Bulk Operation

YUNIKORN-553 Set default scheduling policy to Hard for gang scheduling

YUNIKORN-553 Document Soft/Hard scheduling styles

YUNIKORN-553 Cleaned-up placeholder allocations are still reported by Yunikorn

YUNIKORN-553 Create gangSchedulingStyle field for AddApplicationRequest

YUNIKORN-553 Add gangSchedulingStyle scheduling parameter

YUNIKORN-553 enhanced StateAware gang scheduling

YUNIKORN-553 Publish a pod event to indicate the task is being gang scheduling

YUNIKORN-553 Gang scheduling User Guide

YUNIKORN-553 Allocated resources on a node could become negative

YUNIKORN-553 Enhance placeholder cleanup on timeout

YUNIKORN-553 Create a Failing app state in shim side

YUNIKORN-553 Consider a fallback mechanism to schedule the app in case of gang failure instead of marking the app as failed

YUNIKORN-553 Update state machine doc

YUNIKORN-553 Check if the placeholders and real resources are the same

YUNIKORN-553 Make sure all timer go routines are stopped after removing an app

YUNIKORN-553 Add a label to the placeholder pods the selector

YUNIKORN-553 Gang scheduling Design doc

YUNIKORN-553 Remove the WARN log when placeholderTimeout is not defined in taskGroups

YUNIKORN-553 Expose pod level events when an app is failed in scheduling

YUNIKORN-553 Rename scheduling parameter placeholderTimeout to placeholderTimeoutInSeconds

YUNIKORN-553 Add UT coverage for YUNIKORN-460

YUNIKORN-553 Placeholder pods are not cleaned up even when the job is deleted

YUNIKORN-553 Publish events to app's pods if the app is failed to be scheduled