Cisco Cisco 2000 Series Wireless LAN Controller Release Notes
18
Release Notes for Cisco Wireless LAN Controllers and Lightweight Access Points for Release 3.2.116.21
OL-9542-01
Caveats
•
CSCsd02525—If a Cisco 1200 series access point is reset while the 802.11a network is disabled, the
access point sends a change state event to the controller when it comes back up indicating that its 802.11a
radio is operational, but the power-over-Ethernet (PoE) power detection takes some time to complete.
access point sends a change state event to the controller when it comes back up indicating that its 802.11a
radio is operational, but the power-over-Ethernet (PoE) power detection takes some time to complete.
•
CSCsd03939—Enabling or disabling the 802.11g network causes the access points joined to the
controller to reboot. This is normal operating behavior. However, no messages appear to explain why
the access points suddenly reboot when WCS templates are applied and change the 802.11g state.
controller to reboot. This is normal operating behavior. However, no messages appear to explain why
the access points suddenly reboot when WCS templates are applied and change the 802.11g state.
•
CSCsd04657—The primary controller sends an LWAPP discovery response even when the
maximum number of access points are joined, which causes any new access points to attempt to join
the primary controller without trying the secondary or tertiary controller.
maximum number of access points are joined, which causes any new access points to attempt to join
the primary controller without trying the secondary or tertiary controller.
•
CSCsd09768—The AP1000 loses its Ethernet buffers due to an invalid ARP packet generated by the
neighbor switch. As a result, it is unable to receive any packets, including the DHCP offer, and becomes
stuck in the DHCP Discover state.
neighbor switch. As a result, it is unable to receive any packets, including the DHCP offer, and becomes
stuck in the DHCP Discover state.
•
CSCsd09898—If a client has crypto handles that are set to 0, which is an invalid value, the traffic stops
flowing for that particular client. Other clients with valid crypto handles are unaffected.
flowing for that particular client. Other clients with valid crypto handles are unaffected.
•
CSCsd11774—LWAPP-enabled 1130 and 1230 access points may fragment large packets into two files
after the LWAPP header is added and then do not send the second fragment to the controller.
after the LWAPP header is added and then do not send the second fragment to the controller.
•
CSCsd14113—When the controller receives a delayed Access Request (with an ID out of order) from
an MS-PEAP client, it adjusts the order of the packet and sends the request to the RADIUS server.
However, the client fails to authenticate.
an MS-PEAP client, it adjusts the order of the packet and sends the request to the RADIUS server.
However, the client fails to authenticate.
•
CSCsd21147—When a client device roams from one controller to another, the ARP filters for the client
are not deleted properly.
are not deleted properly.
•
CSCsd21248—A memory leak in buffer pool 0 in the NPU driver causes CPU/NPU communications to
fail.
fail.
•
CSCsd22087—Intel 2200 b/g clients cannot pass traffic with an AP1000 in 802.11g mode. This
condition remains for 5 to 60 seconds and then recovers, but it can occur quite frequently.
condition remains for 5 to 60 seconds and then recovers, but it can occur quite frequently.
•
CSCsd32642, CSCsd23190, and CSCsc95614—The max retry count from the controller is being used
as the total number of retries allowed per packet rather than the number of retries per rate. Therefore, the
controller cannot rate shift down if a client needs lower rates.
as the total number of retries allowed per packet rather than the number of retries per rate. Therefore, the
controller cannot rate shift down if a client needs lower rates.
•
CSCsd44941—The 2.4-GHz radio in some AP1130 and AP1240 units produced after 1/23/2006 can
enter a state in which all clients disassociate and the access points can no longer communicate with the
clients.
enter a state in which all clients disassociate and the access points can no longer communicate with the
clients.
Open Caveats
These caveats are open in operating system release 3.2.116.21.
•
CSCar14535—When configuring a mobility group anchor that is not part of the mobility member
list, the controller displays an “Invalid Parameter Provided” error message.
list, the controller displays an “Invalid Parameter Provided” error message.
Workaround: Make sure that the anchor controller is a mobility group member.
•
CSCek16101—The Controller Network Module does not have the correct time when it is first booted
up because it does not have a real-time clock. Therefore, it contacts the NTP server during the initial
bootup to obtain the correct time. If the controller cannot reach the NTP server, it eventually gives
up and boots without the correct time. When this happens, access points cannot register with the
controller. The controller must wait for at least one hour before trying to poll the NTP server again for
the correct time.
up because it does not have a real-time clock. Therefore, it contacts the NTP server during the initial
bootup to obtain the correct time. If the controller cannot reach the NTP server, it eventually gives
up and boots without the correct time. When this happens, access points cannot register with the
controller. The controller must wait for at least one hour before trying to poll the NTP server again for
the correct time.