You were absolutely right Tom.
Rising work_mem did the trick!
Many thanks :-)
What's the best value for work_mem ?
Gaëtan
Le 7 févr. 2010 à 07:38, Tom Lane a écrit :
> =?iso-8859-1?Q?Ga=EBtan_Allart?= writes:
>> I'm experiencing an interesting issue with PostgreSQL 8.4.2-r2.
>> I was running
Hello,
I'm a long time user and I had always installed postgres by compiling
from source. I recently moved to a new environment where
everything is installed as rpm and plus I need to install a newer
postgres for testing while keeping the existing version in place.
That leads me to repackage postg
This may belong in a bug report, but I'll post it here first...
There appears to be a **SERIOUS** problem with using pg_compresslog and
pg_uncompresslog with Postgresql 8.4.2.
Here's my configuration snippet:
full_page_writes = on # recover from partial page writes
wal_buffers
Koichi-san,
seems, very similar problem.
Regards,
Oleg
On Sun, 7 Feb 2010, Karl Denninger wrote:
This may belong in a bug report, but I'll post it here first...
There appears to be a **SERIOUS** problem with using pg_compresslog and
pg_uncompresslog with Postgresql 8.4.2.
Here's my confi
I doubt pinning buffers ever improve system on any halfway modern system. It
will often *look* like it has improved performance because it improves the
performance of the queries you're looking at -- but at the expense of
slowing down everything else.
There is a use case it would be useful for tho
Marek Lewczuk writes:
> there are a lot of improvements in hstore, that is planned for PG8.5 -
> as far I remember it compiles fine with PG8.4 so is it save to use it with
> PG8.4 ?
See hstore-new, which delivers the same code to be found in 9.0 as a
module against 8.3 and 8.4.
http://pgfound
Could any one please tell me what the overhead will be on the database
server and instance if I set
stats_row_level = on
at Postgesql 8.1.
AI Rumman wrote:
Could any one please tell me what the overhead will be on the database
server and instance if I set
stats_row_level = on
at Postgesql 8.1.
4.3% !
seriously, though, that would be very hard to quantify for a random set
of operations, and would be heavily dependent on what
