What Drives Up Cloud Streaming Costs?
What are the predictable and unpredictable factors that drive up the cost of cloud streaming workflows as complications arise and projects scale? AWS’ John Barber, IMAX’s Abdul Rehman, Zixi’s Eric Bolten, and SVTA’s Jason Thibeault offer practical tips for keeping cloud costs in check in this clip from Streaming Media Connect 2024.
Managing unpredictable costs
Thibeault asks Barber, “Are there things that streaming operators are doing in their workflows, whether it's on purpose or inadvertently, which is causing some unpredictable costs to rise?”
Barber gives several examples of unpredictable costs, including the importance of efficient instance usage, avoiding oversized instances, and the impact of CDN misconfigurations. “I can start with the very tactical people leaving encoders on for the weekend after the Friday night fights,” he says. “Those are very difficult engagements to have with a customer. Another thing that I've seen is a customer license content, and the content provider provided a whole AVR stack, and it was very thick and expensive. They had no idea, and they were on the hook for the CDN cost.”
Emphasizing oversize issues, he says, “Getting the right size content to the right devices to the right personas, that's a magic trick. There's that tripod there that you've got content, and you may be doing manifest manipulation or whatever else and try and tune down the bits that are going to the cell phones, where you're really not getting that much value off of setting that much down the pipe and just kind of massaging that. And there are a lot of tools coming out that are helpful in that.” Addressing CDN misconfigurations, he says, “CDN misconfigurations can cause a lot of pain, and because of the way most of them, you might not know until the end of the month, and that can be a very difficult conversation.”
Right-sizing instances to optimize compute resources
Thibeault asks Rehman of IMAX to discuss his experiences with cloud streaming cost issues.
Rehman highlights right-sizing instances to optimize compute resources. “I think using instances efficiently to get the best bang for the buck is important,” he says. “It could be that a smaller instance would provide you a better turn on investment for the compute that you want to do. So, for example, even if you're doing very compute-intensive tasks like encoding, you want to keep an eye on the workflow that you built. Let's say an encoder and the CPU are used at 10%, and you're assuming that you really need the 18 X-large instance, too, because it's encoding, and you really need that. Whereas, if you look at the usage spec, it may be underutilized significantly. So designing that and making sure you’re [using] your infrastructure to the maximum extent and doing some sort of right sizing and fitting it for the workflow. I think that goes a long way in making sure that you are optimizing it.”
The significance of protocol efficiency and managing egress charges
Bolten of Zixi focuses on the significance of protocol efficiency and managing egress charges. “Zixi is a software-defined,” he says. “We focus very much on efficiency and throughput and different protocols matter. I happen to work for a company that has a protocol, but within the efficiencies we see there, we can be two and a half times more efficient in the payload than at SRT, for example. That matters for what machines you have to keep. You need to find out how you can do more throughput per core. We're getting up to five gigs, and you work on technologies like DPDK, where we can get ten times the efficiency.
“But ultimately, the real meter that's going to be running is on the egress charges,” he says. “And that's going to come down to how you manage the transport layer and how you architect within cloud hybrid [and] multi-cloud type options of customers trying to figure out the optimal way to deliver the signal.”
Join us in November 2024 for more thought leadership, actionable insights, and lively debate at Streaming Media Connect.
Related Articles
What are the best strategies streamers can follow to manage the costs of cloud streaming architecture and production and maximize efficiencies in their workflows? AWS' John Barber, IMAX's Abdul Rehman, Zixi's Eric Bolten, and SVTA's Jason Thibeault discuss cloud streaming cost management in this clip from Streaming Media Connect 2024.
26 Sep 2024
Cloud streaming workflows and infrastructure seemed like the wave of the future 5 years ago, due to the rapid growth of cloud technology, their evident advantages in scalability, and economies of scale, and accelerated adoption during the height of the pandemic. Today, streaming's cloudward progression seems like all but a foregone conclusion-though opinions vary on whether that hypothesis holds for certain live and VOD workflows, as well as productions with essential on-prem components-as this discussion involving key cloud practitioners from Endeavor Streaming, Grabyo, TMT Insights, and LiveX at Streaming Media NYC reveals.
27 Jun 2024
Streaming Media's Tim Siglin interviews United Cloud's Boban Kasalovic at Streaming Media NYC. Kasalovic discusses United Cloud's development of an end-to-end streaming solution and an anti-piracy platform called Gladiator. The platform uses machine learning and AI to analyze data and identify critical behavior that could lead to piracy. The company has also implemented Widevine CAS for its hybrid solution.
19 Jun 2024
Streaming Media's Tim Siglin interviews Grabyo's Clare Butler at Streaming Media NYC 2024. She highlights that Grabyo is not geographically bound and caters to anyone working in the live space, including news broadcasters, sports clubs, and entertainment producers. She also emphasizes the cost-effectiveness of cloud solutions, noting that cloud technology is now at parity with on-prem solutions, offering features like instant replay and SCTE-35 ad insertions.
05 Jun 2024