#1751 [Housekeeping] enhance grafana metrics explanation
Issue created by
ncouronne
Describe the issue
Grafana dashboard for flyte are great but some metrics are hard to understand:
For example:
2021-10-26 11_51_54-Metrics in prometheus_grafana▾
What is the unit ops/s ? How to interpret the value 0,0040 ops/s which in this case correspond to one workflow execution ?
What if we do not do this?
Monitor flyte and get metrics on the executed workflows is crucial for production exploitation of flyte, but more crucial is to get realistic and understandable metrics
The result may be that the metrics cannot be used and dev user will not be able to monitor easily the platform. In consequence it could lead dev user to create their proper metrics or consider flyte as a hard platform to monitor without personal dev.
Related component(s)
grafana dashboard from flyteadmin and flytepropreller exposed metrics
Are you sure this issue hasn't been raised already?
☑︎ Yes
Have you read the Code of Conduct?
☑︎ Yes
flyteorg/flyte