After #32, on startup elector will now wait for Prometheus to be ready before taking part into the election.
Which raises the question: what happens when the Prometheus starts to fail (OOM or crashes)? In that case, we should expect to see the promehteus-elector pod release the leadership if it has it.
We should likely monitor the health endpoint of prometheus just like we do in #32 and pull off from the election whem we detect that our prometheus pod isn't healthy at all.
What's wrong?
After #32, on startup elector will now wait for Prometheus to be ready before taking part into the election.
Which raises the question: what happens when the Prometheus starts to fail (OOM or crashes)? In that case, we should expect to see the promehteus-elector pod release the leadership if it has it.
We should likely monitor the health endpoint of prometheus just like we do in #32 and pull off from the election whem we detect that our prometheus pod isn't healthy at all.