Franco,
When your clients perceive a hang, could you check the status of the bricks by
running,
# gluster volume status VOLNAME (run this on one of the 'server' machines in
the cluster.)
Could you also provide the statedump of the client(s),
by issuing the following command.
# kill -SIGUSR1
OK. I've just unmounted the data2 volume from a machine called tape1,
and now try to remount - it's hanging.
/bin/sh /sbin/mount.glusterfs nas5-10g:/data2 /data2 -o
rw,log-level=info,log-file=/var/log/glusterfs/glusterfs_data2.log
on the server
# gluster vol status
Status of volume: data2
BTW, if I update gluster to 3.5.1-0.1.beta1.el6 using yum,
glusterfs-libs doesn't install properly.
Cleanup: glusterfs-libs-3.5qa2-0.510.git5811f5c.el6.x86_64
11/11
Non-fatal POSTUN scriptlet failure in
Also forgot to mention that it's impossible to change any volume
settings while 3.4 clients are attached but I can unmount them all,
change the setting and them mount them all again.
gluster vol set data2 nfs.disable off
volume set: failed: Staging failed on nas5-10g. Error: One or more
