Completed
Why juggle at all?
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Practical Container Scheduling: Optimizations, Guarantees, and Trade-Offs at Netflix - Lecture
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Reactive stream processing: Mantis
- 3 Container deployment: Titus
- 4 What the cluster needs to support - Heterogeneous mix of workload
- 5 Why juggle at all?
- 6 Scheduling challenge in large clusters
- 7 Our initial goals for a cluster scheduler • Multi goal optimization for task placement . Cluster autoscaling • Extensibility
- 8 Multi goal task placement
- 9 Security
- 10 Capacity guarantees
- 11 Fenzo scheduling strategy
- 12 Fitness functions we use • CPU, memory, and network in packing
- 13 Hard constraints we use • GPU server matching
- 14 Soft constraints we use • Specified by individual jobs at submittime • Balance tasks of a job across availability zones
- 15 Mixing fitness with soft constraints
- 16 Our queues setup
- 17 Sizing agent clusters for capacity
- 18 Reasoning about allocation failures
- 19 What's next?
- 20 Questions?
