cs 744 gandiva
play

CS 744: GANDIVA Shivaram Venkataraman Fall 2019 ADMINISTRIVIA - - PowerPoint PPT Presentation

CS 744: GANDIVA Shivaram Venkataraman Fall 2019 ADMINISTRIVIA - Course project proposal - Midterm Bismarck Supervised learning, Unified Interface Shared memory, Model fits in memory Parameter Server Large datasets, large models (PB


  1. CS 744: GANDIVA Shivaram Venkataraman Fall 2019

  2. ADMINISTRIVIA - Course project proposal - Midterm

  3. Bismarck Supervised learning, Unified Interface Shared memory, Model fits in memory Parameter Server Large datasets, large models (PB scale) Consistency model, Fault tolerance Machine Learning Tensorflow Need for flexible programming model Dataflow graph, Heterogeneous accelerators Ray Reinforcement learning applications Actors and tasks, Local and global scheduler

  4. Applications Machine Learning SQL Streaming Graph Computational Engines Scalable Storage Systems Resource Management Datacenter Architecture

  5. MACHINE LEARNING WORKFLOW?

  6. SHARED ML CLUSTERS Rack

  7. WORKLOAD Feedback-driven exploration

  8. AFFINITY

  9. INTRA JOB PREDICTABILITY

  10. MECHANISMS (1) Rack 1. Suspend-Resume 2. Migration

  11. MECHANISMS (2) Rack 3. Grow-shrink 4. Profiling

  12. SCHEDULING POLICY Goals early feedback cluster efficiency cluster-level fairness? Two modes Reactive Introspective

  13. REACTIVE MODE React to events Job arrivals, departures, failures Hierarchical Preference Nodes with same “ affinity” Nodes with “ different affinity ” Nodes with “no affinity” Suspend-resume …

  14. INTROSPECTIVE MODE Monitor and optimize placement of jobs periodically Actions Packing Migration Grow-shrink

  15. DISCUSSION https://forms.gle/aHYbNcTFdGJtXefj9

  16. What are some guarantees provided by Mesos that are not provided by Gandiva? Explain with an example

  17. Are mechanisms in Gandiva also useful in a cluster running Apache Spark jobs? Provide one example either for or against

  18. NEXT STEPS New module on SQL! Course project introductions Midterm

More recommend