VTP pruning issue

[mbroberson1]

Had an issue a couple of months ago where all of a sudden users connected to a switch on a particular vlan (10) could not access network resources. After some troubleshooting I noticed vtp was pruning traffic on the switch for that particular vlan (10). The moment I disabled vtp pruning users connected to the switch in the vlan (10) could again access network resources. I have not re-enabled vtp pruning yet for the administrative domain but would like to since our environment is fairly large. The switch model this happend on is a 3750 switch stack with nine (9) (the max allowed) switch members, running "3750-ipbase-mz.122-35.SE5.bin". I am wondering if it was just a VTP hiccup?

Just curious if anyone has seen this before.

Thanks

[rahurao]

Hi ,

Pruning increases available bandwidth by restricting flooded traffic to those trunk links that the traffic must use to access the destination devices. so it may have non active members on the Vlan 10 on other switch which could have caused the vlan to get pruned on the trunk.

I have seen this issue quite some times so that may be the reason.

Let me know what changes you have done in the network when the issue happened.

For reference:

http://www.cisco.com/en/US/docs/switches/lan/catalyst3550/software/release/12.1_19_ea1/configuration/guide/swvtp.html#wp1072526

HTH

Rahul

[mbroberson1]

Hi Rahul,

Thanks for your reply. This was a very strange issue. On the 3750 switch stack each member switch has atleast 10 devices in the vlan that was pruned all in an up/up "connected" status. Also there were no changes to the network at the time this occured.

The setup is a 3750 switch stack..each member switch is a WS-C3750-48PS-S (9 member switches) with a 2GB etherchannel back to a Nexus 7010.

Thanks

[Shashank Singh]

Hi,

Sounds like a bug.Following may be a match:

http://tools.cisco.com/Support/BugToolKit/search/getBugDetails.do?method=fetchBugDetails&bugId=CSCth18118

You can check the IOS version on the switch and confirm this.

Hope this helps.

Shashank

Please rate this answer if you found the content useful

[mbroberson1]

Hi Shashank,

Very helpful post. When I searched the bug toolkit before I never came across anything identifying a possible bug. Lucky for me you found something. And yes the version we are on is with-in the known affected releases. Thanks again!

Regards,

-BR

[mbroberson1]

Hi Rahul,

I got sround to upgrading code on our 3750 switch stack and I still seem to be hitting the bug...if it is a bug I am hitting.

Details on the software bug:

http://tools.cisco.com/Support/BugToolKit/search/getBugDetails.do?method=fetchBugDetails&bugId=CSCth18118

I upgraded to 12.2(55)SE1 from 12.2(35)SE5 in hopes of fixing the bug. The bug mentions the issue is fixed 12.2(55)SE, Cisco TAC mentioned the bug is also fixed in 12.2(55)SE1.

As you can see from the below output vlans 65 and 165 are both being pruned for lack of request by neighbor as well as traffic requested of neighbor. Vlan 65 and 165 has many host in these vlans on the switch.

switch065#sh interfaces pruning

Port                Vlans pruned for lack of request by neighbor
Po1                 10,13,20,26,40,50-51,60,65,98-100,104,110,126,150,159,165,198,221-222

Port                Vlan traffic requested of neighbor
Po1                 1,6,8-9,16-17,27,30,65,102,106,116,165,502

I am probably going to re-open my original TAC case and see what they say.

Regards,

Brandon

[rahurao]

Hi Brandon,

can you just try to upgrade to the 12.2(55) version and check ?

HTH

Rahul

[mbroberson1]

Hi Rahul,

You mean the 12.2(55)SE?

Regards,

Brandon

[rahurao]

Yes