FW: homedirs and NFS

[Larry Smith_12]

Hi,

I am running into the same problem described in this old thread (see my entry at the bottom of the thread):

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=202873

In the case above, the author appears to believe that the issue is related to phisical network problems. In my case, I am sure that I am not experiencing any phisicaly network problems. Another difference is that my client is running HPUX 11.11.

Has anyone ran across this same problem and found a solution for it?

[RAC_1]

It would have been good, If you would have posted the problem here. I read the other thread. What makes you think that, you do not have network problem??

Things to check.

1. Are you up to date on NFS patches??
(The error 238 is indeed connection timeout error-check /usr/include/sys/errno.h)
2. Are nic speed settings correct on both ends?? (nic and switch/hub)
3. When you experience this problem, try pinging nfs server and check if you experience any problem. (It would help if you run a ping to nfs server contineously and check if gives time outs when you experience the problem)
4. Did you check nfsstat -s and nfsstat -c
(Check details with man nfsstat)
5. For a test, you may want to put user's .sh_history file in local FS. -modify .profile and check) Does it gives the user prompt fast??

Anil

[Larry Smith_12]

I believe I am not having phisical network problems because we have checked all the cables and jacks. I am going to check them again though. I have to add that the problem is intermitent. The system works fine for days and all of the sudden it presents this behavior and it completely halts for hours. It does sound like something that could be caused by a phisical network problem. I will check the cables again.

>> 1. Are you up to date on NFS patches??
>> (The error 238 is indeed connection
>> timeout error-check
>> /usr/include/sys/errno.h)

Yes, we are up to date on NFS patches. However, I just found out today after posting this question that there is a new patch for the network card that we are using. We will be trying this patch in the next few days. I will let you know how it goes.

>> 2. Are nic speed settings correct on both
>> ends?? (nic and switch/hub)

Yes, we checked this and made sure that everything was setup correctly.

>> 3. When you experience this problem, try
>> pinging nfs server and check if you
>> experience any problem. (It would help if
>> you run a ping to nfs server contineously
>> and check if gives time outs when you
>> experience the problem)

We have a system that pings our NFS servers every 5 minutes and notifies us it detects a problem. The logs of the system show no ping failures. Also, we are having no problems with of our Solaris clients.

>> 4. Did you check nfsstat -s and nfsstat -c
>> (Check details with man nfsstat)

Yes, and I did notice an odd thing. The values of timeout and retrans are very high which would also indicate some sort of network failure. I am not sure why timeout for "Connection Oriented" is 0 though. Should it not be high as well?

> nfsstat -rc

Client rpc:
Connection oriented:
calls badcalls badxids
6655 0 0
timeouts newcreds badverfs
0 0 0
timers cantconn nomem
0 0 0
interrupts
0
Connectionless oriented:
calls badcalls retrans
203 2 1509
badxids timeouts waits
0 1509 0
newcreds badverfs timers
0 0 0
toobig nomem cantsend
0 0 0
bufulocks
0

>> 5. For a test, you may want to put user's
>> .sh_history file in local FS. -modify
>> .profile and check) Does it gives the user
>> prompt fast??

I am going to give it a try.

Thanks

[Andrew Cowan]

Have you checked your other network statistics. Are there a high number of retries and collisions? This shounds to me like a faulty piece of cable or bad switch port.

[T G Manikandan]

I would add with

1.check the NFS version on both sides

2.check also the duplex settings at both the server ends and the switch.