Auto home issue

[Ronelle A van Niekerk]

I am stumped.
At some stage today, suddenly no one could log into "any" of the unix servers.
I've checked a lot of them with my own login and it's sort of true.
Seems the SUN servers are ok, so is the AIX server and the two new itanium 11.31 servers.

Those being affected are all the other HP servers - 11.00, 11.11, 11.23 (itanium and risc)

It seems to be intermittent, but errs to the side of not working.
I've ruled out NIS, because the user does get authenticated.

When I try to su to, or log in directly as, my user, the login goes as far as to when it would bring up the prompt, but then it just hangs.
I can manually mount the home directory.

I have changed nothing on the NFS server, hadn't logged into it for weeks. And obviously didn't change anything on so many of the other servers.

I've checked the /etc/auto_home on a lot of the servers and it's right, besides which, it was working earlier today.

Anyone ahve any ideas?

[Robert-Jan Goossens · Accepted solution]

You can login with a local user (root), but when using a non-local user the login hangs?

Sound a bit like this NFS thread, have a look at Dave Olkers comments.

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1180871

Regards,
Robert-Jan

[Ronelle A van Niekerk]

OK, the ksh lock thing could be a possibility, but why now all of a sudden?
Why all these systems and all the users?
And why not other systems?

And is there another way of getting past it, seems all users are affected, and I can't go edit all their .profile's

[Ronelle A van Niekerk]

OK, it's definitely this lock thing, when I changed my user shell to bash it logged in just fine.

the rpc.lockd and statd log files don't tell me a whole lot.

And I don't understand why this is only affecting some of the servers.

[Ronelle A van Niekerk]

So how do I fix this issue?
I've tried restarting all the nfs services on the nfs server and that didn't help.

[Kenan Erdey]

Hi,

have a look at this thread too. at the and there is a cheat to put /etc/profile. but also some search to fix the problem

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1105150

[Ronelle A van Niekerk]

There are no answers to FIX the problem.
I have run the debug on the rpc daemons, but I haven't got a clue as to what they mean.

[Ronelle A van Niekerk]

I restarted the nfs services on the nfs server - which caused all sorts of problems on the clients and I had to reboot several of them - but it seems to be working a bit better now:
100% on some of them (the ones that weren't working before) and VERY slowly on the others.

[Bill Hassell]

Patches, patches. Start by patching the server to the recommended level--not just NFS but all networking patches. Then patch all the clients using the same criteria. NFS usually works OK as long as you never change anything. That means anything including hosts on the networks. Auto-mounted filesystems are particularly sensitive to change, especially at the server.

[Ronelle A van Niekerk]

That's just great!
I can patch the server, but I can't patch the clients.
No way, no how...

I will see what patching the server does.