[SOLVED] Proxmox 3.1 Lost Mount Point in Cluster

[Raymond Burns]

I have a 3 node cluster. There are two NFS servers with a mount point for each type (ie: KVM, Templates, ISOs, Backups, Containers)
One of the cluster will not mount to Omni02CT, which corresponds to the second NFS Server and mount point for Containers.

This is the error I receive in /var/log/messages:

Nov 12 15:21:59 zwtprox1 kernel: ct0 mount: server 192.168.222.12 not responding, timed out

Omni02CT and Omni02KVM does not seem to mount properly, however, backups02 mounts properly. Just to be sure, I created a new mount point that is called "testCT" which is the same NFS mount to 192.168.222.12 as "Omni02CT"
It works without a problem.

This is my storage.cfg:

dir: local        path /var/lib/vz
        content images,iso,vztmpl,rootdir
        maxfiles 0


nfs: Omni02CT
        path /mnt/pve/Omni02CT
        server 192.168.222.12
        export /vdev1/omni02CT
        options vers=tcp,3
        content rootdir
        maxfiles 0


nfs: Omni02KVM
        path /mnt/pve/Omni02KVM
        server 192.168.222.12
        export /vdev1/omni02KVM
        options vers=tcp,3
        content images
        maxfiles 0


nfs: OpenFiler
        path /mnt/pve/OpenFiler
        server 192.168.222.150
        export /mnt/vg_nfs/nfs-ext3/nfs-datastore
        options vers=3
        content images,backup,rootdir
        maxfiles 1


nfs: backups01
        path /mnt/pve/backups01
        server 192.168.222.11
        export /vdev1/proxBKUP
        options vers=3
        content backup
        maxfiles 3


nfs: backups02
        path /mnt/pve/backups02
        server 192.168.222.12
        export /vdev1/omni02Backup
        options vers=tcp,3
        content backup
        maxfiles 4


nfs: iso01
        path /mnt/pve/iso01
        server 192.168.222.11
        export /vdev1/proxISO
        options vers=3
        content iso
        maxfiles 0


nfs: templates01
        path /mnt/pve/templates01
        server 192.168.222.11
        export /vdev1/proxISO
        options vers=3
        content vztmpl
        maxfiles 0


nfs: Omni01CT
        path /mnt/pve/Omni01CT
        server 192.168.222.11
        export /vdev1/proxCONT
        options vers=3
        content rootdir
        maxfiles 0


nfs: Omni01KVM
        path /mnt/pve/Omni01KVM
        server 192.168.222.11
        export /vdev1/proxKVM
        options vers=3
        content images
        maxfiles 0


nfs: Net30BACKUP
        path /mnt/pve/Net30BACKUP
        server 192.168.222.12
        export /vdev1/omni02Backup
        options vers=3
        content backup
        maxfiles 30


nfs: testCT
        path /mnt/pve/testCT
        server 192.168.222.12
        export /vdev1/omni02CT
        options vers=3
        content rootdir
        maxfiles 0

The pve Version of the node that is not working properly is:

# pveversion
pve-manager/3.1-24/060bd5a6 (running kernel: 2.6.32-26-pve)

The Other Nodes Connect Properly Without Any Issues.
vmbr0 is used to connect to the NFS over bond0
vmbr1 is what I run my standard network

vmbr0 is a storage subnetwork completely seperate from vmbr1

Please, any help Please
I can't force a remount after the system has been booted, so I cannot force the error to happen again without a reboot. Furthermore, just before this happened, I had a CentOS/ZoneMinder CT go crazy with cpu usage, if that helps...

 

[Raymond Burns]

That's ok.
I fixed it.
storage.cfg had "tcp,3" for mount options and it did not work properly. Thanks for reading