List view
Major Feature to be added: Fault Tolerance for 1. Machine failure 2. Temporary Network failure or delay (soft errors) 3. Permanent Network failure? Other workitems: Clean up logging and establish correct logging framework. create a release version so that floe2 can be executed in a distributed mode without compilation. vagrant cluster for easy setup using VMs for testing.
Overdue by 11 year(s)•Due by September 14, 2014•2/3 issues closedx.y.1 milestones will be used to fix any any pending or new issues found in x.y.0 version.
Overdue by 11 year(s)•Due by September 8, 2014•20/20 issues closedFirst working prototype with atleast the following features: 1. Local, psuedo-distributed and distributed mode execution. 2. Scale up/scale down within a container. 3. Scale out/in across containers. 4. Command line tool to launch various components and send commands to the coordinator including but not limited to start application, stop application, scale up/out a given PE by given number of instances (same for scale down/in). 5. ClusterResourceManager - Resource mapping for fixed number of containers. 6. Monitoring framework to monitor (a) app statistics at the level of per pellet/flake (b) Stretch: resource monitoring - CPU and N/W bandwidth. 7. Dynamic Tasks with convenient ApplicationBuilder APIs. 8. Command line too to switch between alternates at runtime. 9. Stretch Goal: Dynamic Paths with convenient ApplicationBuilder APIs and Command line tool.
Overdue by 11 year(s)•Due by September 4, 2014•16/16 issues closed