On Sat, Sep 29, 2012 at 1:21 PM, Alexander Farber
wrote:
>
> About not giving enough information -
> how much information do you want?
> If I list all my databases + source code
> of the scripts, I doubt anyone will read my mail.
Probably not if you just copied and pasted into the body of the ema
On Sun, Sep 30, 2012 at 2:36 AM, Scott Marlowe wrote:
>> Whoa aren't you running pg bouncer? If so then leave pg alone, adjust
>> pg bouncer. Revert that db side change, examine pgbouncer config etc.
>
>
> apache/php -> (500 persistent conns, cheap) -> pgbouncer -> (20
> persistent pgsql conns,
On Sat, Sep 29, 2012 at 4:21 PM, Scott Marlowe wrote:
> Whoa aren't you running pg bouncer? If so then leave pg alone, adjust
> pg bouncer. Revert that db side change, examine pgbouncer config etc.
Let me expand a bit on that point. The reason to use pgbouncer is
that you can have hundreds of
On Sat, Sep 29, 2012 at 2:21 PM, Alexander Farber
wrote:
> Hello Scott and others,
>
> On Sat, Sep 29, 2012 at 9:38 PM, Scott Marlowe
> wrote:
>> On Sat, Sep 29, 2012 at 11:27 AM, Alexander Farber
>> wrote:
>>> I've finally doubled up RAM to 32 GB for my Quad core
>>> CentOS 6.3 server and have
Hello Scott and others,
On Sat, Sep 29, 2012 at 9:38 PM, Scott Marlowe wrote:
> On Sat, Sep 29, 2012 at 11:27 AM, Alexander Farber
> wrote:
>> I've finally doubled up RAM to 32 GB for my Quad core
>> CentOS 6.3 server and have changed postgresql.conf to
>>
>>max_connections = 100
>>share
On Sat, Sep 29, 2012 at 11:27 AM, Alexander Farber
wrote:
> Hello,
>
> I've finally doubled up RAM to 32 GB for my Quad core
> CentOS 6.3 server and have changed postgresql.conf to
>
>max_connections = 100
>shared_buffers = 4096MB
> work_mem = 16M
>
> But don't see any speed improvement
On 09/29/12 11:43 AM, Jeff Janes wrote:
>http://serverfault.com/questions/433281/doubled-up-ram-to-32-gb-now-how-to-speed-up-a-lapp-server
If you expand the "COMMAND" field of the "top" display (by hitting "c"
on my linux), you can probably see what the top "postmaster" process
is doing.
Anyway
On Sat, Sep 29, 2012 at 10:27 AM, Alexander Farber
wrote:
> Hello,
>
> I've finally doubled up RAM to 32 GB for my Quad core
> CentOS 6.3 server and have changed postgresql.conf to
>
>max_connections = 100
>shared_buffers = 4096MB
> work_mem = 16M
>
> But don't see any speed improvement
Here is the 9.0 versionand yes I meant maintenance_work_mem
# Postgresql Memory Configuration and Sizing Script
# By: James Morton
# Last Updated 05/16/2012
#
# Note This script is meant to be used with by the postgres user with a
configured .pgpass file
# It is for Postgres version 9 running on L
On Wed, Sep 5, 2012 at 2:16 PM, jam3 wrote:
> Here is a bash script I wrote to print out mem config ffrom postgresconf.sql
> and os (centos 5.5 in this case). According to Gregory Smith in Postgresql
> 9.0 shared buffers should be appx 25-40% of avail Physical RAM. Also
> considerPostgres uses the
Here is a bash script I wrote to print out mem config ffrom postgresconf.sql
and os (centos 5.5 in this case). According to Gregory Smith in Postgresql
9.0 shared buffers should be appx 25-40% of avail Physical RAM. Also
considerPostgres uses the OS Buffer as it access the physical data and log
fil
On Tue, Sep 4, 2012 at 2:59 AM, Alexander Farber
wrote:
> Hello, thank you for your replies and sorry for the delay in my replying -
>
> On Thu, Aug 30, 2012 at 4:45 PM, Scott Marlowe
> wrote:
>> On Thu, Aug 30, 2012 at 8:42 AM, Scott Marlowe
>> wrote:
>>> users, and currently work_mem is set
On Tue, Sep 4, 2012 at 10:59 AM, Alexander Farber
wrote:
> I'll try changing work_mem to 2MB first - once I upgrade the RAM.
And then I'll increase it up to 16MB every day as Scott proposed.
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscript
Hello, thank you for your replies and sorry for the delay in my replying -
On Thu, Aug 30, 2012 at 4:45 PM, Scott Marlowe wrote:
> On Thu, Aug 30, 2012 at 8:42 AM, Scott Marlowe
> wrote:
>> users, and currently work_mem is set to 1M (the default.) If you
>> increase that to 16M, that'd be max
On Thu, Aug 30, 2012 at 8:42 AM, Scott Marlowe wrote:
> users, and currently work_mem is set to 1M (the default.) If you
> increase that to 16M, that'd be max 1.6G of memory, which you have
> free anyway right now.
Self correction here. Of course that's assuming 1 sort on average per
query. My
On Thu, Aug 30, 2012 at 5:42 AM, Alexander Farber
wrote:
> Hello,
>
> I run CentOS 6.3 server with 16 GB RAM and:
> postgresql-8.4.12-1.el6_2.x86_64
> pgbouncer-1.3.4-1.rhel6.x86_64
>
> The modified params in postgresql.conf are:
> max_connections = 100
> shared_buffers = 4096MB
T
On Thu, Aug 30, 2012 at 6:42 AM, Alexander Farber
wrote:
> Hello,
>
> I run CentOS 6.3 server with 16 GB RAM and:
> postgresql-8.4.12-1.el6_2.x86_64
> pgbouncer-1.3.4-1.rhel6.x86_64
>
> The modified params in postgresql.conf are:
> max_connections = 100
> shared_buffers = 4096MB
>
On 08/30/2012 07:42 PM, Alexander Farber wrote:
Hello,
I run CentOS 6.3 server with 16 GB RAM and:
postgresql-8.4.12-1.el6_2.x86_64
pgbouncer-1.3.4-1.rhel6.x86_64
The modified params in postgresql.conf are:
max_connections = 100
shared_buffers = 4096MB
and the pgbouncer run
Hello,
I run CentOS 6.3 server with 16 GB RAM and:
postgresql-8.4.12-1.el6_2.x86_64
pgbouncer-1.3.4-1.rhel6.x86_64
The modified params in postgresql.conf are:
max_connections = 100
shared_buffers = 4096MB
and the pgbouncer runs with:
pool_mode = session
server_reset_query
19 matches
Mail list logo