Hello, I noticed that from time to time, device id of a gpfs volume is not same across whole gpfs cluster.
[root@kat1 ~]# stat /gpfs/vol1/ File: ‘/gpfs/vol1/’ Size: 262144 Blocks: 512 IO Block: 262144 directory Device: 28h/40d Inode: 3 [root@kat2 ~]# stat /gpfs/vol1/ File: ‘/gpfs/vol1/’ Size: 262144 Blocks: 512 IO Block: 262144 directory Device: 2bh/43d Inode: 3 [root@kat3 ~]# stat /gpfs/vol1/ File: ‘/gpfs/vol1/’ Size: 262144 Blocks: 512 IO Block: 262144 directory Device: 2ah/42d Inode: 3 this is really bad for kernel NFS as it uses device id for file handles thus NFS failover leads to nfs stale handle error. Is there a way to force a device number? -- Lukáš Hejtmánek Linux Administrator only because Full Time Multitasking Ninja is not an official job title _______________________________________________ gpfsug-discuss mailing list gpfsug-discuss at spectrumscale.org http://gpfsug.org/mailman/listinfo/gpfsug-discuss
