Re: Stats collector on rampage (8.2.3)

From: Gaetano Mendola <mendola(at)bigfoot(dot)com>
To: "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>
Cc: Andrew Dunstan <andrew(at)dunslane(dot)net>
Subject: Re: Stats collector on rampage (8.2.3)
Date: 2007-10-29 10:08:08
Message-ID: 4725B108.4080107@bigfoot.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Andrew Dunstan wrote:
>
>
> Gaetano Mendola wrote:
>> hubert depesz lubaczewski wrote:
>>
>>>> On Mon, Oct 29, 2007 at 09:52:24AM +0100, Gaetano Mendola wrote:
>>>>
>>>>>> it seems that the stats collector on my box is using more CPU than
>>>>>> it did in the past.
>>>>>>
>>>> it's well known bug, and it was fixed in 8.2.4:
>>>> http://www.postgresql.org/docs/current/interactive/release-8-2-4.html
>>>> ...
>>>> Prevent the statistics collector from writing to disk too frequently
>>>> (Tom)
>>>> ...
>>>>
>>
>> I saw that, upgrading the DB at this very moment is not doable, killing
>> that process will the postmaster respawn another one? BTW I discover that
>> it was triggered by the time change due the daylight saving.
>>
>>
>>
>
> you do realize that this upgrade wouldn't require a dump/restore, don't
> you? It would be close to instantaneous.

Sure I do, for me it's even easier than that, I use DRBD so I can just shut
down one node upgrade it and then upgrade the other one; however upgrade the
DB means for me "touch" an entire satellite trasmission platform, I have to
schedule a platform maintenance for that...

Regards
Gaetano Mendola

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Magnus Hagander 2007-10-29 11:25:42 pgsql: New versions of mingw have gettimeofday(), so add an autoconf
Previous Message Gaetano Mendola 2007-10-29 10:02:13 Re: Stats collector on rampage (8.2.3)