Hi All,
I run an openmosix cluster using Mandrake, and had
problems with PS hanging or whole system for that
matter. It is still an ongoing issue for a while now,
i took it up to openmosix guys, and response is that
90% of the time its network related, the rest 10%
hardware!
Tracing down the problem today via /proc, trying to
read every cmdline for a process, it hangs on
PID=25583.
Ran strace for `ps`, hanged same location (log below)
open("/proc/25583/status", O_RDONLY) = 7
read(7, "Name:\tchange_missing\nState:\tD (d"..., 511)
= 463
close(7)
munmap(0x40013000, 4096) = 0
open("/proc/25583/cmdline", O_RDONLY) = 7
THEN:
cd'ed into /proc/25583
`ls` (look at exe output)
lrwxrwxrwx 1 stuart stds 0 Oct 28 14:16
exe ->
/targus/linuxfiles/stuart/patches/ver006/exp03/.nfs000a839e00000001
(deleted)
Not sure what that deleted line means, and all other
procs don't have it. However, BIGGER problem is that
this process 25583 cant be killed!! Not kill-9 or kill
-KILL that helps. Its currently owned by root.
Is there any other way to kill this sucker, other than
a reboot? Tired of rebooting every 2-3days like
windoze!
-Thanks All,
Richard
__________________________________
Do you Yahoo!?
Exclusive Video Premiere - Britney Spears
http://launch.yahoo.com/promos/britneyspears/
Want to buy your Pack or Services from MandrakeSoft?
Go to http://www.mandrakestore.com
