2008-04-02 02:19 PM
We have a database app that has its storage on the filer. Once in a while on one of the machines
where it runs we see a stream of messages in /var/log/messages
kernel: lockd: failed to monitor <IP Address of filer>
kernel: lockd: cannot monitor <IP address of filer>
What would cause these error messages? The machine where this is seen is
Linux HOSTNAME 2.6.9-34.ELsmp #1 SMP Fri Feb 24 16:56:28 EST 2006 x86_64 x86_64 x86_64 GNU/Linux
2008-04-04 01:28 PM
I'm not a linux expert, but it sounds to me that you have a problem with the rpc.statd daemon. Can you check if it's running properly?
Is it Oracle or mysql or something else? When this error occur, what's the behavior of your DB at that moment?
2008-04-07 12:20 AM
<span class="small"> Hi Sameer,
Make sure that rpc.statd is running on the Linux machine, and the port is being blocked somehow (tcp wrappers, access lists, etc). On Linux, you should see rpc.statd when you run:
linux-rtp:~# ps -ef | grep stat root 384 1 0 Jun04 ? 00:00:04 /sbin/rpc.statd
root 25487 25410 0 17:17 pts/1 00:00:00 grep stat
To enable NLM lock debugging on Linux, enter:
sysctl -w sunrpc.nlm_debug=65
Messages should appear in your system's log corresponding to the locking events.
To turn off NLM debugging on Linux, enter
sysctl -w sunrpc.nlm_debug=0
If rpc.lockd is not running on the Linux host and mounted with lock ( default option ). Then you will see a message similar to the following ( when debug output is enabled ):
dell2400-rtp2 kernel: NFS locking service started (ver 0.4). dell2400-rtp2 kernel: lockd: nlmsvcmarkresources dell2400-rtp2 kernel: lockd: cannot monitor 10.60.4.20 dell2400-rtp2 kernel: lockd: failed to monitor 10.60.4.20
Hope this helps