MSFC CPU Utilization Pegged at 99%....but a sh proc cpu does not reveal
anything past 1%The problem is that your CPU is spending 82% of it's time
forwarding traffic at interrupt level (the number to the right of the
overall CPU load).
Did you accidentally disable MLS?
If the PFC can't build a flow entry via MLS, then the MSFC will get stuck
forwarding all of the L3 traffic.
David
-----Original Message-----
From: Greene, Patrick [mailto:PJGreene@infotechent.net]
Sent: Friday, February 02, 2001 12:34 PM
To: 'Martin, Christian'; cisco-nsp@puck.nether.net
Subject: RE: [nsp] MSFC CPU Utilization Pegged at 99%....but a sh proc cpu
does not reveal anything past 1%
CPU utilization for five seconds: 99%/82%; one minute: 99%; five minutes:
99%
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
1 176 3721 47 0.00% 0.00% 0.00% 0 Load Meter
2 1436 8675 165 0.00% 0.00% 0.00% 0 OSPF Hello
3 147912 7464 19816 4.54% 0.97% 0.88% 0 Check heaps
4 0 1 0 0.00% 0.00% 0.00% 0 Chunk Manager
5 3416 41959 81 0.00% 0.00% 0.00% 0 Pool Manager
6 0 2 0 0.00% 0.00% 0.00% 0 Timers
7 0 2 0 0.00% 0.00% 0.00% 0 Serial
Backgroun
8 1372 14030 97 0.00% 0.00% 0.00% 0 EnvMon
9 1276 11934 106 0.00% 0.00% 0.00% 0 ARP Input
10 0 2 0 0.00% 0.00% 0.00% 0 DDR Timers
11 0 2 0 0.00% 0.00% 0.00% 0 Dialer event
12 0 1 0 0.00% 0.00% 0.00% 0 Entity MIB API
13 0 1 0 0.00% 0.00% 0.00% 0 SERIAL
A'detect
14 0 1 0 0.00% 0.00% 0.00% 0 Critical Bkgnd
15 96 2500 38 0.00% 0.00% 0.00% 0 Net Background
16 4372 29869 146 0.00% 0.00% 0.00% 0 Logger
17 2144 14026 152 0.08% 0.01% 0.00% 0 TTY Background
18 312 14030 22 0.00% 0.00% 0.00% 0 Per-Second
Jobs
19 22512 84985 264 0.08% 0.06% 0.06% 0 IP Input
20 216 2301 93 0.00% 0.00% 0.00% 0 CDP Protocol
21 0 5 0 0.00% 0.00% 0.00% 0 SCP async:
TCAM
-----Original Message-----
From: Martin, Christian [mailto:cmartin@gnilink.net]
Sent: Friday, February 02, 2001 3:08 PM
To: 'Greene, Patrick'; cisco-nsp@puck.nether.net
Subject: RE: [nsp] MSFC CPU Utilization Pegged at 99%....but a sh proc
cpu does not reveal anything past 1%
Is this spent at the Interrupt level? Recall that the process
utilization adds up to the difference between total 5 second CPU
utilization, and time spent at the interrupt level switching packets. It is
likely that MLS is hosed, or CEF is busted/off on the MSFC2, and the CPU is
switching all the packets in the fast or process path. Since IP_Input
appears idle, then it is likely fast switching. Can you snip over a show
proc cpu?
chris
-----Original Message-----
From: Greene, Patrick [mailto:PJGreene@infotechent.net]
Sent: Friday, February 02, 2001 2:48 PM
To: cisco-nsp@puck.nether.net
Subject: [nsp] MSFC CPU Utilization Pegged at 99%....but a sh proc cpu
does not reveal anything past 1%
Have a 6509 with 2 MSFC's. The MSFC's are running 12.1(2)E. Just had
one
of the MSFC's spike to 99%. A sh proc cpu reveals that total
utlization
should add up to about 5%. A reload of the module causes the exact
same
thing. A no changes have been made. Anybody seen this before. Also,
the
other MSFC and Switch Utilization are both showing about 3-5%, which
is
normal.
Thank You,
Patrick Greene
This archive was generated by hypermail 2b29 : Sun Aug 04 2002 - 04:12:27 EDT