Please provide documentation about Graylog Metrics

[gshill]

Please provide documentation about Graylog metrics for use in Grafana

What?

The Prometheus exporter grabs metrics from Graylog for use in Grafana. I am trying to create some dashboards in Grafana based on those metrics, but no where on Graylog's site/documentation are there any clues as to what they all mean.

Here is a short screen shot:

Why?

I can't make dashboards for context that I do not fully understand what they are pulling

Your Environment

• Graylog Version: 4.3.2
• Elasticsearch Version:
• MongoDB Version:
• Operating System:
• Browser version:

[bernd]

@gshill Thanks for bringing up the topic. We consider adding descriptions to the Prometheus metrics in the future. :+1:

[laurivosandi]

Somewhat related, perhaps deserves separate ticket: I could not find metrics usable for alerting, eg count of messages that JSON extractor failed to parse, message count submitted to/rejected by Elastic eg because of indexing errors. Note that Prometheus metric names could be somewhat self-explainatory, see more here -- for example gl_buffer_usage isn't clear about the unit, what is it being counted here?

[riskersen]

Any news on this? Personally, I like the amount of metrics available, but it's totally unclear how to monitor+graph them.

[coffee-squirrel]

It'd be great to have useful HELP content for metrics as users+customers try to piece together dashboards. If that'd be too much effort, an official reference dashboard (#18275) could help get people started (best practices for what to watch, etc.).