NAB Reflections: FileCatalyst CEO Chris Bailey on Olympic-Size File Transfers, Object Storage, and More
The company is pondering consumption-based billing to serve smaller clients
NAB 2017 marked the debut of FileCatalyst Central Version 3.7, based on workflows and lessons learned in FileCatalyst’s work with NBC Olympics at the Rio Games. Among the enhancements: the ability to remotely administer all remote locations (users previously had to use several tools) and an all-in-one web console with remote admin, real-time network health and statistics, and a map showing the entire file-transfer deployment. In addition, FileCatalyst is integrating with object-storage providers OpenStack Swift, Cloudian, and Caringo,
SVG sat down with FileCatalyst CEO Chris Bailey during the show to discuss the latest release of FileCatalyst Central, how the company’s role in NBC’s Olympics operation has evolved, and the growth of object storage. He also teased the upcoming cloud-based model for FileCatalyst Direct.
Tell us about the new features in version 3.7 of FileCatalyst Central.
Our job has always been to make the plumbing work as fast as possible. We make your editing team at home [at the broadcast center]an extension of your remote kits so that, when you’re in a truck, you have access to everything in the archive. We’re just continuing that mission with the latest release of Central.
For the Sochi Olympics, [NBC Olympics] wanted us to beef up FileCatalyst Central to show more reporting on the system itself. We were able to get to the point where we could report how much data was transferred from this location, on this specific day, from this time to this time. For Rio, [NBC] wanted to administer all the different locations right from the web, and they wanted even more-detailed reporting. So we took what we learned from Sochi and focused [on building] out that capability.
Ultimately, we want to disconnect from our legacy desktop applications and do everything through this nice Central application. In Rio, we did end up using our desktop application at the Olympics for a few reasons, but this release at NAB truly allows you to cut the tether from our desktop applications. We can break out tabs that are monitoring this link or that link and be able to administer all those locations and still clearly see which clips are moving from point A to point B on a given link.
This Central release is all about taking two Olympics’ worth of experience in monitoring an event of that size and making sure that we can do that with one easy-to-use application.
For Pyeongchang 2018, do you plan to add more capabilities for NBC’s FileCatalyst workflow?
Heading into the next Olympics, we’re definitely looking to add even more new features, such as the ability to look into the queue of clips to be transferred and be able to bump something [pertinent] to the front of the queue automatically and give it all the bandwidth to get it across as fast as possible. With an event of that scale, you can’t really anticipate everything that’s going to go right or go wrong, but I think each [Olympics] has gone smoother and the next one will continue that trend.
How is FileCatalyst looking to embrace the growing use of object storage in the M&E sector?
We’re doing a lot of very cool things around object storage. Archiving on tape, which has traditionally been the way archiving is done, takes up a very large footprint in your datacenter with a lot of hardware and mechanical moving parts. Space is expensive and those moving parts break all the time, so you want to remove as much of that as possible. So everybody’s talking about object store.
A few years ago, everybody was looking to move their data to the cloud. Now they’re discovering that, to retrieve that data from the cloud, they have to pay a retrieval fee: for example, 5¢ a gigabyte to pull the data out of Azure Blob or out of S3. A lot of vendors [are] popping up for on-premises and cloud hybrid storage. We are taking the opportunity to integrate with a lot of those vendors, like OpenStack Swift, Caringo, and Cloudian.
Have you completed any real-world use cases integrating FileCatalyst into object-storage workflows?
We have a project with Swift Stack and Tata Communications, Primestream and Vice Media that we worked on this past fall, where Vice Media wants to do archiving to object store from their Primestream MAM. And Tata put together the whole solution. We’re doing the transfers from New York to UK. Basically, the assets are being transferred from a Primestream MAM across a 5-Gb link to the UK and landing in object store at Tata’s links.
Part of what we’re releasing today builds on the things we did on that project to support object storage. I see the industry going that way. We’ve seen Quantum and Spectra release [cloud-based solutions] that can sit on top of your tape archive to allow [assets] to move in and out with lower latency. I missed the overtime goal of the Boston-Ottawa game [Game 6 of the Stanley Cup Playoffs series], and it took 10 or 15 minutes before a clip actually appeared somewhere on my social-media feeds. Having something like the object store and logging in real time that we had with the Olympics provides the ability to get those clips up instantly. All the technologies are moving towards getting things done faster. The things that we’re doing with object store is going to enable that.
File Catalyst has primarily been known for its use on large-scale events like the Olympics. Can it serve smaller sports events?
Absolutely. NBC doesn’t use us just for Olympics; they have Golf Channel deploying it and use it on Sunday Night Football and the NHL and most of their other sports [properties]. And we have a ton of other partners using FileCatalyst.
As we evolve [the platform], we are trying to make it more accessible with a cloud-consumption-based billing approach. That will be coming the next few months. The nice thing about that is, you have a lower entry-point cost, and, on a per-project basis, you don’t want to buy perpetual licenses. This allows someone to come in and use it for a month or use it for whatever data they need to move. We’re moving towards that, and that gives us access to some [customers] that we wouldn’t necessarily have access to because they can’t afford [the perpetual-license approach]. The big guys want to use it with a license because it’s cheaper and longer, but the smaller guys will now be able to get access to the software in more of a hybrid SaaS model.