[YARN-1076] RM gets stuck with a reservation, ignoring new containers

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Duplicate
• Component/s: resourcemanager
• Assignee: Omkar Vinit Joshi
• Reporter: Maysam Yabandeh
• Created: 17/Aug/13 15:15
• Updated: 23/Aug/13 20:54
• Resolved: 23/Aug/13 20:54

Description

LeafQueue#assignToQueue rejects newly available containers if potentialNewCapacity > absoluteMaxCapacity:

  private synchronized boolean assignToQueue(Resource clusterResource, 
      Resource required) {
    // Check how of the cluster's absolute capacity we are currently using...
    float potentialNewCapacity = 
        Resources.divide(
            resourceCalculator, clusterResource, 
            Resources.add(usedResources, required), 
            clusterResource);
    if (potentialNewCapacity > absoluteMaxCapacity) {
      //... 
      return false;
    }
    return true;
  }

The usedResources, which is used to computed potentialNewCapacity, is composed of both actual and reserved containers. So, a prior reservation could causes RM to reject newly available containers, despite the starvation report.

Issue Links

• is duplicated by: YARN-957 Capacity Scheduler tries to reserve the memory more than what node manager reports.