VCS NFS Agent - There should only be one NFS Resource per node.
Article ID: 100020913
Updated On:
Description
Error Message
Sample NFS server error messages on Linux platform:Mar 29 01:30:46 ccing2 nfsd[28259]: nfssvc: Setting version failed: errno 16 (Device or resource busy)
Mar 29 01:30:46 ccing2 nfsd[28259]: nfssvc: unable to bind UPD socket: errno 98 (Address already in use)
Resolution
You should configure only one NFS resource per node. It was found that having duplicate NFS resources per node might cause potential problems. One of the found problems is that if all the NFS resources try to online at the same time, then only the NFS resource started first will successfully online and the rest of the NFS resources will fail. Error messages may be logged by the system like the ones displayed in the Exact Error Message section.
There should only be one single NFS resource configured in a service group on a node. If you have more than one service group that uses the NFS resource, the other service groups should use a Proxy resource. The Proxy resource can point to the NFS resource in the first group.
Please note that the NFS resource is a persistent resource where the resource operation is limited to OnOnly. As a result even when a service group configured with NFS resource is offline, the NFS resource itself will remain online and will be monitored in interval specified in the MonitorInterval attribute which is 60 seconds by default.
If you have duplicate NFS resources configured on your system, please convert them to Proxy resource as advised above. If you can't make the change immediately, please try to lower the NumThreads attribute of the NFS Agent to 1, so that only one agent entry point (e.g. online/offline/monitor/clean) will be executed at one time to minimize the chance to hit potential problems.
There should only be one single NFS resource configured in a service group on a node. If you have more than one service group that uses the NFS resource, the other service groups should use a Proxy resource. The Proxy resource can point to the NFS resource in the first group.
Please note that the NFS resource is a persistent resource where the resource operation is limited to OnOnly. As a result even when a service group configured with NFS resource is offline, the NFS resource itself will remain online and will be monitored in interval specified in the MonitorInterval attribute which is 60 seconds by default.
# hatype -display NFS | egrep 'Operation|MonitorInterval'
NFS MonitorInterval 60
NFS Operations OnOnlyIf you have duplicate NFS resources configured on your system, please convert them to Proxy resource as advised above. If you can't make the change immediately, please try to lower the NumThreads attribute of the NFS Agent to 1, so that only one agent entry point (e.g. online/offline/monitor/clean) will be executed at one time to minimize the chance to hit potential problems.
#haconf -makerw
# hatype -modify NFS NumThreads 1
# haconf -dump-makero
# hatype -display NFS | egrepNumThreads
NFS NumThreads 1
Issue/Introduction
VCS NFS Agent - There should only be one NFS Resource per node.
Additional Information
ETrack: 1651565
ETrack: 1651582
Was this article helpful?
Yes
No
Powered by
