Re: Multiple pg_waldump --rmgr options

From: Daniel Gustafsson <daniel(at)yesql(dot)se>
To: Heikki Linnakangas <hlinnaka(at)iki(dot)fi>
Cc: pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Multiple pg_waldump --rmgr options
Date: 2021-06-28 10:34:32
Message-ID: D94C9152-1362-4637-A529-109EE92D4E74@yesql.se
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

> On 18 May 2021, at 15:50, Heikki Linnakangas <hlinnaka(at)iki(dot)fi> wrote:

> The reason is that if you specify multiple --rmgr options, only the last one takes effect.

That's in line with how options are handled for most binaries, so this will go
against that. That being said, I don't think thats a problem here really given
what this tool is and it's intended usecase.

This patch makes the special case "--rmgr=list" a bit more awkward than before
IMO, as it breaks the list processing, but nothing we can't live with.

> I propose the attached to allow selecting multiple rmgrs

I agree with the other +1's in this thread, and am marking this as ready for
committer.

As a tiny nitpick for readability, I would move this line inside the string
comparison case where the rmgr is selected. Not that it makes any difference
in practice, but since that's where the filtering is set it seems a hair
tidier.
+ config.filter_by_rmgr_enabled = true;

--
Daniel Gustafsson https://vmware.com/

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Ranier Vilela 2021-06-28 11:05:37 Re: Fix uninitialized copy_data var (src/backend/commands/subscriptioncmds.c)
Previous Message Arne Roland 2021-06-28 10:16:04 Re: Rename of triggers for partitioned tables