Details
-
Improvement
-
Resolution: Done
-
Major
-
None
Description
Currently the only alerts shown in the Grafana dashboards will be the ones fired by Cluster Monitor. There is a certain set of alerts that is a better fit for Prometheus (e.g. simple metric thresholds such as resident ratio), and others that are very difficult to do in the Cluster Monitor (e.g. any log alerts on memcached.log, which cannot be got from /sasl_logs and thus must be pulled in via Fluent Bit and analysed via Loki).
We need to look into ways of showing those alerts in our dashboards, ideally in the same tables and totals as the existing ones.
Attachments
Issue Links
- links to