How Much Redundancy Does Your Live Stream Need?
Learn more about streaming redundancy at Streaming Media East.
Read the complete transcript of this clip:
Kiran Patel: There's a point of deciding what level of redundancy or resilience you want or need, and what that means for redundancy, what that means for your kind of failover logic or targets. And then the kind of return point and return time objectives: How quickly do you need to get back? How important is the channel to you? Does it have to be up 100% of the time? If there's a failure, is it okay to be off-air for a few seconds, or a few minutes, or a few hours before you get it back up and running?
Because as with a lot of things, the focus on this was looking at resilience being your top target and a lot of the time, especially for live, it tends to be right up there with your highest priority. But we're seeing other things come into that mix now. There's a lot of talk that multiple sessions over the last few days about latency being a higher and higher priority. And one of the things with latency is the lower you want to get your latency, you do have an impact on how resilient that livestream's going to be. Not necessarily in a sense where it's going to fail, but it's going to give potential users a poorer user experience because they're not going to be able to buffer content as much.
So there's a trade off when you're making your latency. There's a trade off you're making in complexity with the second example I showed you. You had a lot of moving parts. Monitoring all those moving parts gets very complex. You can build something that's much simpler, but then that will make it less resilient as well. And equally the costs, so you can have two, three, four versions of something running in the cloud as well as on premises to make it more and more resilient. But then the cost goes up as well as the complexity going up so all of these kind of factors have to be weighed up to get that kind of perfect solution or architecture for the needs that you have and then I think these are the questions that you kind of ask yourself to figure out how that's going to work.
The last one I had is just for information. I won't expect everyone to read it. It's just questions that come up a lot is that you've got a target you want to hit. Your livestream wants to be up 99% of the time, or five nines is what we hear a lot. And it's hard to get your head around what that really means and so I've just kind of done the Excel maths to figure out what that means. So if you have a target of my livestream wants to be up five nines, so 99.999% of the time, that means you're only willing to accept five minutes, 20 seconds of downtime in a year.
So you have to try and architect to get that level. And if you're going down, that's the difference is that you're designing to be willing to accept, but it's like saying that you're hitting five nines means, yeah, 25 seconds a month is the only acceptable downtime that you're kind of willing to get to. But like I said, most of the time you're trying to hit 100% so it's kind of gambling. We've got leg A and leg B. If one of them fails, the other one won't fail at the same time, or I will fix it within 25 seconds so I'll be back up to 100% before the other one has a chance to fail.
AWS Solutions Marketing Manager Kiran Patel explains the benefits of a cloud native architecture for enhancing stream redundany and resiliency in this clip from Streaming Media West 2019.
Verizon's Joe Einstein and Right Brain Media's Deke Cooper discuss redundancy best practices they've used for webcasts that are too big to fail, such as the Grammys, the Academy Awards, and the Masters.