Discussion:
[c-nsp] 3750 stack with high CPU from IP Input process.
Bill
2005-12-23 22:02:19 UTC
Permalink
Hi,

I have a 3750 stack that has 3 Cisco 3750's in it. Attached to the stack
via a 1 gig trunking port is a Cisco 2950. That 2950 doesn't have anything
plugged into it except a 24 port Alcatel switch. I recently moved that
Alcatel switch from the 2950 to the 3750. That left the 2950 empty. For some
reason the 3750 CPU went from loafing at about 5 or 6% CPU to 50% CPU.
Sometimes even higher. I can't figure out why but I did notice that when I
do a "show interface stats" it shows that all of the interfaces are doing
"process switching". Fast and CEF switching is enabled on all interfaces but
according to the "show interface stats" the stack is doing "process
switching". I assume this is the reason for the high CPU usage.

This switch stack has 1 routed interface as the gateway for the entire
stack. It connects to a 3662 router. And it has 3 VLANs for routing specific
networks.

I've talked to Cisco and the only answer I get is change the "SDM
PREFER" to "routing" and reboot. That's a bunch of bilogna because the stack
ran at 5 or 6% before I moved that ethernet cable to the 3750. Earlier today
I moved the Alcatel ethernet cable back to the 2950 and the CPU continues to
run high. That 2950 loafs along at 1 or 2% but doesn't do any routing. The
load thru the 3750 is the same no matter where that Alcatel ethernet is
plugged in at.

Anyone have any ideas? So far Cisco tech has been no help.

Bill

_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
Brant I. Stevens
2005-12-23 22:41:59 UTC
Permalink
What version of code are you running? Some of the early 12.2.25 code (B
family, IIRC) had bugs with high CPU utilization for no apparent reason.

-----Original Message-----
From: cisco-nsp-***@puck.nether.net
[mailto:cisco-nsp-***@puck.nether.net] On Behalf Of Bill
Sent: Friday, December 23, 2005 5:02 PM
To: cisco-***@puck.nether.net
Subject: [c-nsp] 3750 stack with high CPU from IP Input process.

Hi,

I have a 3750 stack that has 3 Cisco 3750's in it. Attached to the stack
via a 1 gig trunking port is a Cisco 2950. That 2950 doesn't have anything
plugged into it except a 24 port Alcatel switch. I recently moved that
Alcatel switch from the 2950 to the 3750. That left the 2950 empty. For some
reason the 3750 CPU went from loafing at about 5 or 6% CPU to 50% CPU.
Sometimes even higher. I can't figure out why but I did notice that when I
do a "show interface stats" it shows that all of the interfaces are doing
"process switching". Fast and CEF switching is enabled on all interfaces but
according to the "show interface stats" the stack is doing "process
switching". I assume this is the reason for the high CPU usage.

This switch stack has 1 routed interface as the gateway for the entire
stack. It connects to a 3662 router. And it has 3 VLANs for routing specific
networks.

I've talked to Cisco and the only answer I get is change the "SDM
PREFER" to "routing" and reboot. That's a bunch of bilogna because the stack
ran at 5 or 6% before I moved that ethernet cable to the 3750. Earlier today
I moved the Alcatel ethernet cable back to the 2950 and the CPU continues to
run high. That 2950 loafs along at 1 or 2% but doesn't do any routing. The
load thru the 3750 is the same no matter where that Alcatel ethernet is
plugged in at.

Anyone have any ideas? So far Cisco tech has been no help.

Bill

_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
Saku Ytti
2005-12-25 08:46:55 UTC
Permalink
Post by Bill
Sometimes even higher. I can't figure out why but I did notice that when I
do a "show interface stats" it shows that all of the interfaces are doing
"process switching". Fast and CEF switching is enabled on all interfaces but
according to the "show interface stats" the stack is doing "process
switching". I assume this is the reason for the high CPU usage.
Actually if you see process switching only, you're pretty safe, all
traffic that you handle in CPU should be destined to you, which should
mean process switching. Hardware processed traffic doesn't appear in
the counters you mention.
Post by Bill
This switch stack has 1 routed interface as the gateway for the entire
stack. It connects to a 3662 router. And it has 3 VLANs for routing specific
networks.
What SDM profile are you running? How many MAC addresses? How many routes?
Do you have ACL's, if so what do they contain (log statement, lot of
port ranges?).
Post by Bill
Anyone have any ideas? So far Cisco tech has been no help.
What IOS are you running? Can you upgrade to 12.2(25)SED1 (if you do,
remember that it's standard MST since SEC or so)
--
++ytti
_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
VCI Help Desk
2005-12-25 16:22:42 UTC
Permalink
Ok, but my point is why did the CPU go up at all? The same amount of
traffic was going thru the switch no matter what port that ethernet was
connected. The CPU went from 5% to 50% just because I moved an ethernet
cable. When I moved it back the CPU didn't go down.

Bill



----- Original Message -----
From: Saku Ytti
To: Bill
Cc: cisco-***@puck.nether.net
Sent: Sunday, December 25, 2005 2:46 AM
Subject: Re: [c-nsp] 3750 stack with high CPU from IP Input process.
Post by Bill
Sometimes even higher. I can't figure out why but I did notice that when I
do a "show interface stats" it shows that all of the interfaces are doing
"process switching". Fast and CEF switching is enabled on all interfaces but
according to the "show interface stats" the stack is doing "process
switching". I assume this is the reason for the high CPU usage.
Actually if you see process switching only, you're pretty safe, all
traffic that you handle in CPU should be destined to you, which should
mean process switching. Hardware processed traffic doesn't appear in
the counters you mention.
Post by Bill
This switch stack has 1 routed interface as the gateway for the entire
stack. It connects to a 3662 router. And it has 3 VLANs for routing specific
networks.
What SDM profile are you running? How many MAC addresses? How many routes?
Do you have ACL's, if so what do they contain (log statement, lot of
port ranges?).
Post by Bill
Anyone have any ideas? So far Cisco tech has been no help.
What IOS are you running? Can you upgrade to 12.2(25)SED1 (if you do,
remember that it's standard MST since SEC or so)
--
++ytti

_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
Bill
2005-12-25 17:50:18 UTC
Permalink
Here is the "show proc cpu" The CPU usage is low right now. The other
day something was going on and the CPU went 10 100% for no reason I could
find. I can't find any reason why the CPU would have gone up when I moved
the ethernet cable.

Bill


esw-1#show proc cpu
CPU utilization for five seconds: 27%/6%; one minute: 27%; five minutes: 25%
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
1 0 21 0 0.00% 0.00% 0.00% 0 Chunk Manager
2 209 674292 0 0.00% 0.00% 0.00% 0 Load Meter
3 2442168 51948474 47 0.00% 0.02% 0.04% 0 OSPF Hello
4 3181058 399528 7962 0.00% 0.12% 0.07% 0 Check heaps
5 267 262 1019 0.00% 0.00% 0.00% 0 Pool Manager
6 0 2 0 0.00% 0.00% 0.00% 0 Timers
7 493939 1673026 295 0.00% 0.00% 0.00% 0 ARP Input
8 0 1 0 0.00% 0.00% 0.00% 0
AAA_SERVER_DEADT
9 0 2 0 0.00% 0.00% 0.00% 0 AAA
high-capacit
10 0 1 0 0.00% 0.00% 0.00% 0 Policy
Manager
11 17 16 1062 0.00% 0.00% 0.00% 0 Entity MIB
API
12 0 1 0 0.00% 0.00% 0.00% 0 IFS Agent
Manage
13 146 56278 2 0.00% 0.00% 0.00% 0 IPC Dynamic
Cach
14 0 1 0 0.00% 0.00% 0.00% 0 IPC Zone
Manager
15 207 3354448 0 0.00% 0.00% 0.00% 0 IPC Periodic
Tim
16 156 3354448 0 0.00% 0.00% 0.00% 0 IPC Deferred
Por
17 21455 645115 33 0.00% 0.00% 0.00% 0 IPC Seat
Manager
18 10839 842765 12 0.00% 0.00% 0.00% 0 HC Counter
Timer
19 0 1 0 0.00% 0.00% 0.00% 0 HRPC
asic-stats
20 246 3354449 0 0.00% 0.00% 0.00% 0 Dynamic ARP
Insp
21 0 2 0 0.00% 0.00% 0.00% 0 XML Proxy
Client
22 0 1 0 0.00% 0.00% 0.00% 0 Critical
Bkgnd
23 3737 2160039 1 0.00% 0.00% 0.00% 0 Net
Background
24 76 1634 46 0.00% 0.00% 0.00% 0 Logger
25 498 3354398 0 0.00% 0.00% 0.00% 0 TTY
Background
26 1348 3354445 0 0.00% 0.00% 0.00% 0 Per-Second
Jobs
27 525390 56548 9291 0.00% 0.00% 0.00% 0 Per-minute
Jobs
28 0 11 0 0.00% 0.00% 0.00% 0 AggMgr
Process
29 1718 56254 30 0.00% 0.00% 0.00% 0 Net Input
30 16334 674309 24 0.00% 0.00% 0.00% 0 Compute load
avg
31 4891 15248 320 0.00% 0.00% 0.00% 0 Collection
proce
32 0 1 0 0.00% 0.00% 0.00% 0 HRPC
test_hhprpc
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
33 0 1 0 0.00% 0.00% 0.00% 0 HRPC lpip
reques
34 0 2 0 0.00% 0.00% 0.00% 0 HLPIP Sync
Proce
35 0 1 0 0.00% 0.00% 0.00% 0 HRPC Multi-FS
Sy
36 0 27 0 0.00% 0.00% 0.00% 0 HULC multifs
pro
37 0 1 0 0.00% 0.00% 0.00% 0 HRPC test
reques
38 0 2 0 0.00% 0.00% 0.00% 0 MIRAGE RBCP
Moni
39 0 1 0 0.00% 0.00% 0.00% 0 HRPC hsm
request
40 9114 105091 86 0.00% 0.00% 0.00% 0 Stack Mgr
41 11585 210434 55 0.00% 0.00% 0.00% 0 Stack Mgr
Notifi
42 61619 181158634 0 0.00% 0.00% 0.00% 0 Fifo Error
Detec
43 70037 7271 9632 0.00% 0.00% 0.00% 0 Adjust
Regions
44 1246 3354407 0 0.00% 0.00% 0.00% 0 hrpc ->
response
45 19946 3107951 6 0.00% 0.00% 0.00% 0 hrpc ->
request
46 593649 4977303 119 0.15% 0.01% 0.00% 0 hrpc <-
response
47 0 7 0 0.00% 0.00% 0.00% 0 HULC Device
Mana
48 8 23 347 0.00% 0.00% 0.00% 0 HRPC hdm non
blo
49 0 7 0 0.00% 0.00% 0.00% 0 HRPC hdm
blockin
50 50 674291 0 0.00% 0.00% 0.00% 0 HIPC bkgrd
proce
51 17 190 89 0.00% 0.00% 0.00% 0 Hulc
Port-Securi
52 17 421 40 0.00% 0.00% 0.00% 0 HRPC hpsecure
re
53 207 2999 69 0.00% 0.00% 0.00% 0 HRPC hlfm
reques
54 16222 100564584 0 0.00% 0.00% 0.00% 0 HLFM address
lea
55 415 3354402 0 0.00% 0.00% 0.00% 0 HLFM aging
proce
56 9221 100568386 0 0.00% 0.00% 0.00% 0 HLFM address
ret
57 0 1 0 0.00% 0.00% 0.00% 0 HRPC hrcmd
reque
58 0 5 0 0.00% 0.00% 0.00% 0 HRPC hulc
misc r
59 0 5 0 0.00% 0.00% 0.00% 0 HRPC x_setup
req
60 0 5 0 0.00% 0.00% 0.00% 0 HRPC system
mtu
61 265 1124417 0 0.00% 0.00% 0.00% 0 HVLAN main
bkgrd
62 0 2 0 0.00% 0.00% 0.00% 0 HVLAN Mapped
Vla
63 0 2 0 0.00% 0.00% 0.00% 0 Vlan shutdown
Pr
64 8 40 200 0.00% 0.00% 0.00% 0 HRPC vlan
reques
65 0 2 0 0.00% 0.00% 0.00% 0 HULC VLAN REF
Ba
66 9 5 1800 0.00% 0.00% 0.00% 0 HRPC hfbm
reques
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
67 17 18761 0 0.00% 0.00% 0.00% 0 HCMP sync
proces
68 0 1 0 0.00% 0.00% 0.00% 0 HPM Msg Retry
Pr
69 763942 4084 187057 0.00% 0.00% 0.00% 0 crypto sw pk
pro
70 15101 29471384 0 0.00% 0.00% 0.00% 0 hpm main
process
71 41 222 184 0.00% 0.00% 0.00% 0 HPM Stack
Sync P
72 261 1852 140 0.00% 0.00% 0.00% 0 HRPC pm
request
73 407919 5362561 76 0.00% 0.00% 0.00% 0 hpm counter
proc
74 1316182 7464995 176 0.00% 0.01% 0.00% 0 HRPC
pm-counters
75 0 1 0 0.00% 0.00% 0.00% 0 hpm vp events
ca
76 0 1 0 0.00% 0.00% 0.00% 0 HRPC hcmp
reques
77 0 1612 0 0.00% 0.00% 0.00% 0 HCEF ADJ
Refresh
78 0 6 0 0.00% 0.00% 0.00% 0 HL2MM
79 0 15 0 0.00% 0.00% 0.00% 0 HRPC hl2mm
reque
80 0 1 0 0.00% 0.00% 0.00% 0 HRPC hl3mm
reque
81 0 1 0 0.00% 0.00% 0.00% 0
hl3md_rpfq_thrl_
82 2573 2355700 1 0.00% 0.00% 0.00% 0 hl3mm
83 0 1 0 0.00% 0.00% 0.00% 0 HACL Queue
Proce
84 0 1 0 0.00% 0.00% 0.00% 0 HRPC acl
request
85 0 42 0 0.00% 0.00% 0.00% 0 HACL Acl
Manager
86 0 1 0 0.00% 0.00% 0.00% 0 HRPC backup
inte
88 0 1 0 0.00% 0.00% 0.00% 0 HRPC cdp
request
89 0 1 0 0.00% 0.00% 0.00% 0 HRPC dot1x
reque
90 0 1 0 0.00% 0.00% 0.00% 0 HULC DOT1X
Proce
91 0 1 0 0.00% 0.00% 0.00% 0 HRPC sdm
request
92 6010 16770658 0 0.00% 0.00% 0.00% 0 Hulc Storm
Contr
93 0 2 0 0.00% 0.00% 0.00% 0 HSTP Sync
Proces
94 0 85 0 0.00% 0.00% 0.00% 0 HRPC stp_cli
req
95 0 1 0 0.00% 0.00% 0.00% 0 HRPC
stp_state_s
96 0 2 0 0.00% 0.00% 0.00% 0 S/W Bridge
Proce
97 0 7 0 0.00% 0.00% 0.00% 0 HRPC hudld
reque
98 0 1 0 0.00% 0.00% 0.00% 0 HRPC vqpc
reques
99 0 1 0 0.00% 0.00% 0.00% 0 HRPC
iec_load_ba
100 8 406 19 0.00% 0.00% 0.00% 0 HRPC l2pt qnq
rp
101 92 1684265 0 0.00% 0.00% 0.00% 0 hl3mm_rp
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
102 376302 3889509 96 0.00% 0.00% 0.00% 0 HL3U bkgrd
proce
103 159113 2504860 63 0.00% 0.00% 0.00% 0 HRPC hl3u
reques
104 136 771172 0 0.00% 0.00% 0.00% 0 HL3U PBR
bkgrd p
105 17 28135 0 0.00% 0.00% 0.00% 0 HL3U PBR n-h
res
106 49 161 304 0.00% 0.00% 0.00% 0 HRPC hled
reques
107 90555 80466629 1 0.00% 0.00% 0.00% 0 Hulc LED
Process
108 0 1 0 0.00% 0.00% 0.00% 0 HRPC dtp
request
109 0 1 0 0.00% 0.00% 0.00% 0 HRPC
show_forwar
110 0 5 0 0.00% 0.00% 0.00% 0 HRPC snmp
reques
111 5391 674336 7 0.00% 0.00% 0.00% 0 HQM Stack
Proces
112 3341739 3863644 864 0.00% 0.06% 0.09% 0 HRPC qos
request
113 0 1 0 0.00% 0.00% 0.00% 0 HRPC span
reques
114 0 1 0 0.00% 0.00% 0.00% 0 HRPC system
post
115 0 17 0 0.00% 0.00% 0.00% 0 Hulc Reload
Mana
116 0 21 0 0.00% 0.00% 0.00% 0 HRPC
hrcli-event
117 6580186 47460271 138 0.00% 0.12% 0.13% 0 OSPF Router 1
118 0 2 0 0.00% 0.00% 0.00% 0 image mgr

119 0 2 0 0.00% 0.00% 0.00% 0 EAPoUDP
Process
120 0 3 0 0.00% 0.00% 0.00% 0 CEF switching
ba
121 1476 3354401 0 0.00% 0.00% 0.00% 0 PI MATM Aging
Pr
122 0 1 0 0.00% 0.00% 0.00% 0 Switch Backup
In
123 19239 721762 26 0.00% 0.00% 0.00% 0 DTP Protocol
124 0 1 0 0.00% 0.00% 0.00% 0 HRPC dai
request
125 0 1 0 0.00% 0.00% 0.00% 0 HULC DAI
Process
126 0 1 0 0.00% 0.00% 0.00% 0 HRPC dhcp
snoopi
127 9 6 1500 0.00% 0.00% 0.00% 0 HULC DHCP
Snoopi
128 0 1 0 0.00% 0.00% 0.00% 0 HRPC ip
source g
129 0 1 0 0.00% 0.00% 0.00% 0 HULC IP
Source g
130 1124 3376824 0 0.00% 0.00% 0.00% 0 UDLD
131 25 112561 0 0.00% 0.00% 0.00% 0 Port-Security
132 118 244 483 0.47% 0.05% 0.01% 1 Virtual Exec
133 0 1 0 0.00% 0.00% 0.00% 0 Switch IP
Admiss
134 0 2 0 0.00% 0.00% 0.00% 0 L2 ARP
Process
135 8 337506 0 0.00% 0.00% 0.00% 0 Ethchnl
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
136 41 162 253 0.00% 0.00% 0.00% 0 VMATM
Callback
137 0 2 0 0.00% 0.00% 0.00% 0 AAA Server
138 0 1 0 0.00% 0.00% 0.00% 0 AAA ACCT Proc
139 0 1 0 0.00% 0.00% 0.00% 0 ACCT Periodic
Pr
140 0 9 0 0.00% 0.00% 0.00% 0 CEF IPC
Backgrou
141 8 72 111 0.00% 0.00% 0.00% 0 SpanTree
Helper
142 0 1 0 0.00% 0.00% 0.00% 0 HRPC ilp
request
143 0 2 0 0.00% 0.00% 0.00% 0 AAA
Dictionary R
144 17 28143 0 0.00% 0.00% 0.00% 0 DHCP Snooping
145 176888311 748386486 236 15.65% 15.85% 14.66% 0 IP Input
146 0 1 0 0.00% 0.00% 0.00% 0 ICMP event
handl
147 3978 33535377 0 0.00% 0.00% 0.00% 0 MDFS MFIB
Proces
148 3689813 10147379 363 0.15% 0.10% 0.09% 0 Spanning Tree
149 8 56283 0 0.00% 0.00% 0.00% 0 Spanning Tree
St
150 0 1 0 0.00% 0.00% 0.00% 0 CEF RF HULC
Conv
151 3529 425252 8 0.00% 0.00% 0.00% 0 XDR mcast
152 311 56770 5 0.00% 0.00% 0.00% 0 CEF
background p
153 0 1 0 0.00% 0.00% 0.00% 0 IP IRDP
154 407930 6640655 61 0.00% 0.00% 0.00% 0 IPC LC
Message H
155 0 3 0 0.00% 0.00% 0.00% 0 XDR RP Ping
Back
156 18 28141 0 0.00% 0.00% 0.00% 0 XDR RP
backgroun
157 0 1 0 0.00% 0.00% 0.00% 0 XDR RP Test
Back
158 0 3 0 0.00% 0.00% 0.00% 0 MDFS LC
Process
159 0 3 0 0.00% 0.00% 0.00% 0 Routemap RP
IPC
160 0 42212 0 0.00% 0.00% 0.00% 0 Cluster L2
161 58 337504 0 0.00% 0.00% 0.00% 0 Cluster RARP
162 60 421958 0 0.00% 0.00% 0.00% 0 Cluster Base
163 792 35744 22 0.00% 0.00% 0.00% 0 TCP Timer
164 50 194 257 0.00% 0.00% 0.00% 0 TCP Protocols
165 0 1 0 0.00% 0.00% 0.00% 0 Socket Timers
166 0 11268 0 0.00% 0.00% 0.00% 0 HTTP CORE
167 0 1 0 0.00% 0.00% 0.00% 0 RARP Input
168 75 112 669 0.00% 0.00% 0.00% 0 L2MM
169 0 1 0 0.00% 0.00% 0.00% 0 MRD
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
170 0 1 0 0.00% 0.00% 0.00% 0 IGMPSN
171 0 1 0 0.00% 0.00% 0.00% 0 IGMPQR
172 0 2 0 0.00% 0.00% 0.00% 0 L2TRACE
SERVER
173 26 4 6500 0.00% 0.00% 0.00% 0 Crypto CA
174 548812 455899 1203 0.00% 0.00% 0.00% 0 IP RIB Update
175 0 1 0 0.00% 0.00% 0.00% 0 Auth-proxy
AAA B
176 0 11257 0 0.00% 0.00% 0.00% 0 IP Admin SM
Proc
177 76 225013 0 0.00% 0.00% 0.00% 0 MDFS LC
Download
178 0 1 0 0.00% 0.00% 0.00% 0 Crypto
PKI-CRL
179 6127 34124937 0 0.00% 0.00% 0.00% 0 MDFS RP
process
181 0 2 0 0.00% 0.00% 0.00% 0 AAA Cached
Serve
182 0 2 0 0.00% 0.00% 0.00% 0 LOCAL AAA
183 0 2 0 0.00% 0.00% 0.00% 0 TPLUS
184 0 1 0 0.00% 0.00% 0.00% 0 Crypto SSL
185 0 17 0 0.00% 0.00% 0.00% 0 VTP Trap
Process
186 0 2 0 0.00% 0.00% 0.00% 0 VTPMIB EDIT
BUFF
187 0 2 0 0.00% 0.00% 0.00% 0 DHCP Security
He
188 0 1 0 0.00% 0.00% 0.00% 0 HCD Process
189 0 1 0 0.00% 0.00% 0.00% 0 HRPC cable
diagn
190 5438 29925905 0 0.00% 0.00% 0.00% 0 PM Callback
191 8 7 1142 0.00% 0.00% 0.00% 0 VLAN Manager
193 0 3 0 0.00% 0.00% 0.00% 0 RADIUS TEST
CMD
194 0 2 0 0.00% 0.00% 0.00% 0 AAA SEND STOP
EV
195 0 1 0 0.00% 0.00% 0.00% 0 Syslog Traps
196 0 2 0 0.00% 0.00% 0.00% 0 STP FAST
TRANSIT
197 0 2 0 0.00% 0.00% 0.00% 0 CSRT RAPID
TRANS
199 318 1877 169 0.00% 0.00% 0.00% 0 SSH Event
handle
200 6568 5245318 1 0.00% 0.00% 0.00% 0 CEF: IPv4
proces
201 59 83 710 0.00% 0.00% 0.00% 0 ADJ
background
202 6690 56663 118 0.00% 0.00% 0.00% 0 IP Background
203 0 6 0 0.00% 0.00% 0.00% 0 SNMP Timers
204 144871 816084 177 0.00% 0.00% 0.00% 0 IP SNMP
205 44876 409735 109 0.00% 0.00% 0.00% 0 PDU
DISPATCHER
206 188902 409735 461 0.00% 0.00% 0.00% 0 SNMP ENGINE
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
207 0 1 0 0.00% 0.00% 0.00% 0 SNMP
ConfCopyPro
208 8 8 1000 0.00% 0.00% 0.00% 0 SNMP Traps
209 419 3403177 0 0.00% 0.00% 0.00% 0 NTP
210 203 278 730 0.00% 0.00% 0.00% 0 hulc cfg mgr
mas
211 199704 973 205245 0.00% 0.00% 0.00% 0 hulc running
con



----- Original Message -----
From: ***@census.gov
To: VCI Help Desk
Sent: Sunday, December 25, 2005 11:08 AM
Subject: Re: [c-nsp] 3750 stack with high CPU from IP Input process.
Post by VCI Help Desk
The CPU went from 5% to 50% just because I moved an ethernet
cable. When I moved it back the CPU didn't go down.
Which sounds like it wasn't moving the cable that caused the cpu busy.

Can you post the results of a "show proc cpu"?

Regards,
Lee

_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
l***@census.gov
2005-12-25 20:08:05 UTC
Permalink
The next time it happens, post the results of a "sh proc cpu" & maybe
someone can figure out what's causing the cpu busy. Or at least say it's
caused by snmp, spanning tree, or whatever.

Other good stuff to look at is the results from
sh ip traffic
need to do it twice - the deltas are what's interesting
sh int
5 minute in.out rate is interesting, deltas for errors etc. are
interesting

Lee
Post by Bill
Here is the "show proc cpu" The CPU usage is low right now. The other
day something was going on and the CPU went 10 100% for no reason I could
find. I can't find any reason why the CPU would have gone up when I moved
the ethernet cable.
Bill
esw-1#show proc cpu
CPU utilization for five seconds: 27%/6%; one minute: 27%; five minutes: 25%
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
<.. snip ..>

_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
Bill
2005-12-27 13:50:55 UTC
Permalink
I ended up rebooting the switch stack last night and guess what? The CPU
went back to 5%. I've sent this to Cisco asking them if there is a bug in
the IOS.

Bill


----- Original Message -----
From: ***@census.gov
To: Bill
Cc: cisco-***@puck.nether.net
Sent: Sunday, December 25, 2005 2:08 PM
Subject: Re: [c-nsp] 3750 stack with high CPU from IP Input process.


The next time it happens, post the results of a "sh proc cpu" & maybe
someone can figure out what's causing the cpu busy. Or at least say it's
caused by snmp, spanning tree, or whatever.

Other good stuff to look at is the results from
sh ip traffic
need to do it twice - the deltas are what's interesting
sh int
5 minute in.out rate is interesting, deltas for errors etc. are
interesting

Lee
Post by Bill
Here is the "show proc cpu" The CPU usage is low right now. The other
day something was going on and the CPU went 10 100% for no reason I could
find. I can't find any reason why the CPU would have gone up when I moved
the ethernet cable.
Bill
esw-1#show proc cpu
CPU utilization for five seconds: 27%/6%; one minute: 27%; five minutes: 25%
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
<.. snip ..>

_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
Johannes Resch
2005-12-27 16:47:36 UTC
Permalink
FWIW, I've seen a similar issue on a 3750 stack running 12.2(25)SE (4
stackmembers).
If I remember correctly, the 100% CPU load was all due to interrupt CPU,
however "sh int stat", "sh ip cef switching statistics" and friends
indicated traffic from/to the switch itself was only in the order of ~10pps.

regards,
-jr
Post by Bill
I ended up rebooting the switch stack last night and guess what? The CPU
went back to 5%. I've sent this to Cisco asking them if there is a bug in
the IOS.
Bill
----- Original Message -----
To: Bill
Sent: Sunday, December 25, 2005 2:08 PM
Subject: Re: [c-nsp] 3750 stack with high CPU from IP Input process.
The next time it happens, post the results of a "sh proc cpu" & maybe
someone can figure out what's causing the cpu busy. Or at least say it's
caused by snmp, spanning tree, or whatever.
Other good stuff to look at is the results from
sh ip traffic
need to do it twice - the deltas are what's interesting
sh int
5 minute in.out rate is interesting, deltas for errors etc. are
interesting
Lee
Post by Bill
Here is the "show proc cpu" The CPU usage is low right now. The other
day something was going on and the CPU went 10 100% for no reason I could
find. I can't find any reason why the CPU would have gone up when I moved
the ethernet cable.
Bill
esw-1#show proc cpu
25%
Post by Bill
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
<.. snip ..>
_______________________________________________
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
--
Johannes Resch

tel: +43 699 / 11 077 818
mail: ***@xor.at

PGP Key ID: AF93FA9B (pgp.mit.edu)

_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
Ed Butler - RapidSwitch
2005-12-27 21:52:11 UTC
Permalink
I can't be of much help with this, aside from saying we've seen weird CPU
issues on 3750 stacks too.

We isolated one particular problem to the 3750 storm control feature. (When
a storm wasn't occurring.)

Regards,

Ed Butler
RapidSwitch Ltd
DDI: 020 7106 0731

RapidSwitch Ltd, 5th Floor, Sovereign House, 227 Marsh Wall, London, E14 9SD

This email message is intended only for the addressee(s) and contains
information that may be confidential and/or copyright. If you are not the
intended recipient please notify the sender by reply email and immediately
delete this email. Use, disclosure or reproduction of this email by anyone
other than the intended recipient(s) is strictly prohibited. No
representation is made that this email or any attachments are free of
viruses. Virus scanning is recommended and is the responsibility of the
recipient.
-----Original Message-----
From: cisco-nsp-***@puck.nether.net
[mailto:cisco-nsp-***@puck.nether.net] On Behalf Of Johannes Resch
Sent: 27 December 2005 16:48
To: Bill
Cc: cisco-***@puck.nether.net
Subject: Re: [c-nsp] 3750 stack with high CPU from IP Input process.

FWIW, I've seen a similar issue on a 3750 stack running 12.2(25)SE (4
stackmembers).
If I remember correctly, the 100% CPU load was all due to interrupt CPU,
however "sh int stat", "sh ip cef switching statistics" and friends
indicated traffic from/to the switch itself was only in the order of ~10pps.

regards,
-jr
Post by Bill
I ended up rebooting the switch stack last night and guess what?
The CPU went back to 5%. I've sent this to Cisco asking them if there
is a bug in the IOS.
Bill
----- Original Message -----
To: Bill
Sent: Sunday, December 25, 2005 2:08 PM
Subject: Re: [c-nsp] 3750 stack with high CPU from IP Input process.
The next time it happens, post the results of a "sh proc cpu" & maybe
someone can figure out what's causing the cpu busy. Or at least say
it's caused by snmp, spanning tree, or whatever.
Other good stuff to look at is the results from sh ip traffic
need to do it twice - the deltas are what's interesting sh int
5 minute in.out rate is interesting, deltas for errors etc. are
interesting
Lee
Post by Bill
Here is the "show proc cpu" The CPU usage is low right now. The
other day something was going on and the CPU went 10 100% for no
reason I could find. I can't find any reason why the CPU would have
gone up when I moved the ethernet cable.
Bill
esw-1#show proc cpu
25%
Post by Bill
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
<.. snip ..>
_______________________________________________
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
--
Johannes Resch

tel: +43 699 / 11 077 818
mail: ***@xor.at

PGP Key ID: AF93FA9B (pgp.mit.edu)

_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

_______________________________________________
cisco-nsp mailing list cisco-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
Continue reading on narkive:
Loading...