NFS client trouble after upgrade to kernel 2.2

Mark H. Wood (mhwood@ameritech.net)
Sat, 13 Mar 1999 11:31:47 -0500 (EST)


I recently upgraded the box that I use for dialout routing from 2.0.29 to
2.2.1. The box is not totally diskless, but I mount /usr and /home
across NFS from another Linux box (kernel 2.1.129, usermode NFS server
2.2beta37). The router system had no complaints using this approach when
running kernel 2.0. I made sure that I have the latest mount (2.9i).

After the kernel upgrade, however, I now get stalls and errors when it
mounts the NFS volumes:

Mar 12 20:07:44 woodshed kernel: portmap: RPC call returned error 111
Mar 12 20:07:44 woodshed kernel: RPC: task of released request still queued!
Mar 12 20:07:44 woodshed kernel: RPC: (task is on xprt_pending)
Mar 12 20:07:44 woodshed kernel: portmap: RPC call returned error 111
Mar 12 20:07:44 woodshed kernel: RPC: task of released request still queued!
Mar 12 20:07:44 woodshed kernel: RPC: (task is on xprt_pending)
Mar 12 20:07:44 woodshed kernel: lockd_up: makesock failed, error=-111
Mar 12 20:07:44 woodshed kernel: portmap: RPC call returned error 111
Mar 12 20:07:44 woodshed kernel: RPC: task of released request still queued!
Mar 12 20:07:44 woodshed kernel: RPC: (task is on xprt_pending)
Mar 12 20:07:45 woodshed kernel: lockd_up: no pid, 2 users??
Mar 12 20:07:45 woodshed kernel: portmap: RPC call returned error 111
Mar 12 20:07:45 woodshed kernel: RPC: task of released request still queued!
Mar 12 20:07:45 woodshed kernel: RPC: (task is on xprt_pending)
Mar 12 20:07:45 woodshed kernel: portmap: RPC call returned error 111
Mar 12 20:07:45 woodshed kernel: RPC: task of released request still queued!
Mar 12 20:07:45 woodshed kernel: RPC: (task is on xprt_pending)
Mar 12 20:07:45 woodshed kernel: portmap: RPC call returned error 111
Mar 12 20:07:45 woodshed kernel: RPC: task of released request still queued!
Mar 12 20:07:45 woodshed kernel: RPC: (task is on xprt_pending)
Mar 12 20:07:49 woodshed kernel: lockd: failed to monitor 192.168.1.2

After all that, NFS seems to work pretty well anyway (but read on). Any
ideas? I can't even find a '*lockd*' executable -- where is this last
one coming from?

Also I now get errors when trying to run a.out executables fetched via NFS:

N_TXTOFF < BLOCK_SIZE. Please convert binary.

I found some earlier messages about this that suggested problems with
large NFS transfer sizes, but cutting back to 1024 didn't help.

I know, I know -- I should rebuild all the a.out commands as ELF, and
I've done those that I have been able to find. (My systems are Slackware
1.2 plus five years of piecemeal upgrades -- tell me you *know* that
pasting 3.5 over them won't break a basketful of things I've set up in
the interval.) Some of them are proving hard to find.

Please don't tell me to upgrade the server to beta40 unless you also tell
me where to find it. The site listed in CHANGES only has beta37, the
version I'm running now.

aTdHvAaNnKcSe

-- 
Mark H. Wood, radical centrist                             mhwood@ameritech.net
Charlie, put down that Glitter Glue -- it's time to show the audience
some content!

- To unsubscribe from this list: send the line "unsubscribe linux-net" in the body of a message to majordomo@vger.rutgers.edu