Horton

Re: Problem while setting the fpw with SIGHUP

Amit Kapila <amit.kapila16@gmail.com>

View thread View on postgr.es
From: Amit Kapila <amit.kapila16@gmail.com>
To: Michael Paquier <michael@paquier.xyz>
Cc: Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>, Robert Haas <robertmhaas@gmail.com>, hlinnaka <hlinnaka@iki.fi>, Dilip Kumar <dilipbalaut@gmail.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2018-09-27T10:49:02Z
Lists: pgsql-hackers 
On Thu, Sep 27, 2018 at 1:32 PM Michael Paquier <michael@paquier.xyz> wrote:
>
> On Thu, Sep 27, 2018 at 11:18:02AM +0530, Amit Kapila wrote:
> > Your proposed solution makes sense to me.  IIUC, this is quite similar
> > to what Dilip has also proposed [1].
>
> Indeed.  I would just add with the patch a comment like that:
> "Perform this call outside the critical section so as if the instance
> just got out of recovery, the upcoming WAL insert initialization does
> not trigger an assertion failure."
>

I think this is mostly fine, but it seems "if the instance just got
out of recovery" doesn't fit well because it can happen anytime after
recovery, this code gets called from checkpointer.  I think we can
slightly tweak it as below:
"Perform this outside critical section so that the WAL insert
initialization done by RecoveryInProgress() doesn't trigger an
assertion failure."

What do you say?

> Sure, feel free to if you have some room.  I am fine to take care of it
> as well, so that's up to you to decide.

Okay, I will take care of it.

>  Adding a comment like what I
> proposed upthread is necessary in my opinion.

Agreed.

-- 
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com

Commits

  1. Fix assertion failure when updating full_page_writes for checkpointer.

  2. Attach FPI to the first record after full_page_writes is turned on.

  3. Revamp the WAL record format.

  4. Move the backup-block logic from XLogInsert to a new file, xloginsert.c.