I had an NFS client hang today after taking down the server and bringing
it back up.  The server was down for about a half hour.  The client is a
UW-IMAP/POP server and the mail spool is one of the NFS mounts.
After taking down the server I got the usual message:
kernel: nfs: server <server name> not responding, still trying
Then I started to get messages:
kernel: nfs: task <number> can't get a request slot
I looked on Bugzilla and saw a suggestion that these might be related to
the network card.
After I brought the server back up these messages stopped and I got the
usual message:
kernel: nfs: server <server name> OK
This was repeated 100+ times.
Then I started to get messages:
kernel: statd: server localhost not responding, timed out
kernel: lockd: cannot monitor <server IP>
kernel: lockd: failed to monitor <server IP>
over and over again.  IMAP and POP stopped responding.  I had an ssh
connection to the server (Commercial SSH client).  I tried to run "ps
ax" but it hung.  I tried to open a second window on the connection. 
That hung too.  I tried rebooting but init just sat there and didn't do
anything.  Finally I had to push the power button.
I've increased fs/file-max since it appears I had been running fairly
close to it in normal operation.  The "ps ax" hang makes me wonder
whether I'm running out of space in the process table, although I
suppose a network problem would cause similar symptoms (the NFS server
is also the LDAP server).  Should I increase kernel/threads-max?

Server and client are IBM xSeries 230, network card is PCnet/FAST III
79C975, kernel 2.4.9-31, nfs-utils 0.3.1-5.  Naturally RH 7.1 or I
wouldn't be writing to this list.
Thanks,
John Dalbec



_______________________________________________
Seawolf-list mailing list
[EMAIL PROTECTED]
https://listman.redhat.com/mailman/listinfo/seawolf-list

Reply via email to