Cisco 3750E Stack InterVlan Routing Issue

[Ahmet Erkol]

Hi,

We have 2 Cisco 3750E switches. One of them is WS-C3750E-24TD-S and the other one is WS-C3750E-48TD-S. They have  C3750E-UNIVERSALK9-M 15.0(2) SE1 IOS image loaded. They stacked each other and have about 15 SVIs on it. Normally everthing is ok. But when we power off the master switch intervlan routing is not working. We cannot ping Vlans on the same switch. When we shutdown and no shutdown the Vlan interfaces intervlan routing is back again and everything looks normal. We use the persistent mac for stack feature and the mac addresses of the Vlans do not change also. What is the problem we couldn't find a solution.

[Mark Malone]

That sounds like a bug have you tried changing the IOS see if the issue follows ,12.2.55 se10 is the safe harbour images for those devices

[Ahmet Erkol]

We use Right-to-Use license for 48 port 3750 and it is supported 15.0(2) IOS

[Joseph W. Doherty]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages wha2tsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

Agree with Mark.

You might search through Leo Laohoo's posts for a 15.x IOS recommendation (although suspect he too would recommend (55)SE10 if you don't really need 15.x features).

[Ganesh Hariharan]

Hello, 

Bit strange issue, Have you tried changing the role of stack master and see is intervlan routing  is working fine or not in normal scenario.

And as suggested by Mark it may be bug which may required to upgrade on stable version.

-GI

[Ahmet Erkol]

Hi,

At first 24 port switch had an ipservices license and 48 port switch has an ipbase license. Two redundant uplink connections was terminating on the 24 port switch because it is the only stack switch that supports ipservices. After IOS 15.0(2)SE1 Right-to-Use license feature was introduced. For redundancy we took one connection to the 48 port switch and execute a command to use ipservices Right-to-Use license. Then reboot the 48 port switch. Everything was great but when we reload the 24 port master switch we lost connection to the switch and from console we couldn't ping betweeen SVIs. Should I make the stack process from begining with a default configuration?

[Joseph W. Doherty]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages wha2tsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

Ah, well I was wondering about licensing, but I saw both your switches had the -S suffix, so you field upgraded just the one?

When you lose the only (or all) switch(es) with the higher licensed features, the stack loses features.  IPBase should support routing between VLANs on the same stack (and it appears it does, as it does after you shut/no-shut), but I'm guessing when the stack feature set is reduced, the stack becomes "confused".  To me, still sounds like a possible bug, but this is something Cisco might consider a "feature".

A solution might be to run both stack members with the same enabled/licensed features.  (I.e. upgrade or downgrade, so your two stacked switches are as "match pair".)

I've run stacks where not all members have the advanced feature set, but I insure that at least two members always do when that feature set is needed.

[Ahmet Erkol]

Hi,

Actually at the moment two switches running with ipservices license. But at past one of them was working with ipbase. The we upgrade the ipbase one to ipservices right-to-use license. I asked that should I reconfigure the switches from default. Because we have redundant switches and we made a demo platform same with production. We tried the same thing we power off the master switch and after 2 ping lost connection was back.

[Joseph W. Doherty]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages wha2tsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

Losing a few pings, to the switch, when the master fails over, is normal.  What the remaining stack should do is keep forwarding traffic, and if configured for NSF, also not drop dynamic routing protocol to NSF aware neighbors during the master fail over.