text stringlengths 0 1.44k |
|---|
• Jason's startup Vex requires multi-cloud infrastructure due to scalability and reliability needs for video and audio streaming. |
• Traditional cloud providers have limitations in scale, making it difficult to handle large audiences without latency or downtime. |
• Load-balancing between cloud providers is implemented for resiliency, but a seamless failover mechanism is still being developed. |
• The startup's initial focus was on building a scalable and reliable infrastructure, rather than starting with a minimum viable product (MVP). |
• Simulating hundreds of thousands of simultaneous connections required testing frameworks and creative solutions to manage resource costs. |
• The largest test conducted to date involved 500,000 users receiving video and audio from presenters, with constant connections lasting several minutes. |
• Scaling to 500,000 simultaneous connections requires significant resources (15,000 CPUs for load testing and 1,600 CPUs for media servers) |
• Google Cloud was used as the infrastructure provider and required quota system adjustments and multiple region scaling |
• The platform is designed for interactive events, such as virtual conferences, trade shows, and online auctions, which require low latency and real-time communication |
• WebRTC is critical for achieving low latency (under 200 milliseconds) compared to HLS (5-22 seconds) |
• Bandwidth requirements are significant, with gigabits per second of traffic expected for large events, and high-resolution streams contributing to increased costs. |
• Bandwidth costs for global streaming are significantly high due to latency and distance |
• CDNs can help reduce latency, but may not be effective in all cases |
• CPU costs are relatively low compared to bandwidth costs (95-99% ratio) |
• Choosing a cloud provider is influenced by bandwidth cost considerations |
• Efficient transmission of data is crucial for large-scale streaming applications |
• Using WebRTC protocol and smart compression techniques can help reduce bandwidth usage |
• Running own bare metal hosts can provide cleaner control over the network and lower costs |
• Hybrid approach combining on-premise infrastructure with cloud providers offers flexibility and scalability |
• Data privacy considerations, such as GDPR compliance, are important for companies handling sensitive information |
• Concerns about data privacy when using online meeting platforms |
• Discussion on the challenges of developing cross-platform video conferencing solutions |
• Jason Carter's vision for Vex as a platform that provides building blocks for others to create their own applications |
• Importance of providing easy-to-use tools and components for developers to integrate into their applications |
• Overview of Vex's tech stack, including Elixir, Phoenix, LiveView, and Janus for media servers |
• Combining Elixir and Phoenix for scalability and performance |
• Evaluating Golang for high-performance tasks and potential replacement with Rust |
• WebAssembly exploration for browser-based deployment of media servers and other applications |
• Discussion of Rust's advantages over Go and Erlang, particularly in computationally-intensive tasks |
• Jason Carter's background and interests, including technology, music, and e-biking |
• Jason Carter discusses his background as a self-taught engineer and founder of Geometer, an incubator/venture studio working on WebRTC. |
• He explains why he's attracted to complicated problems and learning new things, citing his experience with streaming games and video tools. |
• The conversation turns to DevOps, with Carter recounting how he learned about Kubernetes through a project at Mavenlink, a startup where he worked before joining Geometer. |
• Carter shares his approach to learning new technologies, emphasizing the importance of finding fun in the process and being willing to take on challenges. |
• Gerhard Lazu asks Carter about Fly.io, which Carter uses in conjunction with Kubernetes for certain projects; Carter explains that they needed lower-level access to build machine images and set up firewall rules. |
• Discussion of Fly.io as a platform for deploying apps and handling regionality |
• Use cases for Fly.io in the company's infrastructure (e.g. load testing, media server workloads) |
• Plans for exploring Fly.io's Machine API for booting machines quickly |
• Comparison with container and VMs, noting that Fly.io's use of Firecracker micro VMs is "crazy quick" |
• Upcoming plans for Vex.dev, including launching private alpha and expanding the platform to handle high-scale workloads |
• Future development focus on stability, performance, monitoring, and scalability |
• Team composition, with two co-founders (Jason Carter and Sam Pearson) and a small team of engineers from Geometer |
• Plans to hire frontend developers and WebRTC experts as customers are secured |
• Potential partnership with Geometer and its consulting arm |
• Importance of attitude and approach in building a startup: curiosity, learning, teamwork, and joy |
• Launching the Vex product and gathering feedback from users |
• Adapting to changing customer needs and finding innovative solutions |
**Gerhard Lazu:** The majority of companies use a single cloud provider, and it's usually one of the big three - it's AWS, GCP, or Azure. Few genuinely need multi-cloud. Jason does, for his startup. And the more I learned about it, the more intrigued I became, like "Wow, really? Seriously?" Jason, welcome to Ship It. |
**Jason Carter:** Thanks, Gerhard. It's great to be here. |
**Gerhard Lazu:** So that's what I'm really curious about, why do you need multi-cloud? What is the story behind that? |
**Jason Carter:** Yeah. So I'll start by talking a little bit about Vex, and what it is, and how our unique requirements make it not only advantageous to be multi-cloud, but almost a requirement. So Vex provides APIs for video and audio streaming. So if you're a developer looking to build a video call into your app, or... |
Most of the time, if you go to something like Twitch or YouTube, you're using HLS to stream at scale that way, but you lose the immediacy of it; there's a ton of latency involved. And so for us, trying to build something that scaled really well without customers meant that we had to do a lot of scalability testing, and... |
**Gerhard Lazu:** Yeah. So you mentioned one thing about reliability that I wasn't expecting you to say that. So first of all, it's scale, and certain cloud providers - you cannot achieve certain scale as quickly as you may need it. And I imagine that is a limiting factor. But what about the resiliency? So what happens... |
**Jason Carter:** So when we were first talking to some initial folks that are building - you know, in the context of large virtual events, they would have hundreds of thousands of people joining live on a WebRTC connection. And they really wanted that reliability of "Hey, if a cloud provider goes down, or a region goe... |
**Gerhard Lazu:** \[06:20\] I see. |
**Jason Carter:** So we don't yet have the ability to kind of fail over gracefully if something's happening, like Google goes down - which again, very rare - but we do naively load-balance between the two when we've got our system deployed in kind of a multi-cloud mode. A lot of the times we just run on Google, because... |
**Gerhard Lazu:** So that sounds really challenging, because when I'm thinking of building a product, just like starting out, I'm thinking "Make it work. Make it nice, make it good, and then make it fast." But for you, you seem like to be starting from the "Make it fast" angle, right? Because you need that reliability,... |
**Jason Carter:** Yeah. And in fact, it was almost trickier to build that testing framework than to kind of get the initial prototype up. And it is kind of backwards. Yeah, I think the best way to describe it is that for us, scale is the core feature that we wanted to start with. A lot of other providers in the space -... |
So to test something like that, you have to do two things. You have to scale up your infrastructure to be able to handle that amount of traffic... And there's really sort of two things that you're watching out for with video and audio streaming. It's CPU load of just transcoding, or in a lot of cases forwarding the med... |
\[09:59\] So the next thing that we did is we started going lower level. What if we can just connect WebRTC's process called signaling - that is essentially how you get two peers in a call to know about each other, and establish a connection. So we wrote a much lighter-weight script in Python that could handle the sign... |
**Gerhard Lazu:** These are, by the way, simultaneous connections, and that's really important. It's not like 500,000 requests per second spread over I don't know how many seconds. This is genuinely -- and they're constant, right? the connection remains open. So these are long-running - and you'll tell me what long-run... |
**Jason Carter:** Yeah, that's correct. |
**Gerhard Lazu:** So how many CPUs are we talking about? How much bandwidth are we talking about? Can you give us some numbers? |
**Jason Carter:** Yeah, so to run tests of that size, again, 500,000 simultaneous connections, we sort of had two sets of things that we needed to scale up, both the load test system and the actual media servers. And so we ended up running somewhere in the neighborhood of 15,000 CPUs for the load test users, and about ... |
**Gerhard Lazu:** So that's more than 30,000 CPUs, right? |
**Jason Carter:** Yeah. |
**Gerhard Lazu:** Wow. 30,000 CPUs. Now, that is a very expensive load test. |
**Jason Carter:** Yes. And so you can imagine that we'd do them very quickly, right? we were only running that for 30 minutes or so. We sort of staggered the joins, so that we're not sending 500,000 connection requests at a single second, maybe over the course of several minutes. But at the end, all those connections a... |
**Gerhard Lazu:** That sounds like an awful lot of capacity. Did you have to give some notice to Google about, "Hey, we need like 30,000 CPUs?" how did that work? |
**Jason Carter:** Yeah. So Google has -- like other cloud providers, they all have a quota system and the ability to request more quota. And so we sort of had to step it up over time, because we've been working on it for quite a while; we didn't yet have a Google rep to talk to. And so we started out by taking a Google... |
\[14:05\] And so what was kind of interesting is that even though it was all the same Google account, we would get different results in different Google projects. So we might have our load testing project, and we could easily get more CPUs there. But then we'd have the other project and it'd be much more difficult. So ... |
**Gerhard Lazu:** can you imagine a conference or an event that requires 500 simultaneous connections? which event is big enough? I'm thinking KubeCon, and KubeCon is tens of thousands, maybe up to 30,000... What event requires 500,000 simultaneous connections? NFL? NBA? That's the only one I can think of. |
**Jason Carter:** Yeah. So for many large events - you can think of a trade show. We just had Dreamforce in San Francisco recently; they're broadcasting out to lots of people. The size could be 10,000 people, 100,000 people, and anywhere in between. And in that case, you're mostly just trying to present that informatio... |
Where you might want real live connections is if you have kind of a much more interactive experience, where "Hey, any one of those 500,000 people watching Dreamforce could raise their hand and say, "Hey, I'd like to ask my question on stage." So there are some folks where they really want that sense of interaction. And... |
End of preview. Expand in Data Studio
2022 Ship It Transcripts
Complete transcripts from the 2022 episodes of the Ship It podcast.
Generated from this GitHub repository.
- Downloads last month
- 19