NFS warning [message #22337] |
Thu, 25 October 2007 19:03 |
whately
Messages: 4 Registered: October 2007
|
Junior Member |
|
|
Hello All,
I want to use a NFS server for the /vz partition.
I mounted it on the VE0 of each HN. The server is running CentOS 5, without openvz. The HNs have openVZ RHEL 5 based, release 2.6.18-8.1.8.el5.028stab039.1
Everything is working ok, except that for every VE start and stop, the following messages are issued:
lockd: cannot monitor <nfs_server address>
I disabled the DISK_QUOTA.
I search this forum and the list, but I couldn't find any explanation for this kind of problem.
Should I give any other information ?
I don't have any other idea to fix it. I accept any suggestion.
Thanks in advance,
Lauro.
|
|
|
|
|
|
|
|
Re: NFS warning [message #22757 is a reply to message #22756] |
Fri, 02 November 2007 19:46 |
seanfulton
Messages: 105 Registered: May 2007
|
Senior Member |
|
|
Do you use any sort of HA-failover mechanism?
What I was trying to do was set up DRBD between two machines (works fine), set up HA-NFS (works fine), then put my VEs on the NFS drive. In this way, if NFS server A fails, HA will switch to NFS Server B which has identical data through DRBD and all will be well.
Everything worked extremely well until I started noticing data corruption issues in the VEs. I was told at the that there was a kernel issue (I think it was the 29 kernel). The 31 kernel supposedly fixed this, but at the time one of the OpenVZ guys suggesting not using NFS because there were other problems with it, to use GFS or something else.
Looking for first-hand stories or experience.
sean
|
|
|