show proc cpu misleading RSP4

From: Steven W. Raymond (steven_raymond@eli.net)
Date: Fri Feb 15 2002 - 12:55:59 EST


How come when issuing 'sho proc cpu' one gets a 5 sec, 1 min and 5 min
processor utilization which will be some large number, but that the list
of processes & their corresponding proc time does not add up to the time
interval average?
See below. The 5 minute average is 31%. I have filtered out the zero
values as can be seen. The resulting non-zero time processes add up to
less than say, 5%, but the 5 minute average is 31%. What's taking up
the rest of the CPU? Kernel processes?

Router#sh proc cpu | e 0.00%
CPU utilization for five seconds: 31%/20%; one minute: 32%; five
minutes: 31%
 PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
   3 9134376 452850 20170 5.06% 0.90% 0.72% 0 Check
heaps
  11 4793060 54027297 88 0.49% 0.33% 0.32% 0 IPC Seat
Manager
  41 6189208 33158352 186 0.49% 0.27% 0.29% 0 IP
Input
  90 43215644 744506551 58 1.80% 1.85% 1.87% 0 Multilink
PPP
  91 31468360 229878048 136 1.22% 1.42% 1.39% 0 Multilink
PPP ou
 101 5396328 17159696 314 0.08% 0.26% 0.24% 0 BGP
Router
 102 5369544 8211786 653 0.65% 0.42% 0.39% 0 BGP
I/O

The problem I'm really chasing is PPP Multilink CPU hogging. A small
4-T1 bundle increased an RSP4 from 5-10% avg utilization up to 25-30%
average. DCEF is enabled on all 4 T1s and the "sh run int" for the
multilink2 indicates DCEF is currently turned on. Strangely enough,
router squawks when you say "ip route-cache distributed" when configing
the multilink interface claiming distributed switching is not enabled on
mu2:

Router#conf t
Enter configuration commands, one per line. End with CNTL/Z.
Router(config)#int mu2
Router(config-if)#ip route-cache distributed
Distributed switching not supported on Multilink2
gwRouter#sh run int mu2
Building configuration...

Current configuration:
!
interface Multilink2
...
 no ip directed-broadcast
 ip route-cache distributed
 load-interval 30
 no cdp enable
 ppp multilink
 multilink-group 2
end

Router#

Any recommendations on improving CPU utilization with MLPPP? Thanks for
any advice you have to share.



This archive was generated by hypermail 2b29 : Sun Aug 04 2002 - 04:13:05 EDT