We just got a Dell with 12 x 600G 15K hot swappable disks + 2
internal 146G 15K disks (64G ram)
We will be using postgres 8.4.3 on RHEL 5 64 bit
Does it make sense to use a single Raid 10 stripe across the disks or
split into 2 sets of 3 disk pairs (given that we have 2 channels) ? We
plan to
Hi,
I was installed the Postgresql 8.3 and trying the use the
pg_lesslog_1.4.1_pg83 to reduce the size of WAL file when the WAL file is
archived.
1. Download the pg_lesslog_1.4.1_pg83.tar.gz file from pgfoundry.
2. unpacked the pglesslog source.
3. trying to run the make...facing below issue::
I hope the method with the environment variables will hold on.
For security reasons it is much better to use the variable method with
PGPASSWORD. I can set the environment hidden from any user by a program.
The .pgpass is readable for any admin, opposed to the statement in the docs:
On Microsoft
Alex,
Thanks for the explanation. That helps.
After looking at some of the pgsql_tmp files, It was clear we needed to
increase work_mem. So I switched back to the pgtune suggestions with one
change:
-bash-3.2$ diff postgresql.conf postgresql.conf.local | grep
log_temp_files = 0
On Wed, Jul 28, 2010 at 15:43, Irene Barg ib...@noao.edu wrote:
Alex,
The postmaster wouldn't start with shared_buffers = 5632MB, so we reduced it
to 2GB. The system's shared memory is:
[dpp...@pipen18 ~]$ sudo /sbin/sysctl -a | grep kernel | grep shmmax
kernel.shmmax = 4182863872
You can
I have a question that may be related to connection pooling.
We create a bunch of high-performance lightweight Postgres clients that serve
up images (via mod_perl and Apache::DBI). We have roughly ten web sites, with
ten mod_perl instances each, so we always have around 100 Postgres backends