I have seen this same problem but have not been able to find a workaround other than to delete the file from the server directly. I was not able to figure out a way to reproduce the symptom reliably, but in my case I suspect it was related to heavy concurrent access. Does that seem plausible in light of your access patterns? .. Lana (lana.deere at gmail.com) On Wed, Dec 29, 2010 at 3:30 PM, Dan Bretherton <d.a.bretherton at reading.ac.uk> wrote: > We have an occasional problem that prevents deletion of certain GlusterFS > mounted files. ?See the following, for example, with corresponding log file > message. > > kevin at cd /glusterfs/atmos/users/kih/ECHAM5/TS4-TEMP > kevin at rm HYBRID_TEMP_207212 > rm: cannot remove `HYBRID_TEMP_207212': Invalid argument > > [2010-12-28 00:59:04.298331] W [fuse-bridge.c:888:fuse_unlink_cbk] > glusterfs-fuse: 3997: UNLINK() /users/kih/ECHAM5/TS4-TEMP/HYBRID_TEMP_207212 > => -1 (Invalid argument) > > The file was deleted without error on a machine where the volume was mounted > via NFS. ?I have four compute servers that are using the GlusterFS client > for performance reasons. ?Operating system and GlusterFS package details are > as follows. > > [root at nemo1 TS4-TEMP]# cat /etc/redhat-release > CentOS release 5.5 (Final) > [root at nemo1 TS4-TEMP]# uname -a > Linux nemo1.nerc-essc.ac.uk 2.6.18-194.el5 #1 SMP Fri Apr 2 14:58:14 EDT > 2010 x86_64 x86_64 x86_64 GNU/Linux > [root at nemo1 TS4-TEMP]# rpm -qa | grep -i gluster > glusterfs-fuse-3.1.1-1 > glusterfs-core-3.1.1-1 > > Is there anything I can do to stop this from happening, other than using NFS > instead of GlusterFS client? > > -Dan. > > _______________________________________________ > Gluster-users mailing list > Gluster-users at gluster.org > http://gluster.org/cgi-bin/mailman/listinfo/gluster-users >
