So yesterday I had an nfs mount go bad (thought it would be a great idea
to just use nfs for a few minutes to do some file comparisons across
machines) .. And since then the CPU utilizaion on the machine has been
pegged out. The machine still is very responsive.
When the mount went bad ... I couldn't do anything with any process that
was using anything on that volume. I tried a bunch of stuff, but
eventually just shut down NFS (so the processes got an I/O error and
died) and did an "umount -f" on that volume.
Any ideas on how to make this better?
-Marc
CPU0 states: 14.0% user, 85.0% system, 0.0% nice, 0.0% idle
CPU1 states: 50.0% user, 50.0% system, 0.0% nice, 0.0% idle
Mem: 1048088K av, 1023836K used, 24252K free, 0K shrd, 45404K
buff
Swap: 1052248K av, 148K used, 1052100K free 763964K
cached
PID USER PRI NI SIZE RSS SHARE LC STAT %CPU %MEM TIME COMMAND
20099 hughesm 16 0 940 940 696 3 R 2.6 0.0 0:00 top
20074 root 2 0 1716 1704 1188 2 S 0.3 0.1 0:00 sshd
1 root 0 0 484 484 408 3 S 0.0 0.0 2:31 init
2 root 0 0 0 0 0 2 SW 0.0 0.0 0:18 kflushd
3 root 0 0 0 0 0 2 SW 0.0 0.0 9:22 kupdate
4 root 0 0 0 0 0 2 SW 0.0 0.0 9:00 kswapd