Hi all, with NFS v3 I had following setup: 1 RAID system several LUNs 2 servers providing NFS service (nodeA, nodeB)
I'd set up the two servers making sure the disks they export have identically configured major/minor number. Finally I set up an virtual interface and I was done. With NFSv4 this setup does not work anymore out of the box. If I stop (while having a client writing) nodeA and start nodeB NFS throws ESTALE (stale file handles). Stopping means: 1. unplumb the virtual interface first 2. umount the FS I tried to fiddle with DELEGATION set to off, which appeared to work until I tried an x86 client, which again threw ESTALE This error suggests different major/minor number on the two servers but let me assure you: it is not the case ! Any clues as to where to fiddle and how to do it right ? TIA Oliver This message posted from opensolaris.org
