Re: Process startup infrastructure is a mess

From: pgsql-hackers-owner(at)postgresql(dot)org
> [mailto:pgsql-hackers-owner(at)postgresql(dot)org] On Behalf Of Andres Freund
> I think we should seriously consider doing a larger refactoring of this
> soon. I've some ideas about what to do, but I'd welcome some thoughts on
> whether others consider this a serious problem or not, and what they think
> we should do about this, first.

Thank you for raising this. I think so, too. Some other things that quickly come to mind are:

* The signal handlers are similar in many processes and redundant. It's not easy to see how the processes respond to a particular signal and what signal can be used for new things such as dumping the stack trace in the server log. Maybe we can have an array of process attributes that specify the signal handlers, startup/shutdown order, etc. Then the common process management code handles processes based on the information in the array.

* postmaster.c is very large (more than 6,000 lines) and hard to read.

* It may be easy to forget adding initialization code in the Windows-specific path (SubPostmasterMaain).

* It's confusing to find out which process counts toward max_connections, MaxBackends, the count of auxiliary processes. As a user, I'm sometimes confused about the relationship between max_connections, autovacuum_max_workers, max_wal_senders, max_worker_processes.

Regards
Takayuki Tsunakawa