On Thu, Sep 05, 2019 at 07:09:45PM -0400, Alvaro Herrera from 2ndQuadrant wrote:
> Can't we have pg_ctl just continue to wait indefinitely? So we'd set
> SERVICE_START_PENDING when wait_for_postmaster is out of patience, then
> loop again -- until recovery completes. Exiting pg_ctl on timeout seems
> reasonable for interactive use, but maybe for service use it's not
> reasonable.
The root of the problem here is that the time recovery takes is not
something that can be guessed, and that service registering happens in
the background. It depends on the time the last checkpoint occurred,
the load on the machine involved and the WAL operations done. So it
seems to me that Alvaro's idea is something which we could work on for
at least HEAD. There is also the path of providing a longer timeout,
still that's just a workaround..
My understanding is that this could be qualified as a bug because of
the fact that we require using again pg_ctl after starting the service
from the windows service control center.
So, are there plans to move on with this patch? It is waiting on
author for some time now.
--
Michael