Skaffold should expose metrics/telemetry
AndreasBergmeier6176 opened this issue · comments
Expected behavior
There should be a way of accessing run time of each processing step (e.g. build artifact X, wait on deployment Y).
Bonus points if it is in a format that can be easily pushed/digested via Prometheus.
Actual behavior
In our CI run, we execute Skaffold. To optimize deployment we then need to look at the timestamps for log output. With these timestamps, we can then guess, what processing steps took how long and where to best optimize for swifter deployment.
Information
Our end goal is to push Skaffold metrics to Prometheus and build Dashboards. Then we should easily see, which deployments slow us down the most.
Steps to reproduce the behavior
skaffold run