The bug (thru cisco's bug tool) specifically calls out a customer using
Symbol scanners, and having them all power on at once.

 

I wasn't going to post because I thought you had read the actual bug
text.

 

>From the Cisco bug tool  (Which is working a bit spastically this
morning)

 

Symptom:
symptom
When 200+ wireless clients are trying to associate to a WLC at the same
time,
the WLC starts experience problems:
1. scanners stuck in DHCP_REQD state. The attached sniffer trace 
shows that the WLC receive DHCP offer from an external DHCP 
server, but the WLC does not send out the DHCP offer in LWAPP
2. the following message is logged in show tech:
apf_policy.c:258 APF-1-MOBSTA_ADD_FAILED: Unable to add mobile
xx:xx:xx:xx:xx:x to PEM module:
3. CPU is running high (e.g. 70+%)
symptom
When the partner power off 200+ Symbol scanners, the WLC
starts experience problems:
1. scanners stuck in DHCP_REQD state. The attached sniffer trace 
shows that the WLC receive DHCP offer from an external DHCP 
server, but the WLC does not send out the DHCP offer in LWAPP
2. the following message is logged in show tech:
apf_policy.c:258 APF-1-MOBSTA_ADD_FAILED: Unable to add mobile
00:15:70:32:5a:b5 to PEM module:
3. CPU is running high (77%)
4. A sniffer trace on the WLC shows that the WLC receives DHCP offer
from an
external DHCP server, but the corresponding DHCP offer is not sent to
the AP
in LWAPP.


Conditions:
The problem is verified in 4.0(217.0) and 4.1(171.0).


Workaround:
None


Further Problem Description:

 

From: Lee H Badman [mailto:[EMAIL PROTECTED] 
Sent: Friday, October 05, 2007 9:00 AM
To: [email protected]
Subject: Re: [WIRELESS-LAN] Open Cisco COntroller Caveat/DHCP issue

 

Hi Frank-

 

I would hope. But the wording leaves a lot to the imagination, and we
are seeing enough oddities on occasion that could point at something
like this that clarification is in order, if nothing more than for a
sanity check.

 

Lee

________________________________

From: Frank Bulk [mailto:[EMAIL PROTECTED] 
Sent: Friday, October 05, 2007 8:45 AM
To: [email protected]
Subject: Re: [WIRELESS-LAN] Open Cisco COntroller Caveat/DHCP issue

 

Lee:

 

I think the key phrase is "at the same time".  This may be a bug found
when Cisco or someone else did scalability testing with test tools, not
a likely event in production.

 

Frank

 

________________________________

From: Lee H Badman [mailto:[EMAIL PROTECTED] 
Sent: Friday, October 05, 2007 7:29 AM
To: [email protected]
Subject: [WIRELESS-LAN] Open Cisco COntroller Caveat/DHCP issue

I'm trying to get clarification on this open caveat, but so far can't
get a clear answer on the specifics of the bug:

 

CSCsj25953-When 200 or more wireless clients try to associate to a
controller at the same time, the clients become stuck in the DHCP_REQD
state. The controller receives the DHCP offer from an external DHCP
server but does not send the offer to the access point in LWAPP.

 

Obviously, getting to 200 clients on a single controller is routine
operations on a busy network, especially when one controller has 150
associated access points. Has anyone else dug in on this one, and gotten
any real details? It sounds potentially catastrophic, or that it could
be relatively harmless, but without more detail it's hard to know...

 

Regards-

 

Lee H. Badman

Wireless/Network Engineer

Information Technology and Services

Syracuse University

315 443-3003

 

********** Participation and subscription information for this EDUCAUSE
Constituent Group discussion list can be found at
http://www.educause.edu/groups/. 

********** Participation and subscription information for this EDUCAUSE
Constituent Group discussion list can be found at
http://www.educause.edu/groups/. ********** Participation and
subscription information for this EDUCAUSE Constituent Group discussion
list can be found at http://www.educause.edu/groups/.


**********
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.

Reply via email to