Development PracticesLunch Talk40min
When auto-scaling isn't fast enough: Handling high volume live event traffic
This talk explores the challenges of unpredictable traffic spikes during commercial breaks in large-scale live digital broadcasts. It examines real-world traffic patterns, highlights common pitfalls, and shares engineering strategies for reliably operating advertisement delivery services at massive scale, making it relevant for engineers interested in high-scale system design.
Ivan SchwarzDoubleVerify
talkDetail.whenAndWhere
Thursday, October 9, 12:55-13:35
Room 4
In recent years, viewership is rapidly moving from linear TV (traditional broadcasting) to digital mediums (connected TVs, phones, laptops..). Popular live events such as the Olympics are watched on internet-connected devices by millions of concurrent users, and the number is constantly growing. The content stream is occasionally interrupted by a commercial break. Without strict length or timing these breaks introduce a source of unpredictable, short, massive traffic spikes for the services involved in the advertisement delivery chain.
In this talk, I’ll show you how the live event traffic patterns look like, why they are dangerous, and what you as an engineer can do about it. Based on our first-hand experience, I will shed some light on the common pitfalls and then walk you through the thinking process and techniques we have employed to successfully operate a service at this scale.
This high level technical talk is suitable for every engineer interested in high-scale environments.
In this talk, I’ll show you how the live event traffic patterns look like, why they are dangerous, and what you as an engineer can do about it. Based on our first-hand experience, I will shed some light on the common pitfalls and then walk you through the thinking process and techniques we have employed to successfully operate a service at this scale.
This high level technical talk is suitable for every engineer interested in high-scale environments.
Ivan Schwarz
Ivan Schwarz is a Director of Software Engineering at DoubleVerify. His work is all about scale.
As an engineer with a DevOps background, he thrives in low-latency, high-scale, high-availability environments. His day-to-day experience involves design and operation of services serving billions of daily requests.
As a manager, he focuses on building scalable organizations by empowering self-managed engineering teams that operate independently beyond individual tasks while staying aligned across functions.
In both professional and personal life, Ivan enjoys stories more than anything. Never hesitate to tell him yours!
As an engineer with a DevOps background, he thrives in low-latency, high-scale, high-availability environments. His day-to-day experience involves design and operation of services serving billions of daily requests.
As a manager, he focuses on building scalable organizations by empowering self-managed engineering teams that operate independently beyond individual tasks while staying aligned across functions.
In both professional and personal life, Ivan enjoys stories more than anything. Never hesitate to tell him yours!
talkDetail.transcript.loginToSeeTakeaways
talkDetail.transcript.loginToSeeInsights
comments.speakerNotEnabledComments