Re: Improve description of XLOG_RUNNING_XACTS

From: Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com>
To: sawada(dot)mshk(at)gmail(dot)com
Cc: masao(dot)fujii(at)oss(dot)nttdata(dot)com, pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: Improve description of XLOG_RUNNING_XACTS
Date: 2022-07-21 08:28:40
Message-ID: 20220721.172840.842755856795467411.horikyota.ntt@gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

At Thu, 21 Jul 2022 16:58:45 +0900, Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> wrote in
> > > The patch looks good to me.

By the way +1 to this.

> > The subxids can reach TOTAL_MAX_CACHED_SUBXIDS =
> > PGPROC_MAX_CACHED_SUBXIDS(=64) * PROCARRAY_MAXPROCS. xact_desc_commit
> > also shows subtransactions but they are at maximum 64. I feel like
> > -0.3 if there's no obvious advantage showing them.
>
> xxx_desc() functions are debugging purpose functions as they are used
> by pg_waldump and pg_walinspect etc. I think such functions should
> show all contents unless there is reason to hide them. Particularly,
> standby_desc_running_xacts() currently shows subtransaction
> information only when subtransactions are overflowed, which got me
> confused when inspecting WAL records.

I'm not sure just confusing can justify that but after finding
logicalmsg_desc dumps the whole content, I no longer feel against to
show subxacts.

Just for information, but as far as I saw, relmap_desc shows only the
length of "data" but doesn't dump all of it. generic_desc behaves the
same way. Thus we could just show "%d subxacts" instead of dumping
out the all subxact ids just to avoid that confusion.

However, again, I no longer object to show all subxids.

regards.

--
Kyotaro Horiguchi
NTT Open Source Software Center

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Amit Kapila 2022-07-21 08:36:31 Re: Handle infinite recursion in logical replication setup
Previous Message Junwang Zhao 2022-07-21 08:26:38 [PATCH v1] eliminate duplicate code in table.c