Re: Nicer error when connecting to standby with hot_standby=off

On 2021-Mar-23, James Coleman wrote:

> On Tue, Mar 23, 2021 at 1:46 AM Fujii Masao <masao.fujii@oss.nttdata.com> wrote:

> > Therefore for now what we've not reached the consensus is what message
> > should be logged at PM_STARTUP. I'm thinking it's better to log
> > "the database system is starting up" in that case because of the reasons
> > that I explained upthread.

> I understand your point now, and I agree, that makes sense.

Please note that PM_STARTUP mode is very very short-lived.  It only
starts happening when postmaster launches the startup process, and
before the startup process begins WAL replay (as changed by
sigusr1_handler in postmaster.c).  Once WAL replay begins, the PM status
changes to PM_RECOVERY.  So I don't think we really care all that much
what message is logged in this case.  It changes very quickly into the
CAC_NOTCONSISTENT message anyway.  For this state, it seems okay with
either what James submitted in v7, or what Fujii said.

However, for this one

+       case CAC_NOTCONSISTENT:
+           if (EnableHotStandby)
+               ereport(FATAL,
+                       (errcode(ERRCODE_CANNOT_CONNECT_NOW),
+                        errmsg("the database system is not accepting connections"),
+                        errdetail("Consistent recovery state has not been yet reached.")));

Maybe it makes sense to say "... is not accepting connections *yet*".
That'd be a tad redundant with what the DETAIL says, but that seems
acceptable.

-- 
Álvaro Herrera                            39°49'30"S 73°17'W