Cloud Streaming Workflows for Encoding and Disaster Recovery
Why create a hybrid event stream? Nadine Krefetz, Consultant, Reality Software, Contributing Editor, Streaming Media, asks three leading industry figures about the best use cases for making cloud-based hybrid events versus events that are fully on-prem.
“What are some of the specific things that are done in a hybrid environment?” Krefetz says. “What kind of [media-related] use cases do you we have here?”
Loke Dupont, Solution Architect, TV2, says that encoding is a good media-related use case for hybrid events. “A lot of companies will encode video that are just regular run of the mill, and they'll budget for that and have on-premise equipment for that,” he says. “And then suddenly you might license a new content catalog, and you need to ingest that within a certain time period. In that case, you can run it through your normal process, but it might take a long while because you don't necessarily have spare capacity. It's a relatively easy workload to burst into a cloud environment and run it there and then return it back to on premise. It's also one of the solutions that I think is very well supported by a lot of the encoding vendors…they can send you a physical box, you can plug in and rack somewhere and also the software will handle that automatically and move some of that workload of transcoding video into the cloud. Because it's such a good scenario for something that you do in a batch way, you have all this stuff that you run through and then once you're done with it, you can forget about it, just tear down all the cloud infrastructure and everything is fine. I think that's definitely one of the classic use cases for that.”
Andy Beach, CTO, Media & Entertainment, Worldwide, Microsoft, says that disaster recovery is a common use case regarding workflow management systems for the cloud. “So if you're running 10 to 15 channels of content regularly, and they run 24/7, stand them up, run them in an on-prem environment,” he says. “But that becomes a single point of failure for your content going out. So you would typically set up some sort of secondary site and source, and it's very easy to make that the cloud so that if the signal goes down on the first facility, you get an outage or there's a disaster in that area, an earthquake or something else, then the systems switch over automatically to that secondary space if it's in a cloud. Infrastructure is only spun up for the time period that it's needed until you do the recovery and revert back to your primary sources.”
John Jacobs, Field CISO, Fortinet, remarks, “That is probably the most digestible use case right there. We've seen so many people that just say, ‘We're not going to spend this for a disaster recovery scenario,’ that it's perfect. It's there, it's available in cloud…from the first instantiation, it's on when you want it and then just off.”
Krefetz says, “So for disaster recovery, is it more common to go multi-cloud or single cloud?”
Beach says, “It’s a great question, and it really depends on the comfort level of the engineering group or the ops group. Often people will start with one, and then there will be an outage in that cloud because even clouds have outages at times. And so they will make backups for the backups. And so it becomes a scenario where you likely start with an individual one, but over time you have as many backups as possible. There's a reason I have four keys to my house strategically placed where I know where to find them. I like to have backups!”
Learn more about encoding, hybrid, and cloud streaming workflows at Streaming Media East 2023.