High Utilization on one port of an Etherchannel

[alfonso.cornejo]

Hi to all,

I have an etherchannel using 4 gigabit interfaces and i'm having high utilization just in one of it. It's the interface G6/32 on the SW-CENTRAL1 and SW-CENTRAL2 the ones that are having this situation (one in output and the other in input), the etherchannel is configured as desirable between two WS-C4507R with the IOS version 12.2(20)EWA4.

The configurations for the etherchannel are exactly the same for the two devices:

interface Port-channel2

description PortChannel-Central

switchport

switchport trunk encapsulation dot1q

switchport mode trunk

interface GigabitEthernet6/29

description EtherChannel-2

switchport trunk encapsulation dot1q

switchport mode trunk

channel-group 2 mode desirable

!

interface GigabitEthernet6/30

description EtherChannel-2

switchport trunk encapsulation dot1q

switchport mode trunk

channel-group 2 mode desirable

!

interface GigabitEthernet6/31

description EtherChannel-2

switchport trunk encapsulation dot1q

switchport mode trunk

channel-group 2 mode desirable

!

interface GigabitEthernet6/32

description EtherChannel-2

switchport trunk encapsulation dot1q

switchport mode trunk

channel-group 2 mode desirable

SW-CENTRAL1#sh int port-channel 2

Port-channel2 is up, line protocol is up (connected)

Hardware is EtherChannel, address is 000e.84cd.306e (bia 000e.84cd.306e)

Description: PortChannel-Central

MTU 1500 bytes, BW 4000000 Kbit, DLY 10 usec,

reliability 255/255, txload 7/255, rxload 6/255

Encapsulation ARPA, loopback not set

Keepalive set (10 sec)

Full-duplex, 1000Mb/s, media type is N/A

input flow-control is off, output flow-control is unsupported

Members in this channel: Gi6/29 Gi6/30 Gi6/31 Gi6/32

ARP type: ARPA, ARP Timeout 04:00:00

Last input 00:00:00, output never, output hang never

5 minute input rate 99376000 bits/sec, 20421 packets/sec

5 minute output rate 112217000 bits/sec, 20172 packets/sec

SW-CENTRAL2#sh int port-channel 2

Port-channel2 is up, line protocol is up (connected)

Hardware is EtherChannel, address is 000e.84cd.2cde (bia 000e.84cd.2cde)

Description: PortChannel-Central

MTU 1500 bytes, BW 4000000 Kbit, DLY 10 usec,

reliability 255/255, txload 6/255, rxload 6/255

Full-duplex, 1000Mb/s, media type is N/A

input flow-control is off, output flow-control is unsupported

Members in this channel: Gi6/29 Gi6/30 Gi6/31 Gi6/32

ARP type: ARPA, ARP Timeout 04:00:00

Last input 00:00:00, output never, output hang never

Last clearing of "show interface" counters 11w0d

Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 0

Queueing strategy: fifo

Output queue: 0/40 (size/max)

5 minute input rate 108971000 bits/sec, 19845 packets/sec

5 minute output rate 102179000 bits/sec, 20525 packets/sec

Do you know what could be causing this??

Thanks in advance for your help.

[Jon Marshall]

Alfonso

Any one individual conversation will always use the same link. How the individual conversations are defined is down to the load-balancing algorithm in use on the switch.

So it may just be that 2 machines on either side of the link are doing all the talking at the moment or it may be that your load-balancing algorithm could be changed to get a better distribution across the links. Have a look at this doc for more details on etherchannel load-balancing -

http://www.cisco.com/en/US/tech/tk389/tk213/technologies_tech_note09186a0080094714.shtml

Jon

[andrew.butterworth]

What EtherChannel load-balancing algorithm are you using? Depending on the traffic characteristics it is possible this behaviour is normal.

Have a read of the EtherChannel chapter here:

http://www.cisco.com/en/US/docs/switches/lan/catalyst4500/12.2/50sg/configuration/guide/channel.html

HTH

Andy

[Roberto Salazar]

From the output you've showed, how are you verifying that it's only using one link?

[alfonso.cornejo]

Hi,

My LMS is sending me this message:

EVENT ID = 0008HOB

ALERT ID = 00001L8

TIME = Thu 08-Jan-2009 04:54:10 CST

STATUS = Active

SEVERITY = Critical

MANAGED OBJECT = 128.127.254.1

MANAGED OBJECT TYPE = Switches and Hubs

EVENT DESCRIPTION = HighUtilization::Component=PORT-128.127.254.1/95 [Gi6/32] [EtherChannel-2];OutputPacketRate=67309.25 PPS;Type=ETHERNETCSMACD;CurrentUtilization=75.16374 %;InputPacketRate=16970.86 PPS;TrafficRate=9.395468E7 BYPS;UtilizationThreshold=75;DuplexMode=FULLDUPLEX;MaxSpeed=1000000000;

CUSTOMER IDENTIFICATION = Switches

And i also graph the traffic on all the etherchannel interfaces and is just the g6/32 the one that is having this high utilization

[glen.grant]

Because of the channel distribution method that is configured looks like most is riding that single link , you need to figure out why. Are all the links up in the channel group ? Post a "show ether summary " command.

[alfonso.cornejo]

Hi Glen,

Here is the output from the 2 sw:

SW-CENTRAL1#sh etherchannel summary

Number of channel-groups in use: 2

Number of aggregators: 2

Group Port-channel Protocol Ports

------+-------------+-----------+-----------------------------------------------

2 Po2(SU) PAgP Gi6/29(P) Gi6/30(P) Gi6/31(P)

Gi6/32(P)

4 Po4(SU) - Gi7/3(P) Gi7/4(P)

SW-CENTRAL2#sh etherchannel summary

Number of channel-groups in use: 2

Number of aggregators: 2

Group Port-channel Protocol Ports

------+-------------+-----------+-----------------------------------------------

2 Po2(SU) PAgP Gi6/29(P) Gi6/30(P) Gi6/31(P)

Gi6/32(P)

3 Po3(SU) - Gi7/1(P) Gi7/2(P)

I don't know why most of the traffic is going only throught this single port, one thing important is that this happens on a random period of time, i mean the g6/32 is not all the time with high utilization.

Thanks

[Giuseppe Larosa]

Hello Alfonso,

use

sh etherchannel load-balance

to see what algorythm is in use

example:

sh etherchannel load-balance

EtherChannel Load-Balancing Configuration:

src-dst-ip

mpls label-ip

EtherChannel Load-Balancing Addresses Used Per-Protocol:

Non-IP: Source XOR Destination MAC address

IPv4: Source XOR Destination IP address

IPv6: Source XOR Destination IP address

MPLS: Label or IP

Hope to help

Giuseppe

[Roberto Salazar]

I agree.

Use the option that provides the balance criteria with the greatest variety in your configuration. For example, if the traffic on an EtherChannel is going only to a single MAC address and you use the destination MAC address as the basis of EtherChannel load balancing, the EtherChannel always chooses the same link in the EtherChannel; using source addresses or IP addresses might result in better load balancing.

For more details on configuring channel laod-balancing:

http://www.cisco.com/en/US/docs/switches/lan/catalyst6500/ios/12.2SXF/native/configuration/guide/channel.html#wp1020414

[alfonso.cornejo]

Hi,

I'll try that.

Thanks to all for your help.