From: | "Mason Hale" <masonhale(at)gmail(dot)com> |
---|---|
To: | "Steve Atkins" <steve(at)blighty(dot)com> |
Cc: | "pgsql-general General" <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: vacuum taking an unusually long time |
Date: | 2008-07-16 17:08:46 |
Message-ID: | 8bca3aa10807161008k1934c9cbu15f29cf2828733e4@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
> On Wed, Jul 16, 2008 at 10:26 AM, Mason Hale <masonhale(at)gmail(dot)com> wrote:
>> Here's some of the strace output:
>>
>> select(0, NULL, NULL, NULL, {0, 10000}) = 0 (Timeout)
>>
>> If I read the 'select(2)' man page correctly, it appears this process is
>> waiting indefinitely for a NULL file descriptor.
>> That looks pretty stuck to me.
>>
>
> No, it's just sleeping. For 10 milliseconds or so.
>
Ah, thanks.
> Should I kill this autovacuum process via pg_cancel_backend?
>> Or is there a better way to "unstick" it?
>>
>> Any ideas what may have caused it to get into this condition?
>>
>>
> It looks like "read a page, sleep for 80 milliseconds, repeat".
>
> I'd look at your settings for autovacuum_vacuum_cost_limit /
> autovacuum_vacuum_cost_delay, and maybe the contents of pg_autovacuum.
>
autovacuum_vacuum_cost_limit = -1
autovacuum_vacuum_cost_delay = 20ms
'select * from pg_autovacuum;' returns zero rows
I killed the one autovacuum process already that I thought was hung (via
pg_cancel_backend).
I can see one other autovacuum process running. The other vacuum process was
started manually.
Should I be concerned that pg_autovacuum is empty when I have an autovacuum
process running?
Mason
>
> Cheers,
> Steve
>
>
> --
> Sent via pgsql-general mailing list (pgsql-general(at)postgresql(dot)org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-general
>
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2008-07-16 17:17:32 | Re: vacuum taking an unusually long time |
Previous Message | Scott Marlowe | 2008-07-16 16:57:10 | Re: Which design would be faster ... |