Re: pg_xlogdump --stats

From: Heikki Linnakangas <hlinnakangas(at)vmware(dot)com>
To: Abhijit Menon-Sen <ams(at)2ndQuadrant(dot)com>, Andres Freund <andres(at)2ndquadrant(dot)com>
Cc: Dilip kumar <dilip(dot)kumar(at)huawei(dot)com>, <pgsql-hackers(at)postgresql(dot)org>, <furuyao(at)pm(dot)nttdata(dot)co(dot)jp>
Subject: Re: pg_xlogdump --stats
Date: 2014-08-21 07:06:39
Message-ID: 53F59A7F.5020600@vmware.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 07/07/2014 10:32 PM, Abhijit Menon-Sen wrote:
> At 2014-07-07 09:48:44 +0200, andres(at)2ndquadrant(dot)com wrote:
>>
>> I'd suggest only defining INT64_MODIFIER here and build INT64_FORMAT,
>> UINT64_FORMAT ontop, in c.h.
>
> Oh, I see. I'm sorry, I misread your earlier suggestion. Regenerated
> patches attached. Is this what you had in mind?

Committed the patch to add INT64_MODIFIER, with minor fixes.

The new rm_identify method needs to be documented. Not sure where;
surprisingly I can't find any documentation on the existing methods
either. Perhaps add comments to the RmgrData struct, explaining all of
the methods. In particular, should give guidelines on what output
belongs in rm_desc and what in rm_identify.

I think the names that rm_identify returns should match those that the
rm_desc functions print. As the patch stands, for example when heap_desc
prints out something like:

hot_update(init): xmax 123; new tid 1/2 xmax 456

The corresponding rm_identify output is:

HOT_UPDATE+INIT

It would be better to spell them the same, so that when you look at the
summary stats and the raw pg_xlogdump output, you can tell which records
contributed to which summary line.

- Heikki

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Michael Paquier 2014-08-21 07:20:11 Re: Specifying the unit in storage parameter
Previous Message Dilip kumar 2014-08-21 06:52:31 Re: TODO : Allow parallel cores to be used by vacuumdb [ WIP ]