CBS350 RSPAN defect - CSCva97586

[Dave K]

We may be incurring RSPAN issues related to CSCva97586.

The bug states "RSPAN—If traffic is simultaneously forwarded to a destination port due to a mirror operation and another
operation (such as regular forwarding), not all traffic is mirrored to the RSPAN destination port."

It is unclear to me whether the context of "... and another operation..." is limited to the physical interface which is the RSPAN destination, or whether the context is any other packet forwarding operation within the switch itself even if not associated with an active (R)SPAN source or destination. 

If the meaning is any other operation on the physical port, then removing the port from all VLAN participation and just using the port as an RSPAN destination would be a workaround.  I've done that, yet there are indications that the RSPAN session is still dropping packets.

In this application, the RSPAN destination's source is VLAN '1', most of the access ports and traffic through the switch is simply untagged VLAN 1.  For management, the switches share a common trunk backbone with other switches and routers, so most of the traffic on those trunks is simply passing through.

Could someone from Cisco please elaborate on the scope of this bug, and whether or not RSPAN can work correctly on these products?

Thanks,

--Dave

[marce1000]

                   >....Could someone from Cisco please elaborate on the scope of this bug...
      - Please note that the forum is used by Cisco customers ; to get into contact with Cisco : 
       https://www.cisco.com/c/en/us/support/web/tsd-cisco-small-business-support-center-contacts.html

 M.
                     

[Dave K]

M.,

I tried a total of four times to reach out to Cisco for clarification, but was unable to get past the front line people.  Ultimately either I was unable to guess who Cisco thought owned the products and/or all of the S/N's were outside the time frame they are willing to answer customer's questions.  One Cisco employee suggested that I needed to put the question to the authorized distributor as to how to interpret the referenced bug ID.

--Dave

[Dave K]

Since Cisco chose not to engage on this topic, we dug into this on our own.

The conclusion of this is that there is no work around to the RSPAN problem(s), and there are issues beyond the packet loss described in CSCva97586.

In addition to some packet loss, about every 45 to 60 minutes the switch will start re-ordering the RSPAN packets, over the course of a few minutes the error rate climbs, crests, then falls off, only to repeat this over and over ad infinitum.

We are also seeing a majority of the CBS350's stuck at 100% CPU loading, but that appears to be a different issue as units that are not stuck at 100% exhibit the same cyclic RSPAN behavior.

Bottom line, RSPAN does not work correctly in this product.

--Dave