rpcbind not running

[Andre Ford]

We are running HP-UX 10.20 and have been experiencing some issues with NFS. While trying to troubleshoot the problem, I run rpcinfo -p and see that rpcbind is not running and there is no /usr/bin/rpcbind executable. How do I get this configured or installed? Is there a patch that I need to load? Thanks.

Andre

[Steve Steel]

Hi

rpcbinf is part of the NFS-CORE and
I suspect that the NFS software is not installed on your box

swlist -l file|grep NFS





Steve Steel

[Alex Glennie]

As far as I'm aware rpcbind came in around 10.30 ... at 10.20 portmapper (man portmap) provided it's functionality..... so that explains the missing binary in /usr/sbin.

So I'd suggest we need to know what issues you are having .... apologies but NFS is not my best subject.

[Ravi_8]

Hi Andre

is the flags of NFS_SERVER and NFS_CLIENT set to 1 in /etc/rc.config.d/nfsconf.

check for commented nfs daemons in /etc/inetd.conf and /etc/services files.
and restart the NFS daemons.

[Uday_S_Ankolekar]

You should be having /usr/sbin/portmap daemon running instead of rpcbind.
I think on 10.20 there was no rpcbind.
Look for portmap process by ps -ef |grep portmap

-USA..

[Andre Ford]

The main problem that we are experiencing has to do with NFS server not responding. I have read the Installing and Administering NFS Services document at docs.hp.com and it had some excellent information in it. That was where I received the list of services that should be running when the rpcinfo -p command is executed. All of the services listed are running except rpcbind. We do have NFS services running and they seem to be working properly. Our problem surfaces when users are attempting to move or copy large files across NFS to other locations. I saw a reference to patch PHNE_25234, which we don't installed, but wasn't sure if this would install rpcbind since I didn't see it listed in the output portion of the patch. I don't know a whole lot about NFS but have picked up quite a bit. Any other information would be helpful.

Andre

[Andre Ford]

Hi Ravi - NFS_SERVER and NFS_CLIENT both = 1. Also NFS (nfsd)is enabled in /etc/services.

Hi Uday - Portmap is currently running.

Is there anything else I need to look at to try and resolve this problem? The document I read has some good info but it looks like it may have been written for versions higher than 10.20. I appreciate the help.

Andre

[Alex Glennie]

I'd suggest in the abscence of any errors to go on that you install PHNE_25234 ; from patch text

PHNE_25234: PANIC HANG
Panic due to a null r_unldvp
Hang when doing a large file copy from a local fs to NFS

[Andre Ford]

Thanks for the heads up. We have loaded patch PHNE_24234 and will run some tests to see if this resolves our problem. Thank you all for your prompt assistance. I will reply back with the results.

Andre

[Jim Keeble]

rpcbind replaces portmap in 11.0 of hp-ux, portmap provides that service in 10.20.

If you are getting "NFS Server not responding" as a result of a mount command or automount file system on the client, rpc.mountd on the server side is not answering. You can verify if it is hung with "rpcinfo -u mount".

If the mount goes ok, but the "not responding" message comes later, it is probably nfsd on the server that can't respond. The same test can be used against nfsd, as in "rpcinfo -u nfs".

If both of these tests result in timeouts, you may not have any connectivity at all, and should back up to ping for troubleshooting.

If both are successful, then you may have a problem with dropped packets from the server to client. Refer to the O'Reilly nfs books for troubleshooting hints using "nfsstat -c" on the client side and "netstat -s -p ip" (particularly the "dropped fragments stat), and "netstat -s -p udp" (looking for udp socket overflow on both client and server).

Good luck !

[Mark Greene_1]

Once you get rcp up and running, you can run this to verify all the processes needed:

UNIX95= ps -efH|head -1; UNIX95= ps -efH|egrep "inetd|rpc|mount|nfs"|grep -v grep; rpcinfo -p|egrep "service|nfs"; showmount -e; mount -p


You should get output similar to this:
UID PID PPID C STIME TTY TIME CMD
root 811 0 0 Dec 13 ? 00:00 nfskd
root 907 1 0 Dec 13 ? 00:02 /usr/lib/netsvc/fs/automount/automount -f /etc/auto_master
root 1404 1 0 Dec 13 ? 02:11 /opt/dce/sbin/rpcd
root 806 1 0 Dec 13 ? 00:00 /usr/sbin/rpcbind
root 896 1 0 Dec 13 ? 00:00 /usr/sbin/rpc.lockd
root 890 1 0 Dec 13 ? 00:00 /usr/sbin/rpc.statd
root 1845 1 0 Dec 13 ? 00:00 /usr/sbin/rpc.mountd
root 1856 1 0 Dec 13 ? 00:00 /usr/sbin/nfsd 4
root 1873 1856 0 Dec 13 ? 00:00 /usr/sbin/nfsd 4
root 1857 1856 0 Dec 13 ? 00:00 /usr/sbin/nfsd 4
root 1874 1857 0 Dec 13 ? 00:00 /usr/sbin/nfsd 4
root 27644 20314 0 11:46:35 ttyp3 00:00 more /etc/rc.config.d/nfsconf
root 23330 1 0 Dec 21 ? 00:24 /usr/sbin/inetd

If you don't have all these processes running, you are going to have NFS problems. Also problems with rcp, rsh, etc.

HTH
mark