+--- In our lifetime, dheiland@lucent.com wrote: | | Are these NFS server not responding messages occurring at any specific | time? I'm seeing similar messages about 3am each night - the same time | that a cron runs to remove .nfs files. I'm still looking into and for me | it might just be a matter of changing from soft mounts to hard mounts, or | even problems with NIS binding.
No, they occur at all times:
Sep 4 19:03:09 web22 unix: NFS server maytag.private.cimedia.net not responding still trying Sep 4 19:03:09 web22 unix: NFS server maytag.private.cimedia.net ok Sep 4 19:39:04 web22 unix: NFS server maytag.private.cimedia.net not responding still trying Sep 4 19:39:04 web22 unix: NFS server maytag.private.cimedia.net ok Sep 4 21:01:05 web22 unix: NFS server maytag.private.cimedia.net not responding still trying Sep 4 21:01:05 web22 unix: NFS server maytag.private.cimedia.net ok Sep 4 21:47:04 web22 unix: NFS server maytag.private.cimedia.net not responding still trying Sep 4 21:47:04 web22 unix: NFS server maytag.private.cimedia.net ok Sep 4 21:51:06 web22 unix: NFS server maytag.private.cimedia.net not responding still trying Sep 4 21:51:06 web22 unix: NFS server maytag.private.cimedia.net ok
(this is from a heavy-use client).
And nothing in any of the filer logs (I have most all levels of syslog turned on):
Thu Sep 4 18:00:37 EDT [statd]: 6:00pm up 26 days, 12:03 -1410149907 NFS ops, 0 CIFS ops, 0 HTTP ops Thu Sep 4 19:00:32 EDT [statd]: 7:00pm up 26 days, 13:03 -1405902688 NFS ops, 0 CIFS ops, 0 HTTP ops Thu Sep 4 20:00:17 EDT [statd]: 8:00pm up 26 days, 14:02 -1400850333 NFS ops, 0 CIFS ops, 0 HTTP ops Thu Sep 4 21:00:39 EDT [statd]: 9:00pm up 26 days, 15:02 -1396090158 NFS ops, 0 CIFS ops, 0 HTTP ops Thu Sep 4 22:01:07 EDT [statd]: 10:01pm up 26 days, 16:01 -1391841400 NFS ops, 0 CIFS ops, 0 HTTP ops
It was suggested to me that it may be a client problem. I am hoping to try some things on a client or two to test. But when only the filer stops passing data on the FDDI (non-pingable), my first inclination is to look at what may be wrong between the filer and the rest of the clients.
Since the problems of the filer "falling off" the FDDI are infrequent, it is hard to debug. They have occured at all sorts of times, with nothing in particular triggering the event.
Alexei