From: T. Horsnell \(tsh\) (tsh@mrc-lmb.cam.ac.uk)
Date: Tue Jan 20 2004 - 07:08:05 EST
Thanks to Martin Andersen.
The reason was right under my nose - 5 runaway jobs, probably following
a power-cut elsewhere in the building.
Cheers,
Terry.
Original msg:
==========================================================================
Hi all,
system: ES45, 4cpu, 8GB, 5.1B PK2
I'm getting very slow cpu rsponse to, in particular,
network-related requests. So far, all I can see that is
abnormal is a very high 'sys int' count in vmstat:
alf1_tsh> vmstat 1
Virtual Memory Statistics: (pagesize = 8192)
procs memory pages intr cpu
r w u act free wire fault cow zero react pin pout in sy cs us sy id
14 1K 47 338K 576K 104K 30G 259M 692M 318K 318M 85K 671 367K 4K 28 8 63
13 1K 49 341K 573K 104K 470 92 111 0 130 0 1K 4M 2K 49 51 0
16 1K 50 344K 570K 104K 269 89 76 0 103 0 1K 5M 2K 50 50 0
The 5M system calls a second sometimes reaches 8M and is normally
between 1 and 5K.
Is there some way I can find the process making these system calls?
So far, looking at the ps options:
minflt, majflt, pagein, nsignals, inblock, oublock
has revealed nothing.
Network activity seems normal:
alf1_tsh> /usr/sbin/ifconfig -a
alt0: flags=202c63<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST,SIMPLEX,PFCOPYALL,MULTINET>
inet 10.1.0.1 netmask ff800000 broadcast 10.127.255.255 ipmtu 1500
inet 131.111.85.78 netmask ffff0000 broadcast 131.111.255.255 ipmtu 1500
lo0: flags=100c89<UP,LOOPBACK,NOARP,MULTICAST,SIMPLEX,NOCHECKSUM>
inet 127.0.0.1 netmask ff000000 ipmtu 4096
sl0: flags=10<POINTOPOINT>
tu0: flags=c63<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST,SIMPLEX>
tu1: flags=c63<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST,SIMPLEX>
tu2: flags=c63<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST,SIMPLEX>
tun0: flags=80<NOARP>
tun1: flags=80<NOARP>
alf1_tsh> /usr/sbin/netstat -i 1
input (alt0) output input (Total) output
packets errs packets errs colls packets errs packets errs colls
149881 0 88347 0 0 19347340 0 239370085 220084279 0
367 0 375 0 0 367 0 417 42 0
497 0 360 0 0 497 0 402 42 0
360 0 387 0 0 360 0 429 42 0
302 0 319 0 0 302 0 361 42 0
304 0 335 0 0 304 0 377 42 0
272 0 272 0 0 272 0 314 42 0
Any ideas anyone?
Cheers,
Terry
----- End of forwarded message from T. Horsnell \(tsh\) -----
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:49:49 EDT