From: | Peter Pilsl <pilsl(at)goldfisch(dot)at> |
---|---|
To: | pgsql-general(at)postgresql(dot)org |
Subject: | current_timestamp wrong |
Date: | 2001-08-22 21:50:38 |
Message-ID: | 20010822235038.A63349@i3.atat.at |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
version 7.0.1
In a bigger perl-project I had to hunt down a bug and it turned out that on
very rare occassions the time delivered by 'current_timestamp'
is wrong by a noticeable amount of time. (up to hours sometimes)
The problem occures when using the DBI-module for perl and mod_perl in
apache and the db-handle is stored globally for hours. On the first
call for current_timestamp after long idle-time the returned value is
*sometimes* incorrect.
Creating a subrouting, doing nothing but fetching and printing the
system-time and the database-time in the apps-logfile I get things like this:
2001-08-21-13:59:18 192.168.11.1 databasetime 2001-08-21 13:54:34+02
Is there any chance that this is a postgreSQL-problem or did someone
hear about this before or is it definitely a problem within perl, the
DBI-module or the application ?
It looks like a cache-problem, but no involved part should actually
do caching on such an operation. mod_perl is always a dangerous one
for caching, but in this case I used only local variables ..
thnx,
peter
--
mag. peter pilsl
phone: +43 676 3574035
fax : +43 676 3546512
email: pilsl(at)goldfisch(dot)at
sms : pilsl(at)max(dot)mail(dot)at
pgp-key available
From | Date | Subject | |
---|---|---|---|
Next Message | Murray Hobbs | 2001-08-22 22:01:07 | protected ON DELETE CASCADE |
Previous Message | Johann Zuschlag | 2001-08-22 20:43:03 | Re: Comparing fixed precision to floating (no anwer) |