I have had this kind of problems before, and I seem to remember I even posted a bug about it.
For your information the error message you get does come from mod_ssl:
bash-2.03$ find . -type f -name \*.c -exec grep -l 'Failed to acquire global mutex lock' {} \;
./modules/ssl/ssl_engine_mutex.c
bash-2.03$
This problem appeared when I migrated from Apache 2.0.49 to 2.0.54. The default mutex for my platform (Solaris) was changed somewhere thereabout.
I ended up explicitly setting AcceptMutex and SSLMutex to pthread (the previous default) and never had this problem again. I believe using a file, fcntl or flock mutex on a shared drive is not recommended
-ascs
-----Original Message-----
From: Steely, Bruce (Mission Systems) [mailto:Bruce.Steely@xxxxxxx]
Sent: Thursday, June 08, 2006 4:09 PM
To: users@xxxxxxxxxxxxxxxx
Subject: RE: [users@httpd] Invalid argument: Failed to acquire global mutex lock
Thanks for replying, Axel.
I'm familiar with the SSLMutex issue, but I doubt that it is the problem in this case. The system failed over to another node that uses the same configuration files (on a shared drive), and it runs fine on the other node. It has to be something that is particular to the orignal node. I think it has to do with stopping and starting Apache while the process monitor was also attempting to start it. I don't know how the lock is implemented, but possibly it is trying to access the lock with an old reference that no longer exists?
Bruce
---------------------------------------------------------------------
The official User-To-User support forum of the Apache HTTP Server Project.
See <URL:http://httpd.apache.org/userslist.html> for more info.
To unsubscribe, e-mail: users-unsubscribe@xxxxxxxxxxxxxxxx
" from the digest: users-digest-unsubscribe@xxxxxxxxxxxxxxxx
For additional commands, e-mail: users-help@xxxxxxxxxxxxxxxx
|