NFS readdir+ failed for server <hostname>: RPC: Timed out

[Subhash Pippala]

Hi all,

We are frequently getting the following NFS error messages on our 9000/800 HP-UX servers while accessing remote HP servers,and all the available mount options have been tried in fstab file.

1.NFS readdir+ failed for server : RPC: Timed out
2.NFS getattr failed for server : RPC: Timed out
3.NFS read failed for server : RPC: Timed out

Please suggest.

Thanks & Regards,
Subhash Pippala

[John Waller]

Are the two machines with the NFS directories on a local network or split over a wide area network. The reason I ask is that I've seen this problem before and it was due to a network performance issue. This message would normally appear on the NFS client and it is indicating that it timed out while trying to write data.

I have been wondering and perhaps somebody could confirm it, but would increasing the number of nfsd's running on the server help the reduce the problem.

[Subhash Pippala]

Hi John,

Servers are split over a wide area network. We have even tried the option of increasing the nfsd daemons to 40 but still the problem continues.

Thanks
Subhash Pippala

[Santosh Nair_1]

There is a document available from HP's ITRC which helps troubleshooting NFS problems. The doc id is #KBAN00000261.

-Santosh

[John Waller]

Since the machines are seperated by a Wide Area network , have you tried a ping between the two servers. You may be encountering a high %'age package lost or a long response time. Try setting up a cron job to run a job every 1 or 2 hours which performs and records a ping -n 300 (run a ping for aprox 5 minutes). This will may give you an idea of hour your WAN is performing.

[Sanjay_6]

Hi,

Do you have the latest NFS performance patch on your system. I hope you are using 11.0 since you have not mentioned the OS level.

http://us-support.external.hp.com/wpsl/bin/doc.pl/screen=wpslDisplayPatch/sid=a617afec147fc99266?PACH_NAM=PHNE_24034&HW=s800&OS=11.00

Hope this helps.

thanks