From: | Barbara Stephenson <barbara(at)turbocorp(dot)com> |
---|---|
To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | pgsql-admin(at)postgresql(dot)org |
Subject: | Re: tuning our database by increasing shared buffer |
Date: | 2009-06-23 19:43:09 |
Message-ID: | 4A41304D.10703@turbocorp.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
Thank ypu!
Tom Lane wrote:
> Barbara Stephenson <barbara(at)turbocorp(dot)com> writes:
>
>> We will be consolidating from 4 databases to 2 and want to make sure that
>> these parameters are the only ones that need changing. Please advise.
>>
>
>
>> Current Future
>> ===== =====
>> Max_connection = 50 125
>> Shared_buffers = 16MB 48MB
>>
>
> You will need to make sure that the FSM size parameters are correct for
> the combined databases, too.
>
>
>> Shouldn't we increase the max_locks_per_transaction from 64 to 100 or 128
>> since we have more than doubled the # of connections?
>>
>
> No, because the lock table size automatically scales with
> max_connections. (Probably max_locks_per_transaction should have been
> called max_locks_per_connection ...)
>
>
>> max_prepared_transaction is set at default of 5 which is says if we use it to
>> set it to max_connection.
>>
>
> Are you using prepared transactions at all? If not, I'd actually
> recommend setting that to zero to make sure nobody creates a prepared
> transaction accidentally. You do *not* want anyone doing PREPARE
> TRANSACTION unless there's an XA manager or something in place to make
> sure the prepared xact gets committed or rolled back reasonably soon.
>
> regards, tom lane
>
>
--
Regards,
Barbara Stephenson
/EDI Specialist/Programmer/
*Turbo, division of OHL*
2251 Jesse Jewell Pkwy
*Gainesville, GA 30507*
tel: (678)989-3020 fax: (404)935-6171
barbara(at)turbocorp(dot)com
www.ohl.com
From | Date | Subject | |
---|---|---|---|
Next Message | Jakov Sosic | 2009-06-23 20:18:30 | Re: WAL archiving and backup TAR |
Previous Message | Tom Lane | 2009-06-23 19:25:51 | Re: tuning our database by increasing shared buffer |