Video: The ROI of Redundancy in Large-Scale Live Streams
How much redundancy is enough for mission-critical, mass audience live streams like the Grammys, the Oscars, or the Masters, where snafus happen but webcasts can't fail? And is there such a thing as too much redundancy, as a percentage of the overall investment in the broadcast? Verizon/Go90's Joe Einstein and Right Brain Media's Deke Hooper recount their experiences and best redundancy practices in this excerpt from their panel at Live Streaming Summit 2015.
Learn more about live content flow at Streaming Media West.
Read the transcript of this clip:
Tim Siglin: Is there a rule of thumb in terms of redundancy that if the budget is actually spent 3% on redundancy, 5% on redundancy, or is there not really a model?
Joe Einstein: In my previous life, doing things like Grammys and Academy Awards, where it really can't fail because you're sort of mirroring a broadcast, we would typically present ... You can be diverse out of the venue multiple ways and you can even have diverse encode sites and then multiples on the end, so I don't know that there was necessarily 'It's x-percent more.' It really came down to how much do you want to spend?
The Oscars is one where we sort of did everything that you could possibly do and there were backups to the backups to the backups. The hard part is at what point are you building in so much redundancy that when something does go wrong can you actually recover from something spiraling out of control you didn't think about? I've seen that happen. It's almost having to pull the ripcord to get out of all your recovery scenarios.
Deke Hooper: What it comes down to is the value of the event. It is a free event and it's paid for by ads. You need views. If it's a pay-per-view event, people might be willing to wait. Redundancy might not have a big play. Really it's going to be gauged by the content, what you're putting out there, that's going to control the redundancy level that you're going to put in place.
Some of the larger-scale clients we work with will have three backups. They're paying for the encoders, the equipment, connectivity and multiple CDMs that they might not even use. They just keep warmed up and ready.
Tim Siglin: What kind of client would that be?
Deke Hooper: The Masters event. We've done that the last few years. Every year it's got its new challenges. There's always something that could go wrong, even with as much of an investment that's there. It's live. Anything that could go wrong will go wrong.
Speaker 4: Is there a rule of thumb in terms of redundancy that if the budget is actually spent 3% on redundancy, 5% on redundancy, or is there not really a model?
Speaker 5: In my previous life, coming out of the Service [inaudible 00:00:22] model, things like Grammys and Academy Awards, where it really can't fail because you're sort of mirroring a broadcast, we would typically present ... You can be diverse out of the venue multiple ways and you can even have diverse [in-code 00:00:36] sites and then multiples on the end, so I don't know that there was necessarily 'It's x-percent more.' It really came down to how much do you want to spend? How redundant-
Speaker 6: We're thinking it can't possibly fail three times.
Speaker 5: Right. Oscars is one where we sort of did everything that you could possibly do and there were backups to the backups to the backups. The hard part is at what point are you building in so much redundancy that when something does go wrong can you actually recover from something spiraling out of control you didn't think about? I've seen that happen. It's almost having to pull the ripcord to get out of all your recovery scenarios.
Speaker 6: What it comes down to is the value of the event. It is a free event and it's paid for by ads. You need views. If it's a pay-per-view event, people might be willing to wait. Redundancy might not have a big play. Really it's going to be gauged by the content, what you're putting out there, that's going to control the redundancy level that you're going to put in place.
Some of the larger scale clients we work with will have three backups. They're paying for the encoders, the equipment, connectivity and multiple CDMs that they might not even use. They just keep warmed up and ready.
Speaker 4: What kind of client would that be?
Speaker 6: The [ID Masters 00:01:50] event. We've done that the last few years. Every year it's got its new challenges. There's always something that could go wrong, even with as much of an investment that's there. It's live. Anything that could go wrong will go wrong.
Related Articles
AWS Solutions Marketing Manager Kiran Patel walks streamers through the logic of how much redundancy to build into their live streams in this clip from Streaming Media West 2019.
30 Mar 2020
Accidentally deleted a crucial video file? The good news is it might still be recoverable. Here are steps to take for recovery and to prevent loss from happening in the future.
15 Jun 2018
Redundancy and advanced testing are essential for any live video event. Here are five rules for live streaming that can help save the day.
17 May 2018
Peter Hartz explains how The Switch leveraged fiber technology to cut, switch, and deliver Univision's live soccer broadcasts from their home office, and how the infrastructure can be used for a range of other purposes including live OTT in this excerpt from his Live Streaming Summit panel.
03 Oct 2016
Experts from Akamai, Comcast, Conviva, and Yahoo! discuss key analytics, metrics, and assessment strategies that will help you deliver more effective and profitable live streams.
28 Sep 2016
The Switch's Peter Hartz and Live X's Corey Behnke make the case at Live Streaming Summit for fiber as a reliable, agile, and inexpensive alternative to traditional satellite trucks for large-scale broadcasts.
14 Sep 2016
DaCast VP Greg Ellis reveals his best practices for turning live event streams into profitable undertakings.
12 Sep 2016