Prometheus querying capability and logging


Is Prometheus meant for accepting a huge number of queries (1K/second)? Does anyone have an idea of the higher limit that is been tested?

What is the purpose of Prometheus query logging?

  • Does Prometheus query log serve audit logging as well?
  • Is it intended for troubleshooting performance issues of Prometheus and slow queries?


I don’t know if anyone has done a lot of testing for very high query request rates. Typically it depends a lot on how much data each query needs.

Query logging was meant to help audit and provide information on what queries are being done to help debug high load and slow queries.

If you’re expecting a high query rate for the same queries, you might want to look at Trickster Cache to improve performance.

Thanks Ben, when you say “audit logging”, does the query log print the information about users and services who queried Prometheus?

No, Prometheus has no concept of users.

I thought them to be security-related audit logs. Audit in this context is the information about query events in detail. Is this correct?