Syslog error message

[vselvaraj]

Dear all

I am having hpux-11.00 on L-1000 server.
I am using as a mail server to sendmails, the data stoarage is common to two servers, in this server i am nfs mounting the FSs from another server, these 2 are identical servers in all aspects and MC/SG clustered.

I was facing some problem whcih came suddenly that one of the NFS mounted file system went to read only permission (totally 3 FSs are nfs mounted).Due to that my services get stopped. NFS export and mount all are done through MC/SG control scripts. In script there is no change for the past 6 months. So I could manually stop the cluster and started again , and now it is OK.

I wanted to know why it went to Readonly access suddenly, to avoid the future problems.

The syslog file says some error , what exactly it means.........In what circumstances this message will come? Is there any relation between this message and my problem
Can anyone find any solution/suggestion please welcome
Thanks in advance
V.Selvaraj

[Stefan Farrelly]

There are many reasons that could cause an RPC: Timed out message.
In your case you state your servers are running normally which means there isnt a setup problem, and that they havent changed for months, which is also good.
So, in your case your HP server called chend1 lost contact with the NFS server 202.169.129.35 via RPC (remote procedure call).
ie. chend1 was trying to contact the nfs server daemons on 202.169.129.35 and couldnt. This could possible be because the nfs daemons stopped working normally (they died, someone killed them, they became overloaded)

or the network connection from chend1 to 202.169.129.35 became saturated (a network fault, someone swamping the network with say a massive file copy)

or your nfs client daemons on chend1 stopped working normally due to chend1 being overloaded - too many nfs requests, HP-UX running slowly due to other processes etc.

Can you backtrack and check your performance data on 202.169.129.35 and chend1 and the network at the time of the problem ? this is about the only way to try to ascertain what caused it.

[vselvaraj]

Hi Stefan,
thanks for ur reply.
I found one more thing..
that is the IP address (.35) is the package IP address of
the NFS server, when the package is running ie .35 will be enabled, at that time only
that server is nfs exporting the FSs.

From my NFS client ie chend1 , if I give nslookup for the IP .35 it says unknown host,
and its name if I give IP is comming ie .35, So the one way of resolve is not happening. Now I corrected by making same entry which is there in NFS -server's /etc/resolv.conf to here in NFS client chend1 M/C. This I had done then I restarted the cluster.

In any case this name lookup will create problem....
Selvaraj