[Linux-cluster] Upgrades lead to volumes gone?

[isplist at logicore.net]

I've got LVS running checking on this server so the log goes as such (wish I 
could get rid of this error, have yet to find a way);

Nov 15 15:00:44 cweb92 httpd[3984]: [client 192.168.1.74] PHP Notice:  
Undefined index:  HTTP_HOST in /var/www/html/index.php on line 28

Suddenly, the mount is gone, no errors, no warning, nothing, it's gone on ALL
nodes. I log into each node, remount and it's fine. This doesn't change the 
status of the cluster either.

Nov 15 15:01:03 cweb92 httpd[3978]: [error] [client 192.168.1.74] File does 
not exist: /var/www/html

That's why I'm not sure where to look just yet.

Mike


On Thu, 15 Nov 2007 15:21:51 -0500, Shawn Wells wrote:
> 
> 
> Can you be a little more specific on what you mean?  Once the cluster is
> up do the mounts randomly disappear?  Are they mounted to begin with?
> 
> Please clarify for me.
> 
> --
> Shawn D. Wells
> Solutions Architect, Federal Team
> Red Hat, Inc
> swells at redhat.com
> C: 443-534-0130
> 
> 
> isplist at logicore.net wrote:
>> Based on my other thread about new nodes not coming up, I was able to
>> resolve
>> this by updating the software on all of the nodes to match.
>> 
>> All nodes have now come up, however, the logical volumes keep getting lost
>> now? Actually, it is the mount which keeps going away. There do not seem
>> to be
>> any error messages in the logs either?
>> 
>> Any thoughts on what might be causing this and what else I should be
>> looking
>> for?
>> 
>> Mike
>> 
>> 
>> --
>> Linux-cluster mailing list
>> Linux-cluster at redhat.com
>> https://www.redhat.com/mailman/listinfo/linux-cluster