How HAProxy Forwards 2 Million Requests Per Second - The Backend Engineering Show

Dive into a detailed exploration of HAProxy's impressive achievement of handling 2 million HTTP requests per second in this comprehensive video. Examine the benchmarking process conducted on a 64-core ARM machine, uncovering the key components and low-level factors contributing to this remarkable performance. Learn about latency and throughput differences between HAProxy versions 2.3 and 2.4, understand how TCP connections impact performance, and discover the maximum packet capacity of a 100Gbps network. Explore the distribution of workloads across 64 cores, compare tail latencies between HAProxy versions, and investigate the effects of TLS on overall performance. Gain valuable insights into high-performance load balancing and network optimization techniques through this in-depth analysis.

Intro
Summary of the Article
Latency and Throughput in HAProxy 2.3 vs 2.4
How TCP Connections Affects Performance
Maximum Packets we can get in 100Gbps Network?
How 64 Cores are divided between workloads
Tail latencies HAProxy 2.3 vs 2.4
How TLS Affects Performance?