Hello everyone,
I have a very strange problem on a stack of 8 Catalyst 3750E-48.
The configuration is however pretty simple.
Two VLANS (20 and 24) with inter-vlan routing (no ACL, QoS, whatsoever).
Machines in the the Vlan 20 can see each others with no problem. Same in Vlan 24.
However machines from Vlan 20 can see the ones from Vlan 24 only intermittently!
There is nothing in the log (no interface going up or down that could explain this)
I had a look at the CPU and it seems kinda high although I'm not sure this is the cause of my problems...
CPU utilization for five seconds: 45%/26%; one minute: 40%; five minutes: 45%
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
91 17770597 17477172 1016 7.18% 5.83% 5.85% 0 HLFM address lea
111 13703094 1203510 11385 2.07% 1.75% 1.73% 0 hpm counter proc
151 5449427 18566548 293 0.95% 0.77% 0.99% 0 Hulc LED Process
162 6082598 2006746 3031 0.63% 0.76% 0.75% 0 HRPC qos request
327 3248183 1627290 1996 0.63% 0.50% 0.51% 0 SNMP ENGINE
161 3099356 126352 24529 0.47% 0.38% 0.37% 0 HQM Stack Proces
197 2148272 4136461 519 0.47% 0.55% 0.54% 0 ADJ resolve proc
337 23342 4248 5494 0.31% 0.04% 0.53% 1 Virtual Exec
325 1299928 3259491 398 0.31% 0.23% 0.18% 0 IP SNMP
46 1283570 145905 8797 0.31% 0.16% 0.16% 0 Compute load avg
235 564101 713775 790 0.15% 0.14% 0.14% 0 IPC LC Message H
193 776933 2715979 286 0.15% 0.14% 0.20% 0 IP Input
39 1095189 657808 1664 0.15% 0.12% 0.13% 0 Per-Second Jobs
112 2258147 7973142 283 0.15% 0.19% 0.24% 0 HRPC pm-counters
75 445786 178510 2497 0.15% 0.15% 0.12% 0 Adjust Regions
152 413958 466565 887 0.15% 0.09% 0.09% 0 HL3U bkgrd proce
76 3404112 630469 5399 0.15% 0.04% 0.04% 0 hrpc -> response
17 0 1 0 0.00% 0.00% 0.00% 0 IFS Agent Manage
16 888674 24596449 36 0.00% 0.00% 0.00% 0 EEM ED Syslog
15 200 17 11764 0.00% 0.00% 0.00% 0 Entity MIB API
18 303 10565 28 0.00% 0.00% 0.00% 0 IPC Dynamic Cach
14 0 1 0 0.00% 0.00% 0.00% 0 Policy Manager
It seems that there are lot of interrupts and I cannot see why
On this page, Cisco says that it might be caused by "punted packets"
http://cisco.biz/en/US/docs/switches/lan/catalyst3750/software/troubleshooting/cpu_util.html#wp1028395
The sdm is set to "desktop routing" and there is no TCAM exhaustion :
MyEVS#sh platform tcam utilization
CAM Utilization for ASIC# 0 Max Used
Masks/Values Masks/values
Unicast mac addresses: 3292/3292 447/447
IPv4 IGMP groups + multicast routes: 1120/1120 1/1
IPv4 unicast directly-connected routes: 3072/3072 150/150
IPv4 unicast indirectly-connected routes: 8144/8144 102/102
IPv4 policy based routing aces: 498/498 13/13
IPv4 qos aces: 474/474 21/21
IPv4 security aces: 972/972 32/32
The switch is on its own and as I said the L3 config is very simple (all network directly connected)
MyEVS#sh ip route
172.16.0.0/24 is subnetted, 3 subnets
C 172.16.202.0 is directly connected, FastEthernet0
C 172.16.24.0 is directly connected, Vlan24
C 172.16.20.0 is directly connected, Vlan20
MyEVS#
Maybe is it caused by a lot of multicast traffic ? (I know some application intensively used on this network is based on multicast)
But is my CPU hogged and could drop traffic between Vlan because of that ?
If not.. have you any idea where it could come from ?
Thank you very much for any help
