On Sat, April 2, 2011 21:26, Scott Marlowe wrote:
> On Sat, Apr 2, 2011 at 11:26 AM, Henry C. <henka@cityweb.co.za> wrote:
>
>> On Sat, April 2, 2011 14:17, Jens Wilke wrote:
>>
>>> Nevertheless since at least 8.4 IMO there's no need to bother with
>>> manual vacuum any more.
>>
>> Sadly, in my case, the db is so busy that autovac processes run for weeks
>> and never catch up (insufficient h/w for the app quite frankly - the
>> addition of some more SSD drives have already helped). I eventually run up
>> against the wraparound wall and the only way forward is to stop everything
>> and dump/restore (vacuuming the entire db would take an unknown period of N
>> x weeks - dumping/restoring completes in a day or two).
>
> Have you tried upping the aggressiveness of autovacuum?
Thanks for the suggestion - I'm going to give autovacuum_vacuum_cost_delay=0 a
try (instead of the default 20ms, which if I'm reading the docs correctly,
means the same aggressiveness as a manual vacuum), and see how things go in
terms of the I/O cost/responsiveness and ensuring the damn vacuums finish in a
reasonable time before the wraparound tactical nuke hits :)
