I've been told that to get NFS working you need to start the client(s) with the option "--disable-direct-io-mode". I had a similar problem creating vmdks during backup jobs and this fixed it.
On Mon, Feb 15, 2010 at 1:29 AM, Romans Ščugarevs <[email protected]>wrote: > Hi! > > Yes I've tried reexporting using unsfd too, but problem was the same. > What is even more interesting, when I'm trying to upload an iso image to > the storage using ESXi interface, everything is fine(image size >1GB) so no > problem in creating large files. > > > Quoting "Raghavendra G" <[email protected]>: > > Hi Romans, >> >> Have you tried reexporting glusterfs using unfsd? >> >> regards, >> On Sun, Feb 14, 2010 at 5:02 PM, Romans Ščugarevs <[email protected] >> >wrote: >> >> My configuration: >>> Two node(raid1) cluster based on glusterfs with NFS server. >>> The system is replicating nicely. Changes are synced. The storage is >>> exported over NFS for vmware esx usage. >>> The problem is when I am using vmware ESX to create a VM on the NSF >>> storage, the result is an empty unusable virtual disk. When I export some >>> local ext3 from one of the nodes, correct 8GB VM disk is created. When I >>> reexport the glusterfs mount, I 'm getting unusable VMs(almost zero >>> size). I >>> suppose it has something to do with so called thin provisioning over NFS. >>> Any advice is appreciated! >>> >>> >>> Regards, >>> Roman. >>> >>> _______________________________________________ >>> Gluster-users mailing list >>> [email protected] >>> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users >>> >>> >> >> >> -- >> Raghavendra G >> >> > > _______________________________________________ > Gluster-users mailing list > [email protected] > http://gluster.org/cgi-bin/mailman/listinfo/gluster-users >
_______________________________________________ Gluster-users mailing list [email protected] http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
