December 10, 2018
By Tim Siglin Contributing Editor, Steve Nathans-Kelly
Featured Articles

SMP '18: Ellation's Michael Dale Talks Multichannel OTT

Learn more about AVOD and OTT at Streaming Media's next event.

Read the complete transcript of this interview:

Tim Siglin: Welcome back to Streaming Media West 2018, I'm Tim Siglin, contributing editor with Streaming Media Magazine and the founding executive director of the not-for-profit Help Me! Stream. I’m here with Michael Dale from Ellation. Michael, first of all, tell us what your title and role is with the company, and then where did the name come from.

Michael Dale: Sure. I'm VP Engineering. I work very closely with product and business groups around basically a few different areas of subscription and AVOD-based business. As for Ellation, the name of the company ... We originally were Crunchyroll, as you may have heard of. That's been around for a little bit over 10 years. When Crunchyroll took some funding from a joint venture of Peter Chernin, and AT&T, they retitled it as Ellation. It had to do with an iceberg ... some part of the ice that comes up from the water and the “elation” is the larger iceberg underneath or something. And then we have the pun of “elated.”

Tim Siglin: Right, exactly. As you say, Crunchyroll's been around for about 10 years. What's new now with Ellation to change the focus? Or is the focus the same?

Michael Dale: With Ellation and the AT&T Warner Media acquisition of Otter Media, which is the parent group that owns Ellation, I think we've really seen Crunchyroll scaling up quite a bit. Likewise, the VRV product, the multichannel offering, is also starting to have significant impact on our overall subscriber count. We recently crossed 2 million paid subscribers, putting us well within the top 10, or at least what used to be the top 10. In any case, as a pure OTT play, we're certainly in the top 10 of online subscribers.

Tim Siglin: For you, being in that top 10 grouping, especially with your engineering focus, what are the challenges that you face? Because clearly when you get into that level of the stratosphere or exosphere in terms of the number of subscribers you're delivering to, everything has to work perfectly.

Michael Dale: We're dealing with more than a billion minutes per month streamed. Not only the 2 million subscribers, but also the 10-15 million monthly actives that are on the AVOD side of the business.

It is a fairly massive scale project. We've transitioned a lot of the engineering concerns, the sort of traditional concerns around taking the data center-based monolith and transitioning it to AWS and microservices and those sort of things that we've seen a few talks about here at the show today.

That's certainly been a challenge relative to the just sort of changing the engine as the planes in flight as they say. With the fact that we had to continue to serve our users through this whole duration.

Tim Siglin: When you're talking microservices, I know there's also sort of a debate in the industry too if you're taking the monolith and just shoving it in a kubernetes container, that's not a microservice. How do you define microservices? I think that's a hot topic these days.

Michael Dale: Microservices are more than just the logical code service grouping; it's also about autonomy of the team, so that they can deliver value without being blocked on the rest of the organization. In the context of the lift-and-ship approach that we took with the monolith, just to get it out of the data center.

We don't call that side of our text stack microservice space. We have been building out microservices from the ground up, relative to the nerve product launch, and now we're sort of in a process of consolidating against a multi-tenant-based microservice architecture.

Tim Siglin: In terms, of as you talk about not being sort of slowed down by the other groups, does that help you roll out enhancements and futures faster than any particular?

Michael Dale: You can see things in the old model. In the monolith, there was not ideal an architecture that was in play. I'll give you an example. Basically, it was hitting a database to look up the segment names for the HLS stream. That was pretty much a recipe for disaster as the database would underload as more people were just trying to watch video. Simple things like that, where that's all now decoupled. When a lot of people watch video, you don't take down your login service, so on and so forth.

Tim Siglin: That's actually a really interesting point, a single point of failure if the database is trying to handle both transactions, authentications, and as you say segments for-

Michael Dale: Right. I'm sure you've chatted with other companies that scale up, and they're dealing with Game of Thrones, or some massive cultural event. The video usually works, cause the CDN's are off loading that. Then something will be the playhead position service will take down your off service if you don't have it segmented. An ability to scale it appropriately relative to the surge in demand.

Tim Siglin: It is. A matter of fact, I've written articles in the past about for live events, say a big boxing match, it's never the video deliver, it's always the authentication on the front end. If that fails, the customer can't watch the video, and the same net effect is they want their money back for that.

Michael Dale: Same thing for the customer. Right.

Tim Siglin: Awesome. Michael, thank you for your time. Really appreciate it.