Traffic dropping between Cisco 2960 s and HP Procurve  2910a1

tamray_tech
tamray_tech used Ask the Experts™
on
I am working with a school that uses VOIP. The network ran fine until yesterday. The school does not believe they changed anything, but the network is unstable.  Each desktop runs through the Cisco phone for network connectivity. The phone is vlan 72, the desktop PC is on vlan 1. Pings to anything on vlan 1 will run for a minute, or so and then drop for 15-30 seconds. If I eliminate the Cisco and connect a desktop directly to the HP Procurve, vlan 1 traffic is flawless..

I get the following error on the cisco, but maybe its more than informational:
SPANTREE-2-RECV_BAD_TLV: Received SSTP BPDU with bad TLV on GigabitEthernet1/0/48 VLAN1.
it repeats itself every few minutes.

I have disabled all ports, but one phone (with desktop) and the trunk port. The problem still persists.

Cisco trunk port config:

Tried both ports 47-48

interface GigabitEthernet1/0/47
 description ** Uplink to Core2 **
 switchport mode trunk
 srr-queue bandwidth share 10 10 60 20
 queue-set 2
 priority-queue out
 mls qos trust device cisco-phone
 mls qos trust cos
 auto qos voip cisco-phone
 spanning-tree portfast
 service-policy input AUTOQOS-SRND4-CISCOPHONE-POLICY

The tried this port:

interface GigabitEthernet1/0/48
 description ** Voice Uplink to LCTN **
 switchport trunk allowed vlan 72,1,109,110,111
 switchport mode trunk
 srr-queue bandwidth share 10 10 60 20
 queue-set 2
 priority-queue out
 mls qos trust device cisco-phone
 mls qos trust cos
 auto qos voip cisco-phone
 spanning-tree portfast
 service-policy input AUTOQOS-SRND4-CISCOPHONE-POLICY

HP Port to Cisco switch:

Vlan 1 untagged
Tagged all other Vlans


I believe it could be a spanning tree issue. If so, I would need assistance with the proper commands on each switch to set priority, etc...


The stack consists of 2 Cisco 2960s and 1 Procurve 2960
Comment
Watch Question

Do more with

Expert Office
EXPERT OFFICE® is a registered trademark of EXPERTS EXCHANGE®
View Solutions Only
Bryant Schaper

Commented:
could be a misconfig between the HP and Cisco, what does that trunk port look like in the config.
tamray_tech

Author

Commented:
That is in my original post. Tried port 47 and 48. Same results

The HP doesn't have anything port specific. Just all vlans tagged, but vlan 1.
Distinguished Expert 2018
Commented:
According to CIsco's link:
Error Message    SPANTREE-2-RECV_BAD_TLV: Received SSTP BPDU with bad TLV on [chars]
[chars].

Explanation    The listed interface received an SSTP BPDU without the VLAN ID tag. The BPDU is discarded. The first [chars] is the interface, and the second [chars] is the VLAN that received the SSTP BPDU.

Recommended Action    If this message recurs, copy the message exactly as it appears on the console or in the system log. Research and attempt to resolve the error by using the Output Interpreter. Use the Bug Toolkit to look for similar reported problems. If you still require assistance, open a case with the TAC, or contact your Cisco technical support representative, and provide the representative with the gathered information. For more information about these online tools and about contacting Cisco, see the "Error Message Traceback Reports" section on page 1-9.
Recommended Action    If the message recurs, copy the message exactly as it appears on the console or in the system log, call your Cisco technical support representative, and provide the representative with the gathered information
The strange thing is that this happens to you on VLAN 1 and that one according to what you are saying is native VLAN on both sides. The only thing I can think of is to double check native VLAN (untagged) on every link, and check # sh process cpu history (loop in the network?)
Don JohnstonInstructor
Top Expert 2015
Commented:
As Predrag mentioned, this error is due to receiving a BPDU without a VLAN ID.

Which is going to happen when you connect a Cisco switch running Per-VLAN Spanning-Tree to a switch that isn't.  And HP's don't do Per-VLAN Spanning-Tree so there's no VLAN ID in the BPDU.

Whenever you connect a Cisco to an HP, you have to configure the switches to run the same spanning-tree protocol. In this case, 802.1s Multi-Spanning-Tree Protocol.  

Now you say that nothing has been changed on the network. It's possible that this has been happening all along and just not caused any problems.  

Or maybe something did get changed.  It wouldn't be the first time I have seen that.
tamray_tech

Author

Commented:
Not wanting to make the original post expansive, I only provided detail pertaining to the current errors I was having. Originally, there was a spanning tree event, which I traced to a server with two nics, one on vlan 1, the other on vlan 72.  The school admin had installed some virtual software that ultimately bridged the two nics. Once that was resolved, my errors were reduced to what was mentioned in my ticket. It seemed from my research that those errors were more informational than culprits of an unstable network. Not finding an obvious fix, I restarted the HP Procurve. With that, everything started working properly again, outside of the other errors. It seems the HP was unable to recover from the original spanning tree blocks, the same as the Cisco switches were able to.

Thank you for your assistance!

Do more with

Expert Office
Submit tech questions to Ask the Experts™ at any time to receive solutions, advice, and new ideas from leading industry professionals.

Start 7-Day Free Trial