Bulk Operation

YARN-1011 NM crashes because Fair scheduler promotes a container that has not been pulled by AM

YARN-1011 Fair Scheduler promotion does not update container execution type when the application is killed

YARN-1011 ResourceUtilization cpu is misused in oversubscription as a percentage

YARN-1011 SnapshotBasedOverAllocationPolicy always caps the amount of memory availabe to 4 GBs

YARN-1011 Allow Applications to demand Guaranteed Containers with Capacity Scheduler

YARN-1011 Plumb aggregated application resource utilization from the NM to RM

YARN-1011 Improve debug messages for NM preemption of OPPORTUNISTIC containers

YARN-1011 Refactor AbstractYarnScheduler and CapacityScheduler OPPORTUNISTIC container completion codepaths

YARN-1011 Use aggregate container utilization instead of node utilization to determine resources available for oversubscription

YARN-1011 FairScheduler crashes RM with oversubscription turned on if an application is killed.

YARN-1011 Fair scheduler promotion does not update container execution type and token

YARN-1011 Don't start opportunistic containers at container scheduler/finish event with over-allocation

YARN-1011 Add queue-level control to allow all applications in a queue to opt-out

YARN-1011 Make sure AM is not scheduled on an opportunistic container

YARN-1011 Exclude findbugs warnings on SchedulerNode.numGuaranteedContainers and numOpportunisticContainers

YARN-1011 Preserve execution type for containers to be increased by AM post YARN-1015

YARN-1011 Expose per-node over-allocation info in Node Report

YARN-1011 Add documentation of oversubscription

YARN-1011 Allow resource request to opt out of oversubscription in Fair Scheduler

YARN-1011 Fair Scheduler to explicitly promote OPPORTUNISITIC containers locally at the node where they're running

YARN-1011 Add a configuration to cap how much a NM can be overallocated

YARN-1011 Crash NM at start time if oversubscription is on but LinuxContainerExcutor or cgroup is off

YARN-1011 Refactor ContainerScheduler to make oversubscription change easier

YARN-1011 Add separate NM preemption thresholds for cpu and memory

YARN-1011 Add NM support to launch opportunistic containers based on overallocation

YARN-1011 Add NM preemption of opportunistic containers when utilization goes high

YARN-1011 Add separate NM overallocation thresholds for cpu and memory

YARN-1011 Expose per-application over-allocation info in the Resource Manager

YARN-1011 Handle enforceExecutionType == false in AMRMClient

YARN-1011 Provide a knob to turn on over-allocation

YARN-1011 Common scheduler changes to support scheduler-specific oversubscription implementations

YARN-1011 FS should watch node resource utilization and allocate opportunistic containers if appropriate

YARN-1011 CS should watch resource utilization of containers and allocate speculative containers if appropriate