Wireless WISM and 4402 DMZ controller mobility issue

[kevin.woodhouse]

Hi all,

I have a really weird wireless mobility issue I need help with. I have a Wism installed in our cat 6 with version 7.0.98.0 of software. I also have a 4402 controller in the DMZ acting as an anchor controller to terminate guest traffic, also running the same version of software. The mobility anchor is setup on 4 SSID's and most of the time I don't have an issue with it.

However every week or so the data path or control path will randomly go down, so no guest traffic. Timers are configured as 10 seconds for the keep alive count and 20 seconds for the keep alive interval. I have found the following syslog information from our NMS. Not sure what these errors are and if its anything to do with this random mobility issue.

Any thoughts ??

[Nicolas Darchis]

tracebacks are not cool !

Sounds like a serious problem. I'd advise going for a TAC case at this point.

Nicolas

[kevin.woodhouse]

I raised a TAC yesterday on this issue as the mobilty fell over at leat 6 times. I used our NMS to alert me when the syslog from the controller lost the anchor, I now have six in my mailbox.

Can you explain what a traceback is though ??

[Nicolas Darchis]

A traceback is a device producing a trace of where it was in the software code when the problem occured.

Most often it's when a function fails/crashes. Although that is not always true.

If you have the syslog to the deepest level of debug, you will see harmless tracebacks (functions that returned no result but that might be expected). On higher level of debugs only "crashing" tracebacks are shown. In this case, the fact that you have mobility issues and tracebacks related to mobility is highly suspicious.

TAC can decrypt the traceback hex and find the exact part of code that created the problem. From there, they can link it to an existing bug (or a new one maybe ?).

Hope this clarifies !

Nicolas

===

Please rate answers that you find useful