Hi, looks like I was able to find and solve the cause of our problems. I noticed today that all "permission denied" error messages were connected with the new third cluster node that made rebalancing necessary in the first place. I took a look at the Gluster share directory on that server and found that a lot of the files there were not owned by the backup system user but by root. I don't know why GlusterFS created the files with root as owner, but after a global chown from the client mount point, the storage seems to be working again. If everything keeps running smoothly, the upgrade to 3.2 finally brought us a balanced and working backup storage and I'll be a happy (and very relieved) camper. Best regards, Jochen
