Comcast

OpenStack. In computing, and especially distributed computing, it’s a staple, in conversation and in workflow. People tend to elevator-pitch it as “an open operating system for the cloud,” “Linux on steroids,” and “a framework based around open source software.”

As one software aficionado put it: “It’s a bunch of scripts (translation: instructions) that help create clouds and virtual machines to deploy file systems and storage and a bunch of other stuff.”

Getting clearer? Here’s more. It started in July of 2010 as a collaborative project between NASA and Rackspace, with a goal of making it easier to use regular, off-the-shelf computing hardware to handle public and private cloud activities.

Last month, Time Warner Cable posted a tech blog titled “One Year Later: Setting Up OpenStack at TWC,” penned by its lead “stacker,” Matt Haines (real title: VP, Cloud Engineering and Ops.) In it, he describes how his agile team “designed and deployed an enterprise-grade cloud,” using OpenStack, in its two national data centers.

Comcast began its OpenStack cloud work three years ago, in 2012, to support its X1 rollout — navigation first, then apps, and now video (it’s what’s behind “cloud DVR.”)

Both providers settled on OpenStack as an alternative to buying proprietary set-tops, control components, and servers from the same company. Troubleshooting gets easier, they submit. Rolling out new services, features and bug-fixes gets (way, way) faster.

It’s worth pointing out here that the long-held industrial fears about open anything are rapidly melting away. No longer are concerns about mad coders “doing harm to the network” a definitive reason to not take an open source route.

More, the tech mantra today is one of “disrupt, or be disrupted.”

The vendor community, always in a weird spot when their customers decide to lean toward “build” vs. “buy,” is following suit. Cisco, during the 2015 Consumer Electronics Show, heavily emphasized its investment in, and development of, Openstack-based components for multichannel video providers.

It follows that OpenStack is behind all the tech talk about “transparency,” and the tales about how this-or-that was about to go kaflooey, but because they had visibility into the software (which always comes in “stacks”), they fixed it (in hours, not months), averting disaster. Anecdotes like this abound in OpenStack speak.

Everything about OpenStack is open, even how papers are vetted for its annual conferences, which attract around 5,000 attendees, twice a year, for five days. (The “stackers” met in Atlanta and Paris last year.) For the Paris confab, in November, 1,100 papers were submitted for consideration (by contrast, cable’s tech events typically attract around 300 papers, vetted by committee.) The entire OpenStack community voted on who spoke.

As “open” stuff goes, OpenStack is decidedly one to know. They meet again in Vancouver, from May 18-22; on any given day, regional groups host meet-ups all over the world. Time to get your stack on.

This column originally appeared in the Platforms section of Multichannel News.

“User Defined Sports Ticker” Laura Oberholtzer Director, TV & Media Solutions Ericsson. Filmed June 10-12 2013. Video courtesy The Cable Show.

“RDK and Making Sure Your Kids Get Where They’re Going” Glee Abraham, Solutions Engineering Manager at Tata Elxsi. Filmed June 10-12, 2013. Video courtesy The Cable Show.

Hosts: Leslie Ellis, Ellis Edits, Inc. and Tony Werner, EVP & CTO, Comcast Cable. “Show Don’t Tell: Giving Customers A Taste Of Awesome” Eric Schrag Senior Android Engineer Comcast Cable “Be There from Anywhere with PanaCast” Aurangzeb Khan Co-Founder, President & CEO Altia Systems Lars Herlitz Co-Founder & CMO Altia Systems “Movie Night – A New Way to Pick a Flick, Together” Preston Smalley Product Manager & Entrepreneur Comcast Silicon Valley.
Filmed June 10-12, 2013. Video courtesy The Cable Show. 

Not long ago, I bid farewell to the flood-damaged farmhouse in Longmont, Colo. and moved on to greener, less swampy pastures. Despite the stress of moving and the fact that there are still boxes everywhere, there’s a lot to love about the new digs – a neat old Victorian surrounded by gardening space and fruit trees.

And the best part? I’m back on the cord!

One of the first orders of business at the new house, even before the moving truck pulled in the driveway, was to get Comcast service up and running. After the ultra-slow (<5 Mbps) DSL service at the farm, I was beside myself with joy when I saw this:

So how does the “cord-cutting” experience change now that I’m back on the cord?

For starters, I can watch streaming video and download software simultaneously – at the farm, this same challenge caused everything to grind to a halt for 5 or 10 minutes.

I also don’t see nearly as much buffering — there’s some, of course, but it’s generally limited to when I first start playing a piece of content. For example: Slingplayer, whether on my iPad or another device, will now keep playing without dropping the connection for hours on end (at the farm, Slingplayer would lose sight of the Slingbox at the lab at least once an hour, and every 5 minutes if I was watching something particularly interesting).

I expected to see some improvements in terms of video quality, but found it to be about the same as at the farm. Slingplayer works without interruption, but only in the SD or Auto settings – if I change the picture quality to HD, it’s full of skips and starts just like at the farm.

And the same can be said for Netflix, Amazon, and Hulu Plus, regardless of whether I’m streaming to a Roku, Apple TV, or Chromecast. I can’t say that the picture quality is noticeably sharper than it was on an ultra-slow DSL connection. What I do notice is that videos play smoothly at the new house, with virtually no “buffer breaks” (which, like commercial breaks, were a good time to grab a snack. Now I have to pause the video).

This underscores the fact that our OTT devices are really good at handling streaming video, even when the connection is less than optimal. At the farm, even at <5 Mbps, the video generally looked pretty sharp and the buffer breaks were manageable when using a streaming device connected to my TV. The main difference now that I’m on a 50 Mbps connection is that videos load much faster, and I very rarely see buffering in the middle of a piece of content.

Aside from the faster connection, the biggest difference with my new setup is that I can easily get local channels with an antenna. Finally!

You may recall that I spent hours moving a huge high-powered antenna all over the farmhouse, and tripping over coax in the hallways, only to find I STILL couldn’t get all the major over-the-air networks. When I connected the dinky little Boxee antenna to a TV at the new house, it immediately picked up ~35 channels, including ABC and NBC, two that I tried in vain to pick up at the farm. Of course, I can get those channels (and more) through my cable service, but the rarely used upstairs TV doesn’t warrant its own cable box. And now that Aereo has shut down its service in Denver for the time being, the timing couldn’t be better.

It’s good to be on the cord again. The fast Internet and cable TV feel downright luxurious after doing without for years, and I’m excited to finally be able to explore some of the other technologies that are making their way into homes. Now that we’re in the time of home automation and connected bike helmets, I’m glad to be back on the cable loop.

If you’re a marketer, you already know what drives you nuts about engineers – and vice versa.

Usually, it festers around who calls the shots on new product development.

Marketers, and especially those with packaged goods experience, want earlier involvement.

Engineers usually hear this lament many months (or years) after they began writing densely technical, often inscrutable requirements. They want informed direction, sooner.

And then the wallop of “service velocity” hit. Gone is cable’s purgatory of “one new product every 18 months,” gated by legacy back office, conditional access, or guide issues. As Cox CTO Kevin Hart put it, during a CTAM Summit session last month: “Now, we’re doing 18 products in one year.”

The tech pieces accelerating product rollouts in cable are on a roll: Open standards, the migration to all-IP (Internet Protocol), and the prying open of back office components to remove proprietary hogties.

What’s on now is the workforce and cultural changes. And this is where you run into the lingo of “waterfall” vs. “agile” operations.

Primer: “Waterfall” means serial, step-by-step processes. Write a long requirements document. Get it into silicon. Test. Get it to device manufacturers. Test. Link into provisioning and billing systems. After all that, develop training, installation, customer care and – oh yes! – marketing plans.

“Agile” means working collaboratively, across departments, and in tandem. An “agile sprint” locks a small team into writing code that puts an existing feature into a companion service. Tech people call this “experience threading.” (My favorite example, for so many reasons: Voice mail that comes over as an email transcript.)

Cox, Comcast and others are already retuning the workforce for agility. Cox’s Hart meets all day on Tuesdays with the heads of marketing, product and operations — to review readiness checklists, prioritize resources, and liaison with call centers.

What do engineers want from marketers, these days? John Schanz, EVP of Network Engineering & Technical Operations for Comcast, seeks flexibility through the innovation process. “We need a give-and-take between the business, marketing and technology teams, even when you don’t really know exactly where the destination is.”

“Forge really tight partnerships across departments – that’s where the magic is,” said CableLabs CEO Phil McKinney.

This column originally appeared in the Platforms section of Multichannel News.

by Leslie Ellis // July 09 2012

More than a decade ago, an MSO exec halted a staff meeting to make this exasperated observation: “Tools, tools, tools – can we just have one meeting where I’m not being asked for more tools? How many tools do we really need?”

At the time, Comcast was AT&T Broadband, and the tool in question related to the monitoring of an “open access” (remember that?) trial.

But the question – how many tools do we really need? – is decidedly evergreen.

The latest case in point is the home network, itself an extension of the HFC plant, with gadgets and screens that live better with signal. And they’re all cross-linked.

Today’s home networks make mixed use of MoCA (Multimedia Over Coax Alliance), Ethernet, and Wi-Fi to move stuff around. On top of that, there’s DLNA (Digital Living Network Alliance),  poised to let us share component resources – tuners, hard drives – amongst screens. And that’s just the IP (Internet Protocol) side of the equation.

Here’s how one engine-room guy put it, over a fish taco last week: “So in the home you have a QAM set-top that’s pulling video into the home network. And an advanced wireless gateway, handling data and voice. And lets throw in an IP set-top.

“The IP set-top gets video from the QAM box, but it gets its user interface through the data side.

“A customer calls: Something’s wrong with my set-top. We say, is it a video problem, or a data problem?” (At which point he made the “d’oh!” face.)

Which brings us back to tools. And silos of people — video people, data people, voice people.

One answer getting a lot of play in tech circles is TR-069, where the “TR” stands for “Technical Report.” It’s an outgrowth of what’s now called the Broadband Forum (formerly the DSL Forum; DSL is a telco thing, which might explain why cable’s coming around to it only now.)

TR-69 is sort of like an IP-based SNMP (Simple Network Management Protocol), in that it provides ways to move data back and forth, for purposes of troubleshooting, say, a home network. Or, as the Broadband Forum itself puts it: “The TR-069 standard was developed for automatic configuration of modems, routers, gateways, set-top boxes and VoIP phones.”

Great, right? Yes, if you’re ok with devils and details. While TR-69 can fetch data from different networked devices –assuming they’re plumbed with the right client profile – it lacks the job-specific tools to make diagnostic sense of that data.

What tools are needed? One for bridging into workforce management. One for customer care reps. Engineering tools, to see what’s going on. And some kind of blended video/data tool, because how things work for QAM-based video are vastly different than how they work on IP-based video.

So. How many tools? I’d go with “lots.” (And good luck with that.)

This column originally appeared in the Platforms section of Multichannel News.

In this final Downingtown segment, Charlotte takes me through the facility’s sizeable (15,000 SF) headend. Featured: Automated test equipment that pinpoints problems before technology and products are dispatched into the field. Also: A stroll through the company’s voice and video racks.

Video courtesy Multichannel News.

I traveled to Downingtown, PA (a Philadelphia suburb) in 2008 to tour Comcast’s integration center. In four parts, the tour shows what happens in Downingtown, and why constant and thorough testing is critical, prior to consumer launch. In part one, Charlotte Field, senior VP of test and operations, discusses what goes on at the 60,000 square feet test facility. Produced by the fabulous David Knappe of Glencross Films.

Video courtesy Multichannel News.

In this segment, Comcast’s Charlotte Field discusses how tru2way places the set-top inside the TV. Consumer benefits: No more box cluttering the space underneath the TV (especially important in the age of sets that hang on the wall, like a picture); one TV remote.

Video courtesy Multichannel News.