Request Feedback v1.1.0 - PLEASE

[sirhc]

Chris,

Looks like I'm the only naysayer, Sorry.

When the switch is configured without STP or LAG enabled the switch doesn't handle BPDU packets correctly.
Specifically in the case where I'm noticing this we run RSTP through the switch. We don't want the switch participating in spanning tree and so far have kept STP disabled on it. I'm running RC19. I upgraded it from stable because we were seeing some leakage of BPDUs between VLANs and noticed the notation about the fix. I have MC enabled on all ports.

I haven't been able to wire shark it to be sure about what's happening but it appears the switch is acting as the root bridge even though STP is turned off. To boot with stp turned off in the switch and it acting as root all other bridges don't see the correct topology causing a loop.

PS, I was up till 5am upgrading and total forgot to hit submit.

We will lab this tomorrow and see whats up, please email your config to me, chris@netonix.com

But your switches are functioning OK you just feel BPDU packets are mis-handled?

[LRL]

In follow up to my above post, it acts like once you disable stp on the swtich doesn't always stop or actually disable. I cannot reproduce it all the time.

I've been able to work around the switch not passing BPDUs on most of our towers by configuring the switch for rstp, but there is one tower where we run rstp through the switch on different VLANs. Unfortunately I can't turn rstp on for all ports on the switch because PVRSTP isn't supported. Turning RSTP on for ports in different VLAns causes spanning tree to detect a loop.

[sirhc]

In follow up to my above post, it acts like once you disable stp on the swtich doesn't always stop or actually disable. I cannot reproduce it all the time.

I've been able to work around the switch not passing BPDUs on most of our towers by configuring the switch for rstp, but there is one tower where we run rstp through the switch on different VLANs. Unfortunately I can't turn rstp on for all ports on the switch because PVRSTP isn't supported. Turning RSTP on for ports in different VLAns causes spanning tree to detect a loop.

Please email a copy of the config with an explanation how to replicate to me and I will have Rory and Eric look into it tomorrow.

[LRL]

PS, I was up till 5am upgrading and total forgot to hit submit.

We will lab this tomorrow and see whats up, please email your config to me, chris@netonix.com

But your switches are functioning OK you just feel BPDU packets are mis-handled?

I believe the big issue for me is BPDUs not being passed when STP is disabled. It's pretty easy to lab this one.
RSTP switch or bridge with two links to a WS, disable STP on the WS.

Some of the other stuff is just weird:
The weird stuff has ranged from the WS retaining root bridge even after RSTP was disabled
WS not honoring priorities properly
WS leaving a port in blocking when RSTP was disabled.
This all seems like the changes submitted on the UI didn't get implemented.

I'll send you two of our configs, bare in mind I've changed them to work around the issue.

Part of me thinks implementing PVRSTP or rather MSTP may solve some of the issues better?

As I'm not entirely sure what was causing the BPDU issue before, but it sounds like MSTP might have solved that?

I'll send those configs later tonight after our AF install.

[sirhc]

Also please include your cell number in the email so that we may contact you (specify your time zone)

[adairw]

Running r19 on a switch stuck up a tower with fiber ran to it. No problems.

What's the specific configuration, if you don't mind?

[Rory]

LRL I look forward to seeing your config. I tested a few configurations, most realistic, and some deliberately mis-configured. The only time I saw weird STP behavior was when I deliberately created a double loop through a non-STP bridge (WS-24) between two other stp enabled devices (a WS-12 and another switch) , once all the lines were connected all three were in a ring. That caused me to recieve 163 stp packets that look like they originated from the non-STP switch. If I had to guess I would assume they were in some way being mangled by the switch after some random internal error occurred during the loop. I actually performed the test several times, and I only saw the spurious packets once.

Unfortunately I did not have time to test every scenario. I will try more exhaustive testing tomorrow, after I have some sleep and coffee. Please send your evidence as soon as is convenient, I want to get this tested as quickly as possible.

[LRL]

Running r19 on a switch stuck up a tower with fiber ran to it. No problems.

What's the specific configuration, if you don't mind?

To summarize the problem, bridges after the switch don't receive BPDU updates correctly(or at all) if stp is not enabled in the switch.

[adairw]

Running r19 on a switch stuck up a tower with fiber ran to it. No problems.

What's the specific configuration, if you don't mind?

To summarize the problem, bridges after the switch don't receive BPDU updates correctly(or at all) if stp is not enabled in the switch.

I was actually asking mayheart, sorry.

[LRL]

Ha!

Totally need to quit using my phone for forum replies!

Any update on the stp issue Chris, Rory?