[horde] postresql sessionhandler problem

[COMBES Julien - CETE Lyon/DI/ET/PAMELA]

Le 01/10/07 13:16, Jan Schneider a écrit  :
> Zitat von COMBES Julien - CETE Lyon/DI/ET/PAMELA 
> <julien.combes at i-carre.net>:
>> [...]
>> Sometimes, maybe under connexions peak, horde become more and more 
>> slow to finally become inaccessible for some users and then for all 
>> users. Restarting apache doesn't help. The only thing which help is to 
>> restart postgresql.
>>
>> During the problem :
>>  - it seems that there is many lock on the horde_sessionhandler database.
>>  - all the tables on the postgreSQL's server (of horde and others 
>> applications which use this server) are working fine and respond 
>> correctly except the horde_sessionhandler table.
>>  - the number of apache's processes increase quickly.
>> [...]
> 
> Do you see anything in the logs? Can you determine which statements are 
> being locked when enabling debug logging?

I didn't see anything in horde's log with PEAR_LOG_INFO the last time we 
got the problem.
The only leads I have on this problem are given by the database admin 
team which saw locks on the sessionhandler table.

We have got a problem on our test's platform [1] with PEAR_LOG_DEBUG and 
postgresql sessionhandler which cause apache2 to segfault (no problem in 
PEAR_LOG_INFO or with PEAR_LOG_DEBUG and php session). Is Murphy back ? ;-)
So that we are a little afraid to configure PEAR_LOG_DEBUG on our 
production's plateform with pgqsl sessionhandler. Nevertheless, we will 
try to activate it tomorrow during the lunch's time. After, we must have 
to wait some hours or days for horde becoming slow

[1] this palteform is completly separated from the production's platform 
: other hardware, other apache, other postgresql server and so on.

thanks,
Julien