You are viewing content from a past/completed QCon
Session

Improving Video Encoding System Efficiency @Netflix

What You'll Learn

1Find out what the video encoding system at Netflix looks like.

2Learn the tools and techniques to analyze performance and to improve the system efficiency.


Every show on Netflix is encoded for the device type and bit rate that is optimal for the customers. For any one show that can be tens of different encodes, generated by a large-scale system that scales up to hundreds of thousands of CPU cores concurrently. As the performance team at Netflix, we are motivated to tune the system to generate encodes quickly and efficiently. 

In this talk, we explore the challenges of improving the efficiency of video encoding systems at Netflix scale, and cover actions we take to achieve it by leveraging high-resolution telemetry tools, hardware features, and resource-aware task scheduling.

What is that work you're doing today?

I'm from the Netflix performance engineering team. Our team is responsible for bringing higher efficiency to the Netflix streaming and studio environments, identifying bottlenecks and optimizing systems with creative tooling and analysis. I have been focusing on making video encoding systems fast and efficient, also spending time on building self-service performance analysis tools based on bpftrace.

What is the goal for your talk?

I would like to introduce the video encoding systems at Netflix and some great tools and techniques that we applied to deliver a faster encode and to run them on resources more efficiently.

What do you want people to leave your talk with?

I want my audience to leave with three techniques of improving compute intensive workload performance:  use high resolution tools to know where and how to optimize, leverage hardware features to go full speed and think of how to fit workloads together to maximize throughput. I’d also like to encourage the audience to check out some amazing open-source tools that help in this domain, such as FlameScope, FlameGraph and Vector etc.


SPEAKER

Susie Xia

Senior Performance Engineer @Netflix

Susie works on the Performance Engineering team at Netflix. The team is responsible for bringing higher efficiency to the Netflix streaming and studio environments, identifying bottlenecks and optimizing systems with creative tooling and analysis. On the team, she has been focusing on making video encoding systems fast and efficient.

Prior to Netflix, she worked at LinkedIn and Salesforce, where she helped improve service performance end-to-end and build automations for root-cause analysis and capacity analysis.

DATE

Wednesday Nov 18 / 09:00AM PST (40 minutes )

TRACK Mechanical Sympathy: The Software/Hardware Divide ADD TO CALENDAR Add to calendar
SHARE

From the same track

View full Schedule

3 weeks of live software engineering content designed around your schedule.

Don’t miss out! Save your seat now

Register
TOP