Dear all, I found the problem after some desperate changes by accident, using Saso's recommendation #2. Vcsa apparently uses another interface to access data stores than for management console. After giving the management interface rw access to the store instead of root, the second, undocumented IF hit dmesg with an error while trying to stat the dir in console. The rest was easy: added the second IP to the whitelist too, done.
Adding versions: NFS is v4, VMware is 5.1 currently in evaluation mode. Thanks for helping me with your thoughts. Br Sebastian Sašo Kiselkov <[email protected]> schrieb: >On 01/24/2013 03:57 PM, Sebastian Gabler wrote: >> Hello, >> >> I am using a share via nfs as esxi datastore for more than a year. 3 >> hosts have root access. I have added a vcenter server appliance to >> manage the esxi hosts, and added the vcenter server's IP address to the >> allowed hosts using "zfs set sharenfs=root=host1:host2:host3:vcsa >> dataset/share". >> >> Still, all write actions that are issued to the nfs datastore through >> the vcsa fail with access problems, while they succeed issues through >> the hosts directly. Unfortunately, some actions like cloning or moving >> virtual machines require the vcsa to access the nfs datastore. >> >> I have subsequently re-started the nfs service on my openindiana >> machine, and re-booted the vcsa to no avail. I have browsed the web for >> this issue and did not find a related answer. If anybody has a good idea >> what to do, I would be thankful. > >While I don't know what the answer is, I suggest the following >diagnostic steps: > > 1) snoop the NFS traffic and compare the differences between the > machines that work and those that don't > > 2) does the NFS server print anything in /var/adm/messages or dmesg? > > 3) isn't the NFS server running IPF by any chance? (just guessing) > >Cheers, >-- >Saso > _______________________________________________ OpenIndiana-discuss mailing list [email protected] http://openindiana.org/mailman/listinfo/openindiana-discuss
