High cpu load (system) 1

[jmiturbe]

Hello all,

We have a Sun Fire 280R with one CPU. The application running on it has been stopped but the machine shows a high system CPU load:

machine1:/data># vmstat 2 5
procs memory page disk faults cpu
r b w swap free re mf pi po fr de sr m0 m1 m4 m5 in sy cs us sy id
1 0 0 5009632 921552 0 0 0 0 0 0 0 0 0 0 0 108 40 58 0 69 31
4 1 0 5018768 929528 0 1 0 0 0 0 0 0 0 0 0 84 11 6 0 100 0
4 0 0 5018768 929528 0 0 0 0 0 0 0 0 0 0 0 119 51 12 0 99 1
2 0 0 5018768 929528 0 0 0 0 0 0 0 0 0 0 0 107 13 7 0 100 0
2 0 0 5018768 929528 0 0 0 0 0 0 0 0 0 0 0 141 79 50 0 99 1
machine1:/data>#

I want determine the system process (or processes) that is taking the CPU all the time from yesterday. I supose something is hung, but I don't know how to see what is happening.

Can anybody help me?

Thanks in advance,

jmiturbe

 

[KenCunningham]

Do you have top on your system? If not you can download it from

http://www.sunfreeware.com

 

[mdet]

Try "ps" with options "pcpu" and "time":
ps -e -o pid -o time -o pcpu -o args
Look out for the one with the highest value.

 

[jmiturbe]

Hello,

Do those commands display system or kernel CPU use?
I already tried them, but it seems like they only show user processes:

load averages: 3.78, 3.68, 3.56 12:19:54
24 processes: 20 sleeping, 3 running, 1 on cpu
CPU states: 0.1% idle, 0.0% user, 99.6% kernel, 0.3% iowait, 0.0% swap
Memory: 1024M real, 907M free, 13M swap in use, 4901M swap free

PID USERNAME THR PRI NICE SIZE RES STATE TIME CPU COMMAND
1537 root 1 59 0 2152K 1480K cpu 0:00 0.08% top.5.8
82 root 12 59 0 3072K 2872K run 0:49 0.01% picld
1197 root 1 59 0 1784K 1088K run 0:02 0.00% ttymon
395 root 14 53 0 3288K 2584K sleep 0:02 0.00% nscd
405 nobody 1 59 0 2320K 1752K sleep 0:02 0.00% lic.SUNW
447 root 1 59 0 1032K 680K run 0:00 0.00% utmpd
1494 co01378l 1 49 0 1880K 1320K sleep 0:00 0.00% ksh
464 root 1 49 0 3176K 1272K sleep 0:00 0.00% sendmail
1498 root 1 49 0 1848K 1248K sleep 0:00 0.00% ksh
379 root 1 49 0 1968K 1232K sleep 0:00 0.00% cron
368 root 1 49 0 1808K 1184K sleep 0:00 0.00% inetd
59 root 5 51 0 1336K 824K sleep 0:00 0.00% syseventconfd
57 root 12 52 0 1568K 1240K sleep 0:00 0.00% syseventd
470 root 1 52 0 0K 0K sleep 0:00 0.00% efdaemon
359 root 1 54 0 2776K 1520K sleep 0:00 0.00% sshd


Any ideas on how to look for the process using 99% of CPU?

jmiturbe

 

[grega]

try

top -o cpu

Greg.

 

[jmiturbe]

Hello again,

The top screen I included earlier was ordered by cpu usage, but it seems like it only displays user processes. The user CPU load is nearly 0% because as I said, we've stopped our application.

How can I get more info about the problem that is causing a continuous 99% kernel CPU usage?

Is there any way to look something when the problem is not caused by user processes, or do I have to reboot to solve the situation without determining it's source?

Thanks again,

jmiturbe

 

[daFranze]

if you are running Solaris 8 or 9 try
prstat -a

Regards
-- Franz
Sorry I'm not a native spaeker, I'm from Munich, Germany - "Home of the Whopper", oh no, "Home of the Oktoberfest" ;-)