Tag Archives: Mobility Field Day 4

The Network is Code: Cisco at MFD4

3 Replies

It’s always a bit of a thrill to visit Cisco HQ, and to step within the walls of this global network powerhouse. I got to do that again at Mobility Field Day 4, and as usual the presentations and the visit just went too fast. Such is the way these events go… On this go round, Cisco offered us:

Each is interesting and informative, especially when combined with the delagates questions. You’ll be glad you watched them, if you haven’t yet.

But something else jumped out at me at this event, and it may seem silly to even mention. Have a look at this sticker:
Code Pic

The wording of it got my mind working. In a number of directions.

I’m just sharing what’s in my head as a long-time Cisco wireless customer as I ponder the message on that innocous sticker.

I’m glad to see that CODE is the network, because it hasn’t always been. CODE, as presented like this, implies “reliable code, as surely you don’t want an unreliable network”. To that I would add “especially at the costs charged for licensing the hell out of everything”.  The sticker mentions CODE + the 9000 Catalyst Series, and perhaps sends the message that it’s a new day for reliability? On that topic, the CODE in this case is IOS-XE, which displaces AireOS as what powers the Cisco line of wireless controllers. I do hear often that “IOS-XE has been out a long time so it has to be solid by now” kinda talk.

I’m not sure I buy into that, but am hopeful. If I’m a little skeptical, it’s because IOS-XE packaged as a wireless controller brain is a new paradigm, despite the maturity of the OS. And… despite many, many mea culpa  sessions in private with Cisco’s wireless business unit through the years over wireless code quality, I have yet to see any sort of public-facing commitment to not repeat the development sins of the past as the new magic seeks to gain traction. This bothers me, in that I don’t know that the background culture that allowed so many problems with the old stuff isn’t being carried over into the new. My problem, I know. But I’m guessing I’m not alone with this feeling.

The other thing thing that this sticker has me thinking about is this: if  the network is code, why do I need controller hardware? Yes, I know that the 9800 WLC can run in VM- but VM instances ultimately run on hardware. As a big Cisco customer with thousands of 802.11ac access points that run the latest AP operating system, I would love to be totally out of the controller business (and all the various management servers needed) WHILE KEEPING MY INSTALLED ACCESS POINTS. If the network is code, maybe let me point these things at my Meraki cloud and simplify life?

I’m just one man, with opinions. But that sticker did get me thinking…

 

Code, Heal Thyself: Mist Systems Brings Something Badly Needed to WLAN Market

2 Replies

If you do any profession long enough, you’ll experience all sorts off good and bad along the way. For me, “good” has been the honor of providing reliable Wi-Fi to hundreds of thousands of client devices through the years, and “bad” has been fending off downtime and damage to organizational reputation when code bugs hit. Why focus on code bugs? To me, they are the one huge factor in WLAN system operation that we as wireless professionals can’t control. We can get everything else right from RF environmental design to RADIUS server capacity to onboarding clients, but we can’t defend against what evil lurks in the lines of code that runs the system hardware. Nor should we have to- that’s where we expect vendors to hold up their end of the deal on hardware and software that ain’t getting any cheaper.

Oh, how I have bitched and whined and complained about code bugs through the years. There was “The Horrible Bags We Hold For WLAN Vendors“. And “Code Suck Regulation: Should We Sue Vendors For Major Code Bugs?” I got a bunch of them… and it’s not just me. One of my favorite people, Jake Snyder, laid down a really good video lament on the topic. No one can forget my own video from the Wireless LAN Professional Conference in 2017 where I detailed real-world impact of code bugs. It’s a real thing, ya’ll.

I titled one post on the topic “Will Reliability Be Prioritized Before Wi-Fi’s Whiz-bang Future Gets Here?” (a house built on suck cannot stand).  This one jumped to mind yesterday as I sat in a Juniper Networks conference room in San Jose and heard Mist Systems talk about reliability. What I heard was refreshing.

Mist CTO Bob Friday and his crew presenting at Mobility Field Day 4 detailed how the company’s AI does all kinds of things- but among the most important is finding it’s own system anomalies. The gravity of the point is fairly significant, as one vendor after another wants to put a dashboard in front of you that calls out anything and everything as a wireless problem for you to chase after, but none that I know of will raise their hand and admit “OK- I’m actually the problem here… me, the system. I screwed up… I’ll fix me so we can all move on. Beg your pardon…” But now Mist is promising that, and it’s huge.

CTO Friday not only called out this capability, but was kind enough to give me a shout out for my years of crying like a school girl about code bugs, which was thoughtful.

IMG_3558.jpg

Well done, Mist Systems! There was a hell of a lot more to the presentation- and in the couple of hours I listened, I was impressed that Mist has managed to boil the hype off the concept of AI and actually did a decent job of explaining real-world, practical applications and benefits. There are several videos from the session, and they are worth watching.

More about Mobility Field Day 4 here.