Damn You, CAPWAP Tunnels… Damn You All to Hell

There comes a time in every person’s life when they have to face the truth: maybe their CAPWAP tunnels that have been so good for so long actually have a dark side… Maybe them tunnels make you feel empowered, nay- maybe they make you feel invincible when it comes to creatively using VLANs in your overall Wireless LAN construct… and maybe someday that good thing leaves you in a bad place. Maybe.

Let’s pause for some lyrics from the immortal Waylon Jennings’ song “Wrong”:

I should have known it all along
When the future looks too bright can’t be anything but right
Wrong

Everything was going strong
The sky was always blue I thought my dreams had all come true
Wrong
Wrong

Let’s get right to it: CAPWAP TUNNELS SPOIL YOU.

You’ve been using a WLAN solution for a lot of years. It’s been buggy at times, the vendor has left you frustrated on countless levels. You’re thinking “shit I would freakin love to finally ditch controllers and that bloated, semi-functional NMS and move to a cloud WLAN solution for my thousands of wireless access points” – WAPs for some of you (shut it- you know who you are)… But then you run into the CAPWAP tunnel thing and a big honkin Layer 2 quandary down in your switches.

If I have a controller-based WLAN, I can get away with this at the AP uplink port, which clearly gets the Polly Pony Seal of Approval:

But alas, take away the CAPWAP tunnel construct and you are left with something less savory, and Cactus Mike isn’t digging it:

I gotta agree with Cactus Mike- in very large WLAN environments, the thought of no CAPWAP tunnels sucks ass. Sure, maybe a radical redesign of the LAN that underpins the WLAN would help, by pushing L3 out closer to the edge and reducing the need for VLANs. But such undertakings aren’t always a possibility, and if they are a possibility, the timing of redesign opportunities may not line up. Back to topic.

Am I suggesting that by going to a cloud-managed WLAN solution that CAPWAP tunnels aren’t possible? Yes and no… Some cloud vendors recognize Cactus Mike’s conclusion, others not so much. I have not actually used any of the following solutions, but I do appreciate that they recognize that “switching to cloud” and “ditching the controller” isn’t all that easy for those of us with CTA (CAPWAP Tunnel Addiction):

–Arista:

–Aruba: (link is here)

–Extreme definitely has an answer but I’m not finding the right link. Will edit

–Mist: (link is here)

–Ruckus: (link is here)

By no means is this summary meant to be comprehensive. And, if you were to drill in to any of these, I’m not sure they would each stand up as an answer to “how do we ditch our current controllers, terminate VLANs somewhere, yet move the rest of the show out to the cloud while retaining our CAPWAP tunnels and not doing a massive L2 reconfiguration?” as I have not tested any of them.

But- I do appreciate that the situation is being recognized and addressed by major vendors. AND- I am surprised that at least one long-running pure cloud innovating powerhouse vendor has yet to provide an answer to the situation. As long as the only answer is to configure the uplink to a cloud-managed AP as if it was an old fat legacy access point, they won’t be getting an invite to Cactus Mike’s summer bash…

Your thoughts on the topic?

Mojo (Arista) Answers The Layer 2 Situation for WLAN Migration To Cloud

I recently wrote about the challenges, as I see them, with the Layer 2 aspects of moving from an established controller-based WLAN solution to one like Aerohive, Meraki, Mist, or Ubiquiti that is managed in the cloud. That article is here, at IT Toolbox.

Want the short version of The Layer 2 Situation? Being all about value, I can help you out… Let’s start with the simple view of VLANs that underpin a controller-based WLAN environment:

Betwixt the switch and the AP you have a single VLAN. It’s simple, it’s clean. It’s not a spanning tree asspain. But cut into that single VLAN with your magic network knife, and you’ll find a CAPWAP tunnel with as many VLANs as you need. In large environments, that may be dozens o’ VLANs for various SSIDs scattered across thousands of APs.

Contrast that with the typical fat AP/cloud AP VLAN underlay:

Ugh- see the difference? In those large WLAN environments- where thousands of APs equals hundreds of switches- you might have to configure thousands and thousands of switch interfaces to convert the simple CAPWAP-oriented LAN to the VLAN-heavy LAN needed by fatty-fat APs- AND most cloud APs.

Ugh.

Mojo evidently agrees with that ugh and offers an option that preserves the goodness of the cloud approach (No NMS to keep up, easier code upgrades, no buggy controllers to babysit, etc) while providing an easy way to NOT go down VLAN rabbit holes when converting from controller to cloud. This magical hybrid approach features the Multiservice Platform:

Tres sexy, no? I had heard about Mojo’s Multiservice Platform last year at Mobility Field Day 2, but will admit I lost some of the messaging in the din of all the “Cognitive blah blah blah”. But when I recently wrote about The Layer 2 Situation, two good citizens from WLAN land came forward and reminded me that this nut has indeed been cracked, and by Mojo.

Recall if you will- Mojo has been acquired by Arista Networks since Mobility Field Day 2. I also happened to be present at the Mojorista MFD3 presentation, which I wrote about here.

So… will Arista continue with the Multiservice Platform? I have to say that I really hope so. I hope they promote the heck out of it, and that other cloud Wi-Fi vendors follow suite. I don’t know whether I’ll ever run a massive cloud AP WLAN (I do currently run a massive controller-based Wi-Fi network and a lot of cloud-based branches), but if I do it’s nice to know that there is at least hope for The Layer 2 Situation.

Tanaza and Cucumber Tony- Software Makes Cheap Wi-Fi Cloud Manageable

Quick and dirty- I have an article running at Network Computing about Tanaza and Cucumber Tony, two companies that promise to breath cloud-managed functionality into cheap Wi-Fi gear. Their web sites:

• Tanaza
• Cucumber Tony

I’d love to hear opinions on the paradigm, and whether you’d ever give this sort of software/management/hotspot capability a try.

Thanks for reading!

A Six-Pack Of WLAN Industry Developments

Things are always shaking in Wi-Fi Land. New stuff, company goings on, regulatory drama… it’s never boring. Here’s a quick bundle of interesting hits to consider.

1. Meraki Founders Quit Cisco. I’m not only a Meraki user, I’ve been following the company for years under the brim of my analyst’s hat. I delighted when Meraki came out with their MX line, and later when switches joined the lineup. There’s a lot of power in the Meraki magic, so I can’t say I was totally surprised when Cisco bought them for north of a billion dollars. At the same time, I had my concerns. Far be it for anyone not in the loop to speculate on why Meraki’s Founding Three have opted to split, but it does fuel all sorts of speculation depending on your frame of reference.
   

   ---

2. Xirrus Has Announced a Cloud-Managed 11ac Wallplate AP. This is an industry first (as far as I know) and I hope other vendors follow soon (are you listening, Meraki?)
   

   ---

3. Meru also has new product offering: Xpress Cloud. With 2×2 11ac APs managed via cloud subscription, aimed at SMBs. (Meru ain’t dead, folks.)
   

   ---

4. Fluke Networks’ Air Magnet Enterprise gets an upgrade.  Quoting my brief: “The new version of AirMagnet Enterprise includes several major security enhancements, new 802.11ac functionality, the industry¹s first “No Wireless or Cellular Zone” capability, new PCI 3.0 compliance features,  and more. Enterprise is already unique with its Automated Health Check and Dynamic Threat Update capabilities, but these new features make it even more powerful, and a crucial solution for organizations that can¹t afford to have wireless security loopholes.” Alas- it’s still an overlay…
   

   ---

5. Ruckus Ups Their Smart Wi-Fi Game. A laundry list of beefy feature goodness is aimed at improved Wi-FI calling, among other enhancements.
   

   ---

6. Eero. Interesting promise and premise. We’ll have to see how this one plays out- but promising people that you can solve dead spots in the home without running wires will get attention.
   

   ---

I don’t typically favor scraping press releases into a digest blog, but this mix of topics struck me as a bit profound in showing just how dynamic the Wi-Fi world is at many tiers. Exciting, thought-provoking stuff that can be hard to keep up on.  Don’t blink, things change quick around here!

 

Getting to Know the Very Cool Open-Mesh Wi-Fi System

As I root around at the more inexpensive end of the Suitable-For-Business WLAN space, one product set keeps jumping out at me. Open-Mesh is a big story in a little package, with a pretty crazy feature:cost ratio (as in crazy impressive). It’s just a really neat, innovative framework that offers cloud-managed APs for under $100, with a cloud dashboard that couldn’t be easier to use.

To boot, there is a free Android and iOS app for CloudTrax, and it looks real nice.

This Ain’t Market-Leading Wi-Fi, Nor Does It Try It To Be

One trap that many wireless professionals fall into (in my opinion) is not being agile enough of mind to set aside their loyalties to top-end product sets and simply appreciate what’s out there in the interesting edges of the wireless market. Sure, many of us have million-dollar WLAN environments and appreciate what we get in exchange for Large Costs, but there is life beyond Cisco and Aruba just like there are cars beyond Cadillac and Lexus. If you can open your mind and get over yourself, Open-Mesh fills a cool, low-cost niche for clients that GOTTA have Wi-Fi but DON’T have the dollars or know-how to pull off a pricier install.

Open-Mesh doesn’t really advertise, it sells itself word-of-mouth by satisfied users. It’s roots are largely the same as the early days of Meraki, where “roofnet” low-cost nodes were meant to provide connectivity to the underserved. If you are familiar with Meraki, and look in on Open-Mesh’s CloudTrax dashboard, you’ll see a lot of similarities.

It’s almost like “Meraki in Miniature”! The cloud management account in it’s current form is 100% free, there are NO licensing or account fees, and an overview of features is here. Though I’m not at liberty to say what other features and options are coming, I can say that Open-Mesh has some very cool evolutions on the short horizon.

What About the APs?

There are a number of interesting aspects about the Open-Mesh AP product set. (Again, these are not meant to compete with product like Cisco 3700s, so fight the natural urge to compare and trash the “lesser” product.) The APs are modular in that just a few radios are swappable into different enclosures, letting you “build” the APs that you need. There is no labeling on the APs- if you are in the WI-Fi networking business, your own logo can go on the APs (and in the cloud dashboard, for that matter). And for sparing, you don’t even need an enclosure.

The Open-Mesh APs are detailed here, and make sure you click “Show More Specs” for the full picture. Though you won’t see any dual-band or 11ac APs in the line-up now, take another look at the prices. You’re still getting pretty decent value, and you can expect more impressive hardware spec’d APs to come along soon from Open-Mesh.

So… Who Uses Open-Mesh?

If you get interested enough to learn more, Open-Mesh does have business reference accounts happy to talk about about their success with this unique system. Aimed mostly (but certainly not limited to) housing/hospitality/SMB customers, Open-Mesh has single sites with just a couple of APs all the way to  sites with APs measured in the hundreds.

I personally am working on a potential public WLAN project for my own very small village, and Open-Mesh is at the top of my “to consider” list given the available features, low cost, and decent reputation of the solution. More to follow if I end up pulling it off…

There are a lot- LIKE A WHOLE LOT- of low-end WLAN solutions out there. I’ll be writing up this market niche for Network Computing soon, and will be talking about pros and cons of not spending the big bucks when it comes to WLANs. Meanwhile, take a look at Open-Mesh and see if you don’t find it as intriguing as I do.

Wouldn’t It Be Nice If ALL WLANs Could Move to the Cloud?

Riddle me this, my Wi-FI homeys: What’s missing from today’s cloud-based WLAN paradigm?

I actually (kinda) like my controllers these days. And I like that a single VLAN is all I need to each AP in my CAPWAP world. But, I still yearn for cloud control over the whole thing (sorry, PI- and every other bloated management framework). Here’s how I see it, at my Network Computing blog.

And, as a thank-you for stopping by, here’s a picture of Blue Mountain Lake. We’re all about value here at wirednot.