go to bug id or search bugs for
Every so often (i.e. every month or so) some of our websites stop responding. Looking at Apache 2's error_log reveals a number of the following errors:
PHP Fatal error: Maximum execution time of 30 seconds exceeded in /srv/www/htdocs/applicantday/include.php on line 60
(Line 60 corresponds to the oci_connect call).
Restarting Apache 2 does not resolve the issue. Rebooting the machine does. I'm currently using oci8 pecl from CVS from 2006-02-09.
This *looks* like a bug...?
Add a Patch
Add a Pull Request
I've never seen such problem and I'm sure I'm unable to reproduce it.
So please try to give me some more info.
Are you connecting to a remote Oracle server or it's local for Apache?
What happens on the server side at that moment?
It's a remote Oracle server, i.e. Apache is not on the same box.
Looking at $ORACLE_HOME/network/log/listener.log there's no mention of any login attempts during this weekend for the Apache machine.
Which client library are you using?
Try to use gdb to see where exactly it freezes, maybe that would help to understand the problem.
I will try, but it might be a number of weeks - or even months - before the problem appears again. In the meantime, you might want to change the status of this bug.
Also, it would be good if you try to reproduce it on another server.
Rather irritatingly, the problem has yet to reoccur.
After an investigation the problem seems to be in the Instant Client which freezes at some point when trying to connect to the non-working server.
There is nothing OCI8/PHP can do about it..
I know this comment is pretty late, but who knows! it might save someone from going bald.
There's an Oracle client bug that hits after server is up for 199 days and a server reboot would solve the issue. For folks who can't believe that a reboot would solve a 'client' library issue: