Dimitri Fontaine

Hi @gkokolatos ; thanks for this PR and very detailed architecture presentation! A first level reaction here, not a full review yet even though I did skim through the code...

>> - [supervisor] I think we are missing a notion of static/dynamic services. The services that we have in current releases are all static, the pgbouncer service is dynamic: it...

> Wait. Just to be completely clear. You are suggesting that: * static services are implemented in a static and thus immutable array * dynamic services are implemented in a...

> Hey, I started working on the two things we agreed upon, namely introducing dynamic services and introducing a [services] section in the keeper configuration. Both are great tasks and...

Hi @gkokolatos ; Happy New Year! Thanks for the ping, I didn't realise you added so much to the current PR. Indeed, I think we are ready for a split...

Hi @MeydanOzeri and @zhangyahuai ; thanks for your interest in pg_auto_failover and in this feature to allow controlling the pgbouncer process directly within the pg_autoctl process tree. The sad reality...

Hi, I will have to have a deep look at the issue here when back from vacations in a couple weeks. I already think it's a bug though, so if...

Hi @rheaton ; please see also https://github.com/citusdata/pg_auto_failover/issues/602.

Yeah I'm as confused as you are. I mean, I suppose you're in front of an actual bug here, as in it's not the intended behaviour and I could remember...

That said our pattern is "^pgautofailover_standby_" ; who would use such a pattern in their own replication slots, that they want pg_auto_failover to refrain from handling? ;-)