Re: Hard to Use WAS: Hard limit on WAL space

From: Craig Ringer <craig(at)2ndquadrant(dot)com>
To: Martijn van Oosterhout <kleptog(at)svana(dot)org>
Cc: Josh Berkus <josh(at)agliodbs(dot)com>, "Joshua D(dot) Drake" <jd(at)commandprompt(dot)com>, Robert Haas <robertmhaas(at)gmail(dot)com>, Magnus Hagander <magnus(at)hagander(dot)net>, Daniel Farina <daniel(at)heroku(dot)com>, MauMau <maumau307(at)gmail(dot)com>, Jeff Janes <jeff(dot)janes(at)gmail(dot)com>, pgsql-hackers(at)postgresql(dot)org
Subject: Re: Hard to Use WAS: Hard limit on WAL space
Date: 2013-06-15 10:20:44
Message-ID: 51BC3FFC.9040408@2ndquadrant.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 06/15/2013 05:57 PM, Martijn van Oosterhout wrote:
> On Sat, Jun 15, 2013 at 12:43:10PM +0800, Craig Ringer wrote:
>> Bloat ------
>>
>> Table bloat. Table bloat has been a major issue with PostgreSQL
>> users/admins for years. Anyone care to explain to me in a simple
>> paragraph how to find out if you have table or index bloat issues
>> in your database and what to do about it? (Maybe we need
>> "pg_catalog.pg_index_bloat" and "pg_catalog.pg_table_bloat"
>> views including FILLFACTOR correction?)
>>
>> I think I'll draft up a patch to add exactly that.
>
> Nice list btw. I monitor this by using the excellent check_progres
> nagios plugin, which has stuff to check for things like this.

It's been a vital tool for me too. It probably serves as a reasonable
guide for some things we could really usefully expose in system views.
Once in core we can document them in the main user manual, making them
reasonably discoverable.

Argh. I think my TODO has some kind of horrible disease, it keeps
growing uncontrollably.

> Which makes me think that it might be possible to add some other
> checks like this, in for example pg_ctl. A big fat warning 'your
> data may be eaten' might get noticed at startup.

The users who have this kind of issue aren't the ones running pg_ctl.
They'll usually be using launchd, systemctl, upstart, sysv init
scripts, etc ... whatever, something that sends the warning straight
to the system logs that they likely never read.

I don't have tons of sympathy for these people, but I do think making
fsync=off so easy to set without understanding it is kind of like
handing a grenade to a toddler.

- --
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.13 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBAgAGBQJRvD/8AAoJELBXNkqjr+S2aG8H/jnjATe1M+4O6k8iaS6Okgux
HQU14nDaBh7tbRaNQZUlNFDaVoQA6cynN8Xzq8k5lqJxGnuNRR7SNw8+cZZiZmMe
pS2f9q2IyOMz3T/mBNPuAFcPTbp6pjYrBNpMEGF6FYDhmUMSEfhf4Cp1Ns4FG0kx
o5dIXnhgDpCCTBK4XiYqbijFGe0pqbOH98fTQJLXb2ItgE17t4jU0YoYPJovjjT8
xKnDggN+H3uPMmNTcxn0VL6XcrjM6oDeBQPtzCiePWWxYD4nwP3d0ZIok13jZSHm
KC3NWgYQ7uP8/NJitnqewMQ8RArQjAWsW94deZt28jNDeaKp/vovQlZtrU2M6dQ=
=aysr
-----END PGP SIGNATURE-----

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Cédric Villemain 2013-06-15 10:24:31 Re: [PATCH] Remove useless USE_PGXS support in contrib
Previous Message Andres Freund 2013-06-15 10:20:28 Re: pluggable compression support