Can you sync with me on the CloudNativePG Slack?

I can't review these until the weekend, but I can give you guidance, so it's as easy as possible.

Hi @itay-grudev, sure 😊

In short about this PR: as it contains breaking change it will be merged a bit later.

@itay-grudev can you please help understand what is failing it tests?

@dragoangel You're missing the monitoring section in the per-pooler configuration, hence the error. This wasn't a problem before since all the properties were always set. The best way to fix that would be an extensive use of default I suppose.

P.S. Because the old property is pooler and the new one poolers, this can be made backwards compatible.

@itay-grudev you mean in auto tests?
We can simplify this at all and move monitoring to the . Values.monitoring

And leave only podMonitor relabeling & metric relebling under cluster & poolers

Hi @itay-grudev I pushed proper label example and tests.

I see 3 options:

1. leave as is
2. move .Values.cluster.monitoring and .Values.pooler.monitoring to .Values.monitoring, and leave podMonitor relablings under cluster and each pooler if they need to get this settings.
3. remove .Values.poolers.[0].monitoring.enabled and .Values.poolers.[0].monitoring.podMonitor.enabled and take this info from cluster, while still looks for relablings under specific pooler. It will assume - if we create pod monitor for cluster we want to have pod monitor for pooler too, but this not always can be true?

For me all 3 options are okay, but 1 and 2 looks better.

@dragoangel FYI: I went with option 1 as it was the simplest.