Re: Is Recovery actually paused?
Dilip Kumar <dilipbalaut@gmail.com>
From: Dilip Kumar <dilipbalaut@gmail.com>
To: Yugo NAGATA <nagata@sraoss.co.jp>
Cc: Masahiko Sawada <sawada.mshk@gmail.com>,
Robert Haas <robertmhaas@gmail.com>, Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>, Kyotaro Horiguchi <horikyota.ntt@gmail.com>,
Simon Riggs <simon@2ndquadrant.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2021-01-28T04:25:42Z
Lists: pgsql-hackers
Attachments
- v8-0001-Provide-a-new-interface-to-get-the-recovery-pause.patch (application/octet-stream) patch v8-0001
On Wed, Jan 27, 2021 at 2:28 PM Dilip Kumar <dilipbalaut@gmail.com> wrote: > > On Wed, Jan 27, 2021 at 2:06 PM Yugo NAGATA <nagata@sraoss.co.jp> wrote: > > > > On Wed, 27 Jan 2021 13:29:23 +0530 > > Dilip Kumar <dilipbalaut@gmail.com> wrote: > > > > > On Wed, Jan 27, 2021 at 12:50 PM Masahiko Sawada <sawada.mshk@gmail.com> wrote: > > > > > > > > On Tue, Jan 26, 2021 at 2:00 AM Robert Haas <robertmhaas@gmail.com> wrote: > > > > > > > > > > On Sat, Jan 23, 2021 at 6:10 AM Bharath Rupireddy > > > > > <bharath.rupireddyforpostgres@gmail.com> wrote: > > > > > > +1 to just show the recovery pause state in the output of > > > > > > pg_is_wal_replay_paused. But, should the function name > > > > > > "pg_is_wal_replay_paused" be something like > > > > > > "pg_get_wal_replay_pause_state" or some other? To me, when "is" exists > > > > > > in a function, I expect a boolean output. Others may have better > > > > > > thoughts. > > > > > > > > > > Maybe we should leave the existing function pg_is_wal_replay_paused() > > > > > alone and add a new one with the name you suggest that returns text. > > > > > That would create less burden for tool authors. > > > > > > > > +1 > > > > > > > > > > Yeah, we can do that, I will send an updated patch soon. > > > > This means pg_is_wal_replay_paused is left without any change and this > > returns whether pause is requested or not? If so, it seems good to modify > > the documentation of this function in order to note that this could not > > return the actual pause state. > > Yes, we can say that it will return true if the replay pause is > requested. I am changing that in my new patch. I have modified the patch, changes - I have added a new interface pg_get_wal_replay_pause_state to get the pause request state - Now, we are not waiting for the recovery to actually get paused so I think it doesn't make sense to put a lot of checkpoints to check the pause requested so I have removed that check from the recoveryApplyDelay but I think it better we still keep that check in the WaitForWalToBecomeAvailable because it can wait forever before the next wal get available. -- Regards, Dilip Kumar EnterpriseDB: http://www.enterprisedb.com
Commits
-
Be clear about whether a recovery pause has taken effect.
- 32fd2b57d7f6 14.0 landed
-
Pause recovery for insufficient parameter settings
- 15251c0a60be 14.0 cited
-
Remove most volatile qualifiers from xlog.c
- 6ba4ecbf477e 9.5.0 cited