Re: New WAL record to detect the checkpoint redo location

Dilip Kumar <dilipbalaut@gmail.com>

From: Dilip Kumar <dilipbalaut@gmail.com>
To: Michael Paquier <michael@paquier.xyz>
Cc: Andres Freund <andres@anarazel.de>, Heikki Linnakangas <hlinnaka@iki.fi>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2023-08-25T05:38:25Z
Lists: pgsql-hackers

Attachments

On Fri, Aug 18, 2023 at 10:12 AM Dilip Kumar <dilipbalaut@gmail.com> wrote:
>
> On Fri, Aug 18, 2023 at 5:24 AM Michael Paquier <michael@paquier.xyz> wrote:
> >
> > On Thu, Aug 17, 2023 at 01:11:50PM +0530, Dilip Kumar wrote:
> > > Yeah right, actually I was confused, I assumed it will return the
> > > start LSN of the record.  And I do not see any easy way to identify
> > > the Start LSN of this record so maybe this solution will not work.  I
> > > will have to think of something else.  Thanks for pointing it out.
> >
> > About that.  One thing to consider may be ReserveXLogInsertLocation()
> > while holding the WAL insert lock, but you can just rely on
> > ProcLastRecPtr for the job after inserting the REDO record, no?
>
> Yeah right, we can use ProcLastRecPtr.  I will send the updated patch.

Here is the updated version of the patch.

-- 
Regards,
Dilip Kumar
EnterpriseDB: http://www.enterprisedb.com

Commits

  1. During online checkpoints, insert XLOG_CHECKPOINT_REDO at redo point.

  2. Unify two isLogSwitch tests in XLogInsertRecord.