cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1492
Views
0
Helpful
7
Replies

Best practices for trunking with cisco devices in a mix environment.

JeanHarvey
Level 1
Level 1

Hi all,

We are in the process of updating our network with new cisco 2960x switches.  We have 2 Brocade core switches in MCT cluster configured with lag links with our old Nortel Baystacks.  We started installing new cisco 2960x stacks (4 total in different locations).  I just uplinked them with our management vlan for remote config purposes (Port in access mode on both sides).  When I configured a new port-channel (Po1) in Active mode for the first time on one of the stacks to our brocade core switch, it worked for 1-2 minutes but then all other cisco stacks lost communication.  I tried many things, but the only 2 things that really worked is to change the default PVID to a specific VLAN within all trunks in the network or to disable all spanning-tree for all vlans on the trunk ports (On the Po1 channel for exemple).  I can see messages in the logs: 

Nov 17 02:24:15: %SPANTREE-2-RECV_PVID_ERR: Received BPDU with inconsistent peer                                                                                                                                                              vlan id 1 on GigabitEthernet1/0/1 VLAN999.
Nov 17 02:24:15: %SPANTREE-2-BLOCK_PVID_LOCAL: Blocking GigabitEthernet1/0/1 on                                                                                                                                                              VLAN0999. Inconsistent local vlan.
Nov 17 03:21:27: %SPANTREE-2-UNBLOCK_CONSIST_PORT: Unblocking GigabitEthernet1/0                                                                                                                                                             /1 on VLAN0999. Port consistency restored.
Nov 17 03:26:45: %SPANTREE-2-RECV_PVID_ERR: Received BPDU with inconsistent peer                                                                                                                                                              vlan id 1 on GigabitEthernet1/0/1 VLAN999.
Nov 17 03:26:45: %SPANTREE-2-BLOCK_PVID_LOCAL: Blocking GigabitEthernet1/0/1 on                                                                                                                                                              VLAN0999. Inconsistent local vlan.
Nov 17 10:54:27: %CDP-4-NATIVE_VLAN_MISMATCH: Native VLAN mismatch discovered on                                                                                                                                                              GigabitEthernet1/0/1 (1), with XX-XXX-ST1 GigabitEthernet1/0/1 (333).
Nov 17 10:55:16: %CDP-4-NATIVE_VLAN_MISMATCH: Native VLAN mismatch discovered on                                                                                                                                                              GigabitEthernet1/0/1 (1), with XX-XXX-ST1 GigabitEthernet1/0/1 (333).
Nov 17 10:56:13: %CDP-4-NATIVE_VLAN_MISMATCH: Native VLAN mismatch discovered on                                                                                                                                                              GigabitEthernet1/0/1 (1), with XX-XXX-ST1 GigabitEthernet1/0/1 (333).
Nov 17 10:57:11: %CDP-4-NATIVE_VLAN_MISMATCH: Native VLAN mismatch discovered on                                                                                                                                                              GigabitEthernet1/0/1 (1), with XX-XXX-ST1 GigabitEthernet1/0/1 (333).
Nov 17 10:58:06: %CDP-4-NATIVE_VLAN_MISMATCH: Native VLAN mismatch discovered on GigabitEthernet1/0/1 (1), with XX-XXX-ST1 GigabitEthernet1/0/1 (333).

I then started reading about how BPDU's traveled around the network and native vlan for cisco and my brain fried.

I apologize for being lost here and i'm scared of taking all the network down in the future because of a misconfiguration accident with some vlan or some native vlan.  At the sametime,  I want to make sure were protected against any loop.

I searched the net alot trying to find the best practices of implementing cisco devices, but could not find a clear path.

Any advice would be greatly appreciated.

Best regards,

1 Accepted Solution

Accepted Solutions

No the Cisco is not aware of other vendor equipment, it follows the dot1q standard which other vendors follow.

With regards why it's worked elsewhere, there's nothing to say it won't work, packets sent from one side that were coming from 999 will be treated as 1 on the other side, the Cisco merely points out the fact that they're different!  The problem you have is spanning tree in different environments causing issues with a bpdu mismatch.

View solution in original post

7 Replies 7

Simon Brooks
Level 1
Level 1

Change your native vlan from 333 to 1.

Yes  best practice use 1 native vlan not 2 or you will have issues on local layer 2 network , 1 side is set to 333 and one side is default 1 cdp has picked this up , these need to match at layer 2 on either side of the trunk connection

Sorry for replying so late, thanks very much for your responses.  What i've posted earlier was some logs of testing.  I have tested with native Vlan identical on both sides of that trunk (Brocade on one side, cisco on the other) but without success.  101$ buck question:  Does it matter if other trunks of other brands in other parts of the network don't match?  I did further testing and what's strange is that the only setup that works without dropping the link is setting native vlan to 1 on cisco side and 999 vlan on the other brand side.  999 Vlan is a normal common vlan we use in our network.  That vlan is sometimes tagged in some trunks and sometimes untagged in other trunks (VMware etc.)

Hi

Does it matter if other trunks of other brands in other parts of the network don't match?

If you mean can you use cisco to hp or dell to cisco trunked no it does not matter mixed brands can trunk together once they both use the dot1q standard trunking protocol , using mixed native vlans will cause issues though throughout same lan its not recommended practice

link is setting native vlan to 1 on cisco side and 999 vlan ...are you specifically setting both sides as native here ?

Ok so if I understand correctly, we can trunk everything if it is dot1q.  But we need to set native vlan the same on both sides of a given trunk, specially if we have Cisco on one side of that trunk.  I solved the issue but I don't know why it works and that's what scares me.  My setting that work is:  Cisco Device (native vlan set to 1 and .dot1q) trunked to a Brocade (tagged Vlan x,x,x and untagged 999).  I've also solved other parts of our net like that:  Cisco Device (native vlan set to 1 and .dot1q) trunked to an Avaya Baystack (Other tagged vlans + Default PVID set to vlan 999).  Is there something magical about that 999 vlan other then that it's untagged?  Also, my question was, is cisco equipement aware of let's say Avaya trunked to Brocade?  And is there a specific way of doing things (Best practice) in mix environment?  Thanks for the help, it is much appreciated!

No the Cisco is not aware of other vendor equipment, it follows the dot1q standard which other vendors follow.

With regards why it's worked elsewhere, there's nothing to say it won't work, packets sent from one side that were coming from 999 will be treated as 1 on the other side, the Cisco merely points out the fact that they're different!  The problem you have is spanning tree in different environments causing issues with a bpdu mismatch.

Thanks to all again! You've got it all right!  Were going to work with BPDU filters.  I figurered that out today.  Thanks alot!

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: