The database slows down after a few weeks
Hello
We have a problem with our database. It is accessed by
Java using Hibernate. When the program has been
running for a couple of weeks some transactions are
delayed and are executed up to an hour later than then
they are first called.
If the computer is restarted it will take about a week
before the problems appear again.
If the following sequence of commands is executed
(reinserting the data) then it will take a couple of
weeks before the problem appears again.
pg_dump > file.sql
dropdb
createdb
psql < file.sql
Is this behaviour something seen before and in that
case: how to get around it? I also need to know why it
is happening so that when I have a solution I can
guarantee that the problem will not appear again.
The system with the problem is unfortunately placed
abroad with very limited remote access. Therefore "Try
this, and try that"-solutions are not possible in this
case. I have a hard time reproducing the error on my
local system which makes the problem so much harder.
If no cause can be found: Are there any other useful
tips?
Versions running:
Linux Fedora Core 3
psql (PostgreSQL) 7.4.8
jdk1.5.0_02
____________________________________________________________________________________
Luggage? GPS? Comic books?
Check out fitting gifts for grads at Yahoo! Search
http://search.yahoo.com/search?fr=oni_on_mail&p=graduation+gifts&cs=bz
In response to Bekar Bartaia <bbartaia@yahoo.com>:
Hello
We have a problem with our database. It is accessed by
Java using Hibernate. When the program has been
running for a couple of weeks some transactions are
delayed and are executed up to an hour later than then
they are first called.
What is your maintenance schedule? Specifically: vacuum
and analyze? What is the output of vacuum verbose when
you are experiencing the problem?
If the computer is restarted it will take about a week
before the problems appear again.If the following sequence of commands is executed
(reinserting the data) then it will take a couple of
weeks before the problem appears again.pg_dump > file.sql
dropdb
createdb
psql < file.sqlIs this behaviour something seen before and in that
case: how to get around it? I also need to know why it
is happening so that when I have a solution I can
guarantee that the problem will not appear again.The system with the problem is unfortunately placed
abroad with very limited remote access. Therefore "Try
this, and try that"-solutions are not possible in this
case. I have a hard time reproducing the error on my
local system which makes the problem so much harder.If no cause can be found: Are there any other useful
tips?Versions running:
Linux Fedora Core 3
psql (PostgreSQL) 7.4.8
jdk1.5.0_02____________________________________________________________________________________
Luggage? GPS? Comic books?
Check out fitting gifts for grads at Yahoo! Search
http://search.yahoo.com/search?fr=oni_on_mail&p=graduation+gifts&cs=bz---------------------------(end of broadcast)---------------------------
TIP 4: Have you searched our list archives?
--
Bill Moran
http://www.potentialtech.com
Bekar Bartaia <bbartaia@yahoo.com> writes:
If no cause can be found: Are there any other useful
tips?
Versions running:
Linux Fedora Core 3
psql (PostgreSQL) 7.4.8
jdk1.5.0_02
The first tip I'd give you is to get onto something newer than PG 7.4.
Fedora 3 is pretty long in the tooth as well.
regards, tom lane
On 7/27/07, Tom Lane <tgl@sss.pgh.pa.us> wrote:
Bekar Bartaia <bbartaia@yahoo.com> writes:
Versions running:
Linux Fedora Core 3
psql (PostgreSQL) 7.4.8
jdk1.5.0_02The first tip I'd give you is to get onto something newer than PG 7.4.
Fedora 3 is pretty long in the tooth as well.
Although unlikely to be related to your problem or its solution, but
jdk1.5.0_02 is certainly not recommended. The "5" series is currently
at 1.5.0_12. Also, it is likely that an update to JRE6 (currently
1.6.0_02 Build 06) would yield few problems and significant
performance gains for a 10 minute effort.
-Damian