Missing monitoring documentation

[zamazan4ik]

Hi!

After reading the documentation, for it is not clear, how I should monitor the database. Ideally, I want to see some explanation, how to configure monitoring, which monitoring ways are supported (e.g. Prometheus), ready-to-use Grafana dashboards, etc.

Would be awesome, if you add such information into the documentation.

Thanks in advance.

[mponomar]

I can tell you how we monitor Comdb2 databases at Bloomberg, generally. That uses some internal tooling, so the code is not part of this repository.

• Poll comdb2_metrics for general database performance counters (lock waits, page requests, etc.)
• Poll comdb2_sql_client_stats for client stats (where requests are coming from, what the requests are)
• Poll comdb2_fingerprints for query performance metrics (the cost of queries, how often they run, etc.)

We build dashboards from those metrics.

[zamazan4ik]

@mponomar Could you please add this information somewhere into the documentation? Thanks!