Dimitri Fontaine

Hi @jnehlmeier ; see https://github.com/citusdata/pg_auto_failover/pull/399 which is an attempt at having pg_autoctl take care of the "alter user ... password ..." step directly. We never got around to finish that...

Hi @jnehlmeier, thanks for another bug report. This one is troubling, having two nodes get both to wait_primary should never happen. Would you have kept pg_autoctl logs from the nodes...

On a first reading it looks like the secondary node is failing right after a manual failover has been started. Is that correct? If that's the case, I will play...

Hi @gfphoenix78 ; we would consider such a PR for compatibility purposes yes. What I think should be done then is to change all the `*_id` names in the functions...

Alternative thinking: is there a way to get the caller of the function in the logs when that happens? When using `-vv` or `-vvv` we output __FILE__ and __LINE__ in...

> When trying this out locally with `pg_autoctl stop --pgdata node1` (the primary). I got this error in the logs: I can't reproduce at the moment. I suppose that's because...

Hi @jnehlmeier ; thanks for opening this issue. The situation is a little complex with the HBA and the roles we use, and I believe we have a good opportunity...

The property of the password encryption (either md5 or scram at this point) is decided by the `password_encryption` GUC, see https://www.postgresql.org/docs/current/runtime-config-connection.html#GUC-PASSWORD-ENCRYPTION. At the moment we don't change that GUC from...

I think I would start with the following place in the code, where we create the `pgautofailover_monitor` user with hard-coded password and HBA rules, and then navigate up from there....

Hi @trecloux ; I could see merging such a PR yes. I think this would have to be optional though, as the existing tooling for Postgres logs analysis are usually...