3550 flapping of ports

[JMCNEL]

We have several 3550 switches still in our our network. I know, its old but please continue to read.  All the 3550 switches are on the same IOS and feature set. Recently users on 2 of the 3550 switches started to experience bandwitch issues - specifically upload speed are so slow to the point where the users cannot work. Nowhere else on the  network except for the 2 labs that have the 3550 switches are seeing the problem. The other 3550's dont have the same issue. The 2 switches in question had on multiple occasions flapping of all ports connected when the users were trying to access large files. when the 2 switches were powercyled they were fine just until the next time. there were no errors in the logs or show tech that would explain this. Why would these two switches go into the flapping state? do you think it would be worth trying to upgrade these 2 switches to the below IOS? Is there perhaps a bug on the IOS that we have? If so why would it be just the 2 switches and not all the 3550 switches that is having the issue.

The lastest IOS that is available for download was published in early 2009. (c3550-ipservices-mz.122-44.SE6.bin)
IP SERVICES W/O CRYPTO 

3550-ipservices-mz.122-44.SE6.bin
Release Date: 17/MAR/2009
Size: 7936.82 KB (8127303 bytes)
Minimum Memory: DRAM:64MB   Flash:16MB 

information on our 3550's
Cisco IOS Software, C3550 Software (C3550-IPSERVICESK9-M), Version 12.2(35)SE5, RELEASE SOFTWARE (fc1)
System image file is "flash:c3550-ipservicesk9-mz.122-35.SE5/c3550-ipservicesk9-mz.122-35.SE5.bin"
WS-C3550-48-SMI

Any help would be great

[sleepyshark]

Are you using VLAN's?

It could be an iOS bug, however bugs generally don't just manifest themselves out of thin air.... I would be inclined to look at changes made over the past which may have changed how the switches are configured.  Also, i'd like to take a look at the sh int output of the trunk interfaces and port interfaces along with the sh vtp s (if you're using VLANs) to see if we have any errors coming into play.

Essentially, it's VERY important for you to start a log of when you reboot the switches, when the problems start happening, and the configs and/or sh int outputs. 

Just my two cents....

Sean Brown

http://www.sleepyshark.com

[JMCNEL]

we havent change any config in the past 5 months. we use vlans.

VTP Version                     : 2

Configuration Revision          : 0

Maximum VLANs supported locally : 1005

Number of existing VLANs        : 8

VTP Operating Mode              : Transparent

VTP Domain Name                 : WPS

VTP Pruning Mode                : Disabled

VTP V2 Mode                     : Disabled

VTP Traps Generation            : Disabled

MD5 digest                      : 0x6A 0x48 0xAB 0x84 0x66 0x40 0xED 0xF7

Configuration last modified by 0.0.0.0 at 0-0-00 00:00:00

I will attach file with sh int output

[glen.grant]

  To me it almost sounds like you are getting some kind of spanning tree loop between these 2 switches.  I would check to see if spanning tree looks ok and the roots are where they are supposed to be.  Look  at cdp when this happens and see if you see  ports on the switch where the switch can see itself.  Do a show spanning tree blocked ports command. Look at the cpu when they start flapping and see if its really high and what process is driving it .   Look at port utilization at time of trouble using show controllers utilization command.

[Leo Laohoo]

The 2 switches in question had on multiple occasions flapping of all ports connected when the users were trying to access large files. when the 2 switches were powercyled they were fine just until the next time.

Can you post the output to the following commands:

1.  sh logs;

2.  sh controller e