×

Warning message

  • Cisco Support Forums is in Read Only mode while the site is being migrated.
  • Cisco Support Forums is in Read Only mode while the site is being migrated.

2924M XL with %RTD-1-ADDR_FLAP:FastEthernet 2/1 relearning x addrs per min

Unanswered Question
Sep 9th, 2002
User Badges:

Have recently upgraded the OS on my 2924XL switches to 12.0(5)WC3b from 12.0(5)XP. Following the upgrade, started receiving the RTD...error message. My topology has two fast ethernet interfaces on the switch connected to two fast ethernet interfaces on a 7500 series router which is running version 12.1(13). The switch interfaces are configured for trunking with ISL encapsulation. They are also fully duplexed. We have been running version 5XP for two years without any problems. What is our problem now and what can we do to correct it. Thanks for your assistance.


Ken Lee

Network Engineer

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
efrahim Mon, 09/09/2002 - 10:01
User Badges:
  • Cisco Employee,

These messages means you the mac addresss are flapping between interface-


Please use the command " debug ethernet-con add" to see what address are flapping and between what ports....



Do you have any bridging on the routers- How about hsrp - If you are running hsrp and no bridging on the router, please make sure you have the portfast on those ports. It could be reboot that cause this problem , not the upgrade.


here is the good link


http://www.cisco.com/warp/customer/473/62.shtml#casestudy5





eecunningham Mon, 09/09/2002 - 10:45
User Badges:

We are not running HSRP on the routers. I have two identical switches, configured the same and connected to two different routers and both of these switches started getting this error message right after being upgraded to the new OS. I am apprehensive about running debug on these switches because they are in a production environment and support about 2000 users each. Any other ideas? Thanks...

efrahim Tue, 09/10/2002 - 07:05
User Badges:
  • Cisco Employee,

These messages are just mentioning that the mac addresses are flapping. We need to find the reason why they are-


Use the debug ethernet-contro address to see what mac addresses are flapping and between what ports and how they are connected why that mac addresse are flapping.between two ports. kind of indicating the loop. It seems like the root of the problem is something else. I think open up the TAC case for further investigation

eecunningham Tue, 09/10/2002 - 11:24
User Badges:

I ran the debug. As expected ports fa1/1 and 1/2 are flapping. These are the two ports which are connected to the router on ports 1/0/0 and 1/0/1. The two router ports are configured to trunk over multiple vlans in the form of ISLs and are configured as port channel group 1. I have just configured my switch ports, fa1/1 and fa1/2, to port group 1 and it seems that has resolved my flapping problem. However, a new problem has emerged. If I disconnect either port fa1/1 or fa1/2 from the router, the switch loses all connectivity to the network. This didn't happen before I configured them to port group 1. In addition, if I administratively shut down either port, they both shutdown. Is there some way to fix this problem because we want to be able to run both interfaces and can't afford to have one of them going down causing both interfaces to go down. Thanks again. Ken

Prashanth Krishnappa Tue, 09/10/2002 - 16:26
User Badges:
  • Cisco Employee,

If you configure port-group(etherchannel) on the XL, you'll need to do the same on the router(port channel) as well. What router is this? Etherchannel on routers is both hardware and software dependent

eecunningham Wed, 09/11/2002 - 06:25
User Badges:

We are using 7500 series routers with the 12.1(13) operating system. The router's port channels have always been configured, only the switches weren't configured for port groups. The remaining problem is that if we lose one of the switch interfaces connected to the router, whether physically or administratively, we effectively lose the entire switch. This is unacceptable. Is there a work around for this?

Actions

This Discussion