What’s new in WLC 7.6.110 – 8.0.100

A while back I wrote an article that covered the changes from WLC 7.1 – 7.6.100 . Let’s catch up to 8.0, shall we?

7.6.110

  • Bugfix release.
  • Fix for issues with WMM with Broadcom clients (no 802.11n for you)
  • Fix for an issue with the AP3700 and replay counters which apparently causes major performance problems on 5GHz.

7.6.120

  • Adds support for 2700 series and 700W series APs.
  • Adds “Cisco WLAN Express Setup” for 2500 series controllers. The notes say: “It includes easy to use GUI Configuration Wizard, an intuitive monitoring dashboard and several Cisco Wireless LAN best practices enabled by default.” Sounds nifty. I need to get a 2504 for my lab… If you are upgrading a 2500 to this release, there’s a decent chunk of steps involved to enable this feature. More info about the feature and the steps here: Cisco WLAN Express Setup for Cisco 2500 Series Wireless Controller.
  • Of course, lots of bugfixes.
  • Several crashes fixed.
  • Obligatory security fixes.
  • False DFS positives fixed.
  • If you really want to see if your favorite bug is fixed, check here: 7.6.120 Resolved Caveats

7.6.130

  • Bugfix release with an even longer list of resolved caveats (7.6.130 Resolved Caveats).
  • More crash fixes and obligatory security fixes. The ones that jumped out at me below.
  • 99% CPU usage fix.
  • Apple auth problems fixed.
  • Fix for CAPWAP disassociation due to DTLS errors
  • Vocera broadcast failure fix
  • RAID volumes get proper status codes
  • vWLC Service Port issue with distributed vSwitch fixed
  • Annoying MFP anomaly messages fixed (but I’m still seeing them)

Now the really interesting stuff. Major releases are always fun, at least once the first round or two of bug fixes come in. Let’s see what’s shiny, shall we?

[Note: I've since found this Cisco Wireless Release 8.0 document, which has a nice summary of the features.]

8.0.100

  • Cisco Aironet AP and Scale Features
    • Keep-alives now sent over both control and data CAPWAP tunnels.
    • New Flex+Bridge mode enabled FlexConnect functionality across mesh APs. This means if the wired link goes down, your AP can failover to mesh backhaul. (Not supported on 1130 and 1240. No surprise.)
    • Mesh fast convergence. Automatically sets faster convergence timers. Convergence time per hop down to 20s.
    • AP700W gets VLAN tagging
    • FlexConnect APs can be a PPPoE client. Was in 7.3/7.4, but not 7.5/7.6. Now it’s back. And it’s angry…
    • Dynamic Channel Assignment (DCA) on RF Profiles. Enables multi-country support using AP groups and simplifies mixed channel environments (40MHz/80MHz mix). Sounds like this could be useful for those of us with a mixed .11n and .11ac environment, which will probably be just about everybody soon… See Configuring RF Profiles for more information. (Not supported for mesh/bridge APs.)
    • Rx-SOP: Receiver Start of Packet threshold. #shiny Particularly helpful in high density environments. This helps reduce CCI by controlling what frames the AP will decode. The No Strings Attached Show has a nice whitepaper about it. Config information is here: Configuring Receiver Start of Packet Detection Threshold.
    • Optimized Roaming. Ooh, more #shiny! This helps with sticky clients by disassociating them based on RSSI and data rate. This will also help prevent clients from associating as they pass by. Config info: Configuring Optimized Roaming.
    • Side note: Good article covering Rx-SOP, Optimized Roaming, and RSSI low at Revolution Wi-Fi: Optimized Roaming, RSSI Low Check, RX-SOP, Oh My!
    • AP1700 support added
    • CleanAir Express for AP1600 and AP1700
    • OEAP gets basic firewall support, split tunneling, VoIP QoS
    • Increased scale of vWLC (now up to 6000 clients)
    • 2500 WLC now supports wired guests
  • Native IPv6 (if you need the exhaustive list see Native IPv6 Support)
    • Finally!
    • SLAAC for the service port
    • Full support for all the services and ways of accessing the WLC that you would expect out of v6 support.
    • DHCPv6 option 52 for controller discovery
    • CAPWAP preferred mode – you can choose v4 or v6 as preferred. v4 is preferred by default
    • List of things not supported, which will take away your initial joy:
      • FlexConnect-local switched, mesh/outdoor, teleworker/OEAP, converged acces
      • Services: mDNS, AVC, and TrustSec
      • Bridge mode APs with 64MB of RAM: 600 OEAP, ISR 800/802, 1130, 1240, 1250, 1310, 1410, 1520
      • Internal DHCPv6 server, DHCPv6 proxy, auto-configuration, dynamic interfaces, RA interfaces, OSCP and CA server URL, VLAN pooling
      • NTPv4 (typo?), MLDv2, IPSec v3 and IKEv2, RLDP and CIDS, PMIPv6, mDNS IPv6 clients, and New Mobility
      • IPv6 is not supported for HA Redundancy Interface configuration
      • Auto-RRM, Dynamic Anchoring, DNS RADIUS/TACACS+, core dump
  • Security and RADIUS enhancements
    • SPs can configure new VSAs and tell the WLC how to handle them.
    • WLC can be configured to use the realm value to determine the RADIUS server for a client.
    • WebAuth now works for HTTPS.
    • 802.1X and EAP WLANs now support sending the WLAN ID to the RADIUS server.
    • SHA256 certificate support
  • Ease of Management Features
    • SSID and WLAN profiles can be renamed (Yay! Now you can cleanup the mess!)
    • “ping” can be sourced from a dynamic interface.
    • “show ap summary” now shows the AP’s IP address. Also can search for APs based on IP in the GUI.
    • Bunch of new show system commands. They provide more info about how WLC is running.
    • show run-config startup-commands – Finally, something you can copy and paste into a controller!
    • You can globally enable/disable SSH/telnet for all APs on a controller.
    • Choice of color themes for the GUI (default and red). Helps distinguish between controllers.
    • You can now flash the LEDs on an AP to identify it. About time…
    • “show client detail” now shows AP and WLAN
    • “show ap join stats” corrects output for renamed APs
    • “debug client” now shows the AP connected and RSSI.
    • You can now update the OUI list without upgrading the controller. But it requires a reboot…
    • 802.11v. My understanding is it’s supposed to leverage 802.11k info to control client associations. Not sure if any clients actually support this.
    • 802.11r mixed mode. Yes, bold. No need for a seperate SSID for 802.11r and non-11r clients. This is very shiny. And useful.
  • High Availability Enhancements
    • 802.11ac is now supported in HA. I hadn’t realized it wasn’t supported before.
    • Handful of enhancements to HA, including faster sync and more configurability.
    • Internal DHCP now works with client SSO. The database is synced between the active and standby controllers.
  • Better policy control for mDNS
  • AVC
    • NBAR 2 protocol pack updated to 11. Heh.
    • Per app, per client rate limiting. Nice.
    • QoS marking can choose the direction instead of only bidirectional. I wonder what the use case for that is.
  • Q-in-Q support. Outer tag for AP group. Inner tag assigned by AAA.
  • VideoStream now supported for FlexConnect locally switched mode.
  • WPA/TKIP now only configurable from the CLI.

Closing Notes

WLC 8.0 is supported on PI 2.1.1. A number of the new features aren’t supported, though I expect those will be available in PI 2.2. ISE 1.2 is supported. Obviously, MSE 8.0 is supported with it. It’s not clear is MSE 8.0 upgrade is required, but it is at least implied.

I suspect this will be the last release to support the 1130 and 1240 series.

8.0.100 has a LONG list of resolved caveats (many of which are also resolved in the 7.x code base) and a decent list of open ones. If you are considering 8.0, I recommend going over those carefully: WLC 8.0 Caveats. Personally, I will likely wait for 8.0.110 before going into production with it.

FIN

This Is Not The Flash You Are Looking For

A while back, I was trying to install an IOS-XE update on an ASR1001 and run into something weird.

asr1001#request platform software package expand file ?
 bootflash: RP-relative file path
 flash: RP-relative file path

OK, sounds good, right? Nothing obviously weird, until you discover that only bootflash: actually works… Let me show why this is really confusing.

asr1001#request platform software package expand file flash:?
flash:.installer
flash:.prst_sync
flash:.rollback_timer
flash:archive
flash:asr1000-rommon.153-1r.S.pkg
flash:asr1001-universalk9.03.09.00.S.153-2.S.bin
flash:asr1001-universalk9.03.11.00.S.154-1.S-std.bin
flash:core
flash:lost+found
flash:pp-adv-asr1k-153-1.S-14-4.0.0.pack
flash:pp-adv-asr1k-154-1.S-17-8.0.0.pack
flash:tracelogs
flash:vman_fdb

asr1001#request platform software package expand file bootflash:?
bootflash:.installer
bootflash:.prst_sync
bootflash:.rollback_timer
bootflash:archive
bootflash:asr1000-rommon.153-1r.S.pkg
bootflash:asr1001-universalk9.03.09.00.S.153-2.S.bin
bootflash:asr1001-universalk9.03.11.00.S.154-1.S-std.bin
bootflash:core
bootflash:lost+found
bootflash:pp-adv-asr1k-153-1.S-14-4.0.0.pack
bootflash:pp-adv-asr1k-154-1.S-17-8.0.0.pack
bootflash:tracelogs
bootflash:vman_fdb

It looks like either one should work, doesn’t it? Let’s see what happens if you choose incorrectly.

asr1001#request platform software package expand file \
     flash:asr1001-universalk9.03.11.00.S.154-1.S-std.bin to flash:test
/usr/binos/conf/provfunc.sh: line 1991: cd: flash: No such file or directory
Verifying parameters
  FAILED: Specified package file flash:asr1001-universalk9.03.11.00.S.154-1.S-std.bin does not exist

This can lead to a bunch of wasted time replacing images, verifying checksums, and scratching your head. Then you finally try bootflash…

asr1001#request platform software package expand \
     file bootflash:asr1001-universalk9.03.11.00.S.154-1.S-std.bin to bootflash:test
Verifying parameters
Validating package type
Copying package files
SUCCESS: Finished expanding all-in-one software package.

Yep. Pretty annoying. Check out the directory listings. (I’m doing the directory listings in this odd way so they fit my WordPress theme better, just in case you were wondering.)

asr1001#dir bootflash:test/?
bootflash:test/asr1001-espbase.03.11.00.S.154-1.S-std.pkg
bootflash:test/asr1001-packages-universalk9.03.11.00.S.154-1.S-std.conf
bootflash:test/asr1001-rpaccess.03.11.00.S.154-1.S-std.pkg
bootflash:test/asr1001-rpbase.03.11.00.S.154-1.S-std.pkg
bootflash:test/asr1001-rpcontrol.03.11.00.S.154-1.S-std.pkg
bootflash:test/asr1001-rpios-universalk9.03.11.00.S.154-1.S-std.pkg
bootflash:test/asr1001-sipbase.03.11.00.S.154-1.S-std.pkg
bootflash:test/asr1001-sipspa.03.11.00.S.154-1.S-std.pkg
bootflash:test/packages.conf

asr1001#dir flash:test/?
flash:test/asr1001-espbase.03.11.00.S.154-1.S-std.pkg
flash:test/asr1001-packages-universalk9.03.11.00.S.154-1.S-std.conf
flash:test/asr1001-rpaccess.03.11.00.S.154-1.S-std.pkg
flash:test/asr1001-rpbase.03.11.00.S.154-1.S-std.pkg
flash:test/asr1001-rpcontrol.03.11.00.S.154-1.S-std.pkg
flash:test/asr1001-rpios-universalk9.03.11.00.S.154-1.S-std.pkg
flash:test/asr1001-sipbase.03.11.00.S.154-1.S-std.pkg
flash:test/asr1001-sipspa.03.11.00.S.154-1.S-std.pkg
flash:test/packages.conf

So, just remember to use bootflash and you’ll save yourself some headache and confusion!

FIN

#vBrownBag Book Contest (not) Still Going!

The contest is still going until June 30, 2014 over! There hasn’t been a huge number of entries, so your odds are very good. Get your entries in!

Contest Details

ShowCover.aspI have copies of both the CCNA Routing and Switching 200-120 Official Cert Guide Library and Networking for VMware Administrators to give away. These are courtesy of Cisco Press and VMware Press, so a big thanks goes to them!

Here’s how you can win one:
Send out a tweet with your CCNA R&S question, including the hashtag #vBrownBag and @scottm32768. If your question too long for twitter, you can post it as a comment here, then link to it on twitter. The best questions by the end of the month (June 30 2014 23:59) will win. Myself and others related to the #vBrownBag podcast will make this decision. In the event we cannot reach a decision, we will use the contents of a hermetically sealed envelope kept in a #2 mayonnaise jar on Funk and Wagnall’s back porch. Or maybe just choose winners at random.

FIN

Cisco Live 2014 Day 1

Pedometer Stats

Todays stats for the Official Unofficial Cisco Live Pedometer were:

Steps: 7611
KCals: 411
Miles: 3.6

Not as much as I expected for today. Maybe things are closer than I at first thought.

VIRL & CML Update

I had a chance to talk with some of the CML/VIRL team today and learned a few things of note:

  • VIRL (Virtual Internet Routing Labs) is back as the personal/developer edition and will be available through the Cisco DevNet. For free. Yes, you read that right!
    • I believe this will support up to 15 Cisco VMs and as many other VMs (Linux guests, etc) as fit in your RAM, but I need to verify the 15 number is still accurate.
    • The quid pro quo for this deal is they will be collecting data. They will be completely up front about this, the data they are collecting, and will transmit it in clear text. The data they will be collecting is going to be things like what kinds of virtual devices you launch, how many, what specific features you use, and that sort of general statistics. They will not collect any configuration information.
    • No NX-OS at this time. The virtual NX-OS is just not ready, yet.
    • Still no L2, but they have a prototype already for adding this in the future. It will be coming, just not right away.
    • This will be community supported.
  • CML (Cisco Modeling Labs) is the corporate version. This is licensed by the node, can scale much larger, and will have Cisco support. I have not seen pricing for this, but expect it to be priced for corporate use.
  • I have no release dates for either of these, but it does sound like they are pretty close to being able to ship it.

I’m really looking forward to this. It’s going to be a great tool for both validation and learning.

There was more to the day, but those were the bits that are probably most interesting for the moment.

FIN

Cisco Live 2014 Day 0

The Official Unofficial #CLUS Pedometer

The Official Unofficial #CLUS Pedometer

So It Begins

It’s that time of year again when I break out the Official Unofficial #CLUS Pedometer. Today wasn’t a very busy day and I didn’t actually have any technical sessions, so the count is pretty low.

Steps: 5272

KCals: 285

Miles: 2.499

Not stellar numbers, but I assure you they will be bigger tomorrow!

Today consisted of a morning of last minute studying for the CCIE R&S written, followed by failing the CCIE R&S written, and in turn followed by regret for wasting my “free” exam on a test that I was pretty sure I wasn’t going to pass. Oh well, hind sight is 20/20. If I can find some time later this week, I have a couple other exams I’m thinking of taking that I’m much more likely to pass. I don’t think I’ve left a Cisco Live without passing a test, yet.

Social Media Hub

I spent most of my afternoon catching up with old friends and meeting new ones at the Social Media Hub as people started to arrive for the conference in force. It was a good time talking about all kinds of mostly geeky stuff. Later in the afternoon was the official tweet up with probably a couple hundred people powered by caffeine and sugar. Meeting more people and doing more geek talk.

Cisco Empowered Women’s Network

After meeting with people for a bit, a group of went over to the Cisco Empowered Women’s Network session. We went over primarily to see Amy Lewis (@commsninja) speak, but ended up staying to hear Carlos Dominguez (@carlosdominguez) speak as well. Both were very good talks. Amy spoke about social and how it can be used for you and your career. Carlos’ talk was about how these tools are both disruptors and enablers and how they can be used for the greater good and to create change. Both were humorous and informative. I suspect some of the earlier talks must have been pretty good, too. If you want to check out the video, it should show up at this URL sometime in the next few weeks: CEWN-1000 – Cisco Empowered Women’s Network – Today. Tomorrow. Transformed.

On Monday, the real learning begins.

FIN

Prime Infrastructure 2.1 Upgrade

Cisco Prime Infrastructure Login Screen

Cisco Prime Infrastructure Login Screen

Cisco recently released Cisco Prime Infrastructure 2.1. Here’s a link to the release notes. I run my instance as a VMware appliance. That helps to make upgrades relatively painless and risk free because you can take a snapshot before you start. If everything goes heinously wrong, you can restore the snapshot and everything is hunky dory again. Then you can go figure out why it didn’t work without worrying about a service being down.

The Repository

With ADE-OS, you can’t just reference the URL for the download like you can with IOS. There’s no command like copy tftp://192.0.2.1/upgrade.tgz flash:. You must pre-configure a “repository” and reference that. For some reason, I almost always seem to have issues using a tftp repository. It may just be something about the way I have my machine setup, but it seems to only happen with these ADE-OS appliances. In my case, I have configured a repository for my workstation:

repository httpRepo
  url http://192.0.2.1:8000/

When I want to transfer an upgrade or patch to an ADE-OS appliance, I just fire up a python HTTP server with python -m SimpleHTTP in the directory where the upgrade file lies.

The repository will also let you configure FTP with a username and password. I use this method for my backup repository that points at an appropriate server. You configure that something like this:

repository backupRepo
  url ftp://backupserver.example.com/
  user ncs password plain <password>

The password will be stored as a hash in the config.

SimpleHTTPServer

So here’s what my Python http server looks like:

scottm@scottm-mac:~/tftp$ python -m SimpleHTTPServer
Serving HTTP on 0.0.0.0 port 8000 ...

It’s pretty exciting. When you’re ready for it to quit, hit ^C in that window. The same thing works if you’re running on Windows. Don’t forget that you may need to allow the port through your host firewall.

The Upgrade

Before you start the upgrade, verify that you are running a version that us supported. As of this writing, you can upgrade from 2.0.0.294 or 1.3.0.20 with the point patch (see the Quick Start Guide for more info). Sorry, still no path from 1.4. Here’s how you check your version:

prime/admin# show application version NCS

Cisco Prime Infrastructure
------------------------------------------
Version : 2.0.0.0.294

All good, so now we stop the NCS processes:

prime/admin# ncs stop

Stopping Network Control System...

This may take a few minutes...

Network Control System successfully shutdown.
Plug and Play Gateway is being shut down..... Please wait!!!

Stop of Plug and Play Gateway Completed!!
SAM daemon process id does not exist
DA daemon process id does not exist
DA syslog daemon process id does not exist

prime/admin#

Here’s where the real stuff starts. You took a snapshot of your VM and/or have a valid backup, right?

prime/admin# application upgrade PI-Upgrade-2.1.0.0.87.tar.gz httpRepo
Save the current ADE-OS running configuration? (yes/no) [yes] ?
Generating configuration...
Saved the ADE-OS running configuration to startup successfully

Please ensure you have a backup of the system before proceeding.
Proceed with the application upgrade ? (yes/no) [yes] ?

DO NOT press ^C while the upgrade is in progress
Aborting upgrade with a ^C may leave the system in a unrecoverable state

Initiating Application Upgrade...
Stage 1 of 7: Transferring file ...

At this point, you should see it hit your HTTP server if you used the Python SimpleHTTPServer:

scottm@scottm-mac:~/tftp$ python -m SimpleHTTPServer
Serving HTTP on 0.0.0.0 port 8000 ...
192.0.0.10 - - [05/May/2014 13:49:55] "GET /PI-Upgrade-2.1.0.0.87.tar.gz HTTP/1.1" 200 -

This is what the rest of the install looks like. This takes some time… The time stamps in the logs I am posting are from an actual upgrade, so it gives you some sense of the time involved when running the upgrade on a lightly loaded C210 M2 host.

Initiating Application Upgrade...
Stage 1 of 7: Transferring file ...
-- complete.
Stage 2 of 7: Unpacking file ...
-- complete.

*** System will reboot after a successful installation of this package ***
After reboot, please login again into the server to check status

Stage 3 of 7: Executing pre-install ...

[WARNING] System will reboot after a successful installation of this package (after Stage 7).
After reboot, please login again into the server to check status.
No action required at this time. Continuing with Stage 3.

-- complete.
Stage 4 of 7: Upgrading binaries ...
-- complete.
Prime Infrastructure Application installation completed
Stage 5 of 7: Retrieving system version ...
-- complete.
Stage 6 of 7: Updating Database Schema ...
: This could take long time based on the existing data size.

Stage 1 of 5: Pre Migration Schema Upgrade ...
-- completed at: 2014-05-05 13:58:01.026, Time Taken : 0 hr, 0 min, 19 sec
Stage 2 of 5: Schema Upgrade ...
: This could take long time based on the existing data size.
-- completed at: 2014-05-05 14:02:07.149, Time Taken : 0 hr, 4 min, 6 sec
Stage 3 of 5: Post Migration Schema Upgrade ...
-- completed at: 2014-05-05 14:03:34.82, Time Taken : 0 hr, 1 min, 27 sec
Stage 4 of 5: Enabling DB Constraints ...
-- completed at: 2014-05-05 14:07:02.317, Time Taken : 0 hr, 3 min, 25 sec
Stage 5 of 5: Finishing Up ...
-- completed at: 2014-05-05 14:07:20.315, Time Taken : 0 hr, 0 min, 17 sec
-- complete.
Stage 7 of 7: Re-enabling Database Settings ...
-- complete.
Upgrade Finished. Server is restarting . Please wait ..

% This application Install or Upgrade requires reboot, rebooting now...

Broadcast message from root (pts/0) (Mon May 5 14:10:31 2014):

The system is going down for reboot NOW!

Application upgrade successful
prime/admin#

This takes quite a while to start up after the reboot, but eventually you’ll be able to login and use the new features. Enjoy!

FIN

Nexus 5k NTP Failure

Ran into this bug today. Went to a pair of Nexus 5500s to debug a vPC link and noticed the timestamps were off. I thought that was odd. I tried show ntp peer-status and received no output in return, which I thought was even more odd. I poked at NTP for a while and decided it had to be a bug. I found the bug in the release notes and it was fixed in 5.2(1)N1(6). Here’s the Cisco bug report:

Symptom:
Nexus 5k acting as an NTP Client can’t sync with any NTP server(s).
when issuing a “show ntp peer-status” or a “show ntp peers” it does not display any of the servers/peers configured.

Conditions:
Nexus 5500/5000 running 5.2(1)N1(5).

Workaround:
Proactive workaround to prevent from this issue is none.
Reactive workaround to recover this issue is below. However, after reloading system, same issue may happen again.

#conf t
#clock protocol none
#clock protocol ntp
#copy run start

Fun. At least it can be fixed without reloading, which is a good thing in a data center switch.

FIN