers have meaning. For example:0.5and0.7indicate no alerts are firing, while1.2indicates exactly one alert is firing and3.0indicates exactly three alerts firing.
Labels:
level: eithercriticalorwarning, as defined above.service_name: the name of the service that fired the alert, one of the following constants:"frontend""github-proxy""gitserver""precise-code-intel""query-runner""replacer""repo-updater""searcher""symbols""zoekt-indexserver""zoekt-webserver""syntect-server"
name: the name of the alert that the service fired (chosen by the service)description: a human-readable description of the alertinstance: identifies the Kubernetes pod, Docker container, or host machine from which the alert came.
Examples:
To get examples of how you might consume this metric in your own alerting system, see: Custom consumption of Sourcegraph alerts.
A complete reference of Sourcegraph's vast set of Prometheus metrics is not yet available. If you are interested in this, please reach out by filing an issue or contacting us at [email protected].
