8.1MR3 Beta Availability

Javier Contreras · ‎10-02-2015

8.1.123.x Available - 8.1MR3 Beta

Beta process is closed, final release 8.1.131.0 is now posted in CCO

Support

For any new issue during this test, please write to wnbu-mrbeta@external.cisco.com, pre-existing problems would be directed through normal TAC support channels, the image is TAC supported until CCO release of final code

Image types

AS_5500* can be used for Wism2/2500/5508 controller types. Not 5520

AS_7500* can be used for 7500/8500/5520

AS_CTVM can be used for vWLC

ap1g4-ME can be used for AP 1850 Mobility Express

This release is exclusively bug fixes

Issues fixed in this release

8.1.123.20 CSCuv92719    vWLC crash on 8.1.111.0 serving the RF dashboard web page
CSCuw29419    Cisco Wireless LAN Controller Radius Packet of Disconnect Vulnerability
CSCuu68490    duplicate radius-acct update message sent while roaming
CSCuv44568    Downloading RADIUS AVP via GUI does not load Vendor IDs correctly.

8.1.123.16 CSCuu86587 DN1/DN2: Need closed loop Niagara/Cavium fan control

CSCuv72545 Mobility Express Controller takes a long time to come up in Day 0 mode

CSCuv92719 vWLC crash on 8.1.111.0 serving the RF dashboard web page

CSCuw48570 ME UI: Apply button in creating WLAN is not working

CSCtu45614 Spectrum Management Bit Should be set to 1 all the time

CSCuv72545 Mobility Express Controller takes a long time to come up in Day 0 mode

CSCuw30799 Kernel panic at IPReassembler on internal AP , while pre-image download

CSCuw47234 ME: Transfer upload datatype coredump broken

CSCus53495 DFS detection due to broadcom spurious emissions 2700/3700

CSCut98006 DFS detections due to high energy profile signature on 2600/3600

CSCuu34220 Observed traceback in wlc doing rogue test.

CSCuv43057 ME wlc crash on ConfigSynctoAp

CSCuv74590 Please delay reboot if image upgrade is in progress

CSCuv88543 ME-UI]: U/P with special characters under Admin acc GUI Kicks off.

CSCuw37896 Corsica 8.2 MR2 M.E. AP cert date is newer than controller build date

CSCuw38455 sanity: corsica ap1832 unable to detect rogue ap

CSCuw41770 ME: HTTP download through web UI

8.1.123.12 CSCur71315 1552 bridge Transmit voice queue stuck

CSCuq79283 Sanity-Beni: ap1142n 1042n 1140 traceback dot11_set_frame_templates

CSCus61679 Problem in Client Stats Reports - Follow up CDETS

CSCut07170 AP1532 WGB intermittently doesn't see incoming probe responses

CSCut33114 LAG with 1G does not work in 5520 and 8540

CSCut83422 vwlc SN changed after mgmt interface ip change

CSCuu05565 NDP packets not tx'ed on secondary20 channels

CSCuu32049 WGB attempts to send IAPP change channel event during scanning

CSCuu37437 WLC8510 crashing while NMSP polling in progress.

CSCuu54100 Switching between SSIDs fails with FAST SSID enabled on PMIPv6 WLANs

CSCuu72366 WLC 5508 Silent Crash 8.0.110.11 mmListen process

CSCuu82416 Evaluation of wlc for OpenSSL June 2015

CSCuu85603 flex+MeshAP shw's Incorrct chnnel Info in O/P 'sh mesh convergence' list

CSCuu86265 Flex AP Local switching WLAN specific losing VLAN Mapping Configuration

CSCuu98667 vWLC: Low Scale VM auto reboots while downloading image -2GB default Mem

CSCuv00107 PMIPv6 Client MAC Address shows up on the mac address table of Switch

CSCuv00598 Optimized Roaming per WLAN feature

CSCuv00856 CleanAir process fails to start for 1572 AP when using 169

CSCuv28555 3702 - Voice Queue stuck

CSCuv34946 EOGRE and PMIPv6 client fails to move to Run state

CSCuv37613 Apple devices failing 802.11r FT roam

CSCuv47241 5500 WLC crashed with SNMP walk on 8.2.1.43

CSCuv47679 EoGRE client traffic packet with DF bit set is getting fragmented

CSCuv48219 SCTG ST:Module doesnot come up after AP reboot in Centralised mode

CSCuv51521 Active WLC should send GARPs when HA Re-Paring after Active-Active state

CSCuv54337 ME: "Error while opening a file" before switchdrvr bootup

CSCuv57159 ME: MC2UC conversion not happening

CSCuv61883 Error message to be corrected for error popped up(a generic issue)

CSCuv62540 Adding -S domain support for Hong Kong

CSCuv63073 ME: Confusing msg when next-preferred-master is the current Master AP

CSCuv74068 AP3700 low throughput with 3700 WGB 40mhz

CSCuv80866 CMEB-52 WLAN Security 'WPA2 Personal' has WPA AES policy enabled

CSCuv81907 ME: Show time has garbage on NTP auth status

CSCuv83856 Read/Write user is unable to edit settings from GUI

CSCuv86624 ME: HTTPs Day-1 GUI connection presents WebAuth cert instead of WebAdmin

CSCuv87657 8.0/8.1 WLC's fail to send FRAMED-IP attribute to AAA server

CSCuv90042 Fastlocate distant clients have strong rssi value from AP intermittently

CSCuv90508 Wlan Count: Deleting last wlan does not update wlan count

CSCuv90567 ME: Changing controller country code does not update Internal AP

CSCuv90769 Unable to schedule pre-image download for 12 - 12.30PM

CSCuv92645 SNMP walk on clsDownloadFileType returns invalid enum for ME controller

CSCuv93732 Hyperlocation not working after adding WLC to CMX

CSCuv96344 ME: WLC crash files shouldn't be in /storage/cores

CSCuv97538 Mu-beamformee field in VHT Cap IE is enabled in assoc resp

CSCuw02126 HALO: Last digit of NTP server is dropped when address has 12 digits

CSCuw02922 show rules is empty for mobility express controller

CSCuw04344 Module doesnot come up after AP reboot in Centralised mode on AP3700

CSCuw13898 RLAN client drops from the WLC when b radio Admin status get disturbed

CSCuw16735 ME emWeb Crash

CSCuw19551 ME : show time not shown as "AUTH SUCESS" when NTP Sync happens

CSCuw22214 11v shows 'not supported' under client detail for 11v capable client

Leo Laohoo · ‎10-02-2015

CSCut33114

LAG with 1G does not work in 5520 and 8540

Thanks Javier. I find this bug ... weird.

According to the 8540 Deployment Guide it states "A single LAG across the 4 x 10 G interfaces is supported in software versions 8.1 and later." This statement leads me to believe LAG with SFP is not supported.

Javier Contreras · ‎10-05-2015

Hi Leo

the DG mentions "a single lag", to emphasize that multiple lag interfaces are not supported at this time

The 10g is not a restriction on 8540/5520, as they physical ports support 1G/10G SFP

it was a restriction on 8510/7500, so probably a leftover in documentation that may need correction

The bug fixes handling of the 1G link state for 8.1

regards

Leo Laohoo · ‎10-06-2015

Ok, thanks for the explanation.

So this means that 5520/8540 will (version 8.1.123.X and later) support 1G LAG? Am I reading this correct?

Javier Contreras · ‎10-06-2015

Yes.

Leo Laohoo · ‎10-06-2015

Ok. Thanks for the confirmation.

I hope someone can fix up the "document bug".

ajc · ‎03-09-2016

Hi Javier,

We are hitting bug CSCuw06153 and we are planning to move into 8.1.131.0. Any feedback about the reliability on this new release?. thanks. We have 5508 and 8510 in our environment (over 125k+ users).

thanks

Javier Contreras · ‎03-09-2016

hi Abraham

we do not plan to do further MR releases for 8.1 at this time, so if a migration is needed, it is recommended to invest certification/validation time into 8.2, if you need to upgrade

if you absolutely need 8.1 based code, send me private message, and I can share the list of known "bad" issues for your evaluation

if you are on 8.0 code now, you can migrate directly to 8.0.121.0 for this bug

regards

ajc · ‎03-21-2016

Apple TV does not work on 8.0.121.0, 8.1.131.0. Something was changed so looks like mDNS is no longer supported. I have 5500 and 8510. We have a 500+ school environment with over 120k+ endusers and many AppleTV devices being used so I cannot upgrade to fix the security breach. Any thoughts?

majhensgens · ‎10-09-2015

Is there coming any fix to te bug causing the 802.11a radio resets due the DCA interval? We've tried 8.1 MR1 and MR2 but the bug persists. I had to freeze the RRM DCA cycle to prevent the bug from occuring.

Regards.

Javier Contreras · ‎10-13-2015

Hi

A channel change "requires" a radio reset. if you see a large channel change count, please open TAC case to investigate. The change maybe a symptom not the root cause

regards

majhensgens · ‎10-13-2015

I'm know that an DCA channel change causes an radio reset. The problem is that there is no channel change, tx level change or radar and it still resets the radio with an Unknown reason. (The timing is the at or after the DCA interval).

An overview of the resets code is presented below:

AP_109360#show controllers dot11radio 1 radio-reset
Last radio reset code: 88
Radio resets - total:65 retries:0 failed:0
Reset Stats: Start Cnt: 10, Recovery: Cnt 0, Last Ret: 0, Fails: 0, Recvry Status: Stalled NO, In Prog NO
Code/Count: 37/00059 1 62/00003 24AC 67/00001 0 88/00002 24AC

Javier Contreras · ‎10-13-2015

hi

37 -> IDB interface reset (main contributor in you list, with 59 events)

This is generic reset, it may have several triggers, not all pointing to DCA

Things like RLDP or auto contention can be directly responsible for this, independently of DCA

Check if you have those features enabled

62 is manual channel assignment

67 is due to hostname change

88 is due to client deauth (Special recovery case due to pending packets)

There are possibilities of troubleshooting this further, as AP can register the reset caller, not just the type, that is why I recommended TAC process. That should be part of separated discussion, outside 81MR3 release thread. ;-)

regards

Mathieu Hofstra · ‎10-22-2015

We have vWLC with several 1602i and 1702i APs in flexconnect local switching mode.
Previous week we upgraded from version 7.6.x to 8.1.122.0 and we enabled 802.11r and 802.11k on a 802.1x secured SSID.
After a few days some users with Apple devices were complaining of reconnect issues.
After investigation it looked like we ‘hit’ bug CSCuv37613.
Yesterday I installed version 8.1.123.20 but after some running around in the building with my iPhone 5s the issue still occurred. So it seems to me bug CSCuv37613 is not solved (or we are running in another issue). After disabling 802.11r the reconnect problems seem to be resolved (but no fast roaming anymore…).

Charles Dahlgren · ‎11-04-2015

Any idea when the actual release will be out? The TAC told me "end of October" two weeks ago.