Help Rather Than Recycle: Alleviating Cold Startup in Serverless Computing...
summary
System re-purposes idle warm containers from other functions instead of booting new ones, implemented as Pagurus with an intra-function manager, an inter-function scheduler and a cluster-level sharing-aware balancer. It builds a reusable zygote container that contains common packages and forks secured helper containers per function, and uses a similarity-filtered weighted random sampling policy to pick donors. Evaluation on Azure traces reduces cold startups by 84.6% and cuts alleviated cold-start latency to 16 ms.
AI-written summary · about
chapters (6)
- 0:00Intro
- 0:43Introduction & Background
- 4:30Motivation
- 10:02Methodology & Design
- 14:33Evaluation
- 16:34Conclusion
related talks
27mServerless Performance on a Budget
23mServerless in the Wild: Characterizing and Optimizing the Serverless Workload
37mGo FaaSter: Cold Start Optimization in a Serverless Platform
21mHarmonizing Efficiency and Practicability: Optimizing Resource Utilization in...
37mGo FaaSter: Cold Start Optimization in a Serverless Platform
33mGoing FaaSter: Cost-Performance Optimizations of Serverless on Kubernetes
48mServerless beyond functions
57mServerless or Containers? Who Cares Pick Both!
