Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn to detect and mitigate memory-related contention between containers that can increase latency by 4-13x in this conference talk from the Linux Foundation's Open Source Summit. Explore why traditional observability tools fail to identify noisy neighbor problems that cause some requests to take significantly longer than others, particularly those triggered by garbage collection. Discover real-world interference patterns from production environments at Google, Alibaba, and Meta, and understand why measuring at millisecond frequency rather than the standard few-second intervals is crucial for accurate detection. Examine the architecture of a new open source collector that combines CPU performance counters, eBPF, and high-resolution timers to identify resource contention between containers. Understand why current mitigation strategies of scaling out and running at low utilization are both expensive and ineffective at solving response time problems. Gain insights into the technical challenges of building effective noisy neighbor detection systems and learn about future development opportunities in this emerging area of container observability.
