[Gluster-users] Outages serverl times a day

Alex Sanderson alex at greenshifteurope.com
Tue Jan 25 08:06:03 UTC 2011


With GlusterFS 2 we experienced many problems with similar high load.  
One of the contributors was updatedb which runs at least once per day 
and scans all filenames on most mounted filesystems by default on some 
linuxes.  I would imagine that with metadata caching this would be less 
of a problem.

Sorry I don't have any real help.

On 2011-01-19 16:06, Jürgen Winkler wrote:
> Hi Peoples,
> i hope any one can help me to solve the Problem we are facing here.
> We have 7 clients connected two Servers (distributed and replicated) ,
> the File system is EXT3.
> The used Glusterfs Version is 3.0.7
> On the Servers is one Raid5 Lun mounted on the Folder X, in that folder
> are 5 Sub-folders witch are exported to the clients.
> 3 clients are connected to Sub-folder A, 1 is connected to the
> Sub-folder B, 1 to C and so on.
> Each of the Clients is running a webserver , and several times a day all
> of the Client are having load peaks and stop Serving content for about 3
> - 7 minutes.
> also on the Servers there is a smaller load peak at the same time.
> On the 2 Servers are more raidluns and glusterfs exports witch do not
> have these problems .
> Thx in advance
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users

More information about the Gluster-users mailing list