[AMBARI-16888] Handle the scenario when 'capacity-scheduler' configs is passed in as dictionary to Stack Advisor (generally on 1st invocation) in order to create 'llap' queue for Hive Server Interactive.

• Type: Bug
• Status: Resolved
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 2.4.0
• Fix Version/s: 2.4.0
• Component/s: ambari-server
• Assignee: Swapan Shridhar
• Reporter: Swapan Shridhar
• Due: 25/May/16
• Created: 25/May/16 22:10
• Updated: 26/May/16 04:51
• Resolved: 26/May/16 01:47

Description

• The 1st call to SA gets capacity-scheduler configs as dictionary compared to "/n" separated single string of all the configs in subsequent calls.

• 1st invocation, passed-in capacity-scheduler looks like :

          "capacity-scheduler" : {
            "properties" : {
              "capacity-scheduler" : "null",
              "yarn.scheduler.capacity.root.accessible-node-labels" : "*",
              "yarn.scheduler.capacity.maximum-am-resource-percent" : "1",
              "yarn.scheduler.capacity.root.acl_administer_queue" : "*",
              "yarn.scheduler.capacity.queue-mappings-override.enable" : 'false',
              "yarn.scheduler.capacity.root.default.capacity" : "100",
              "yarn.scheduler.capacity.root.default.user-limit-factor" : "1",
              "yarn.scheduler.capacity.root.queues" : "default",
              "yarn.scheduler.capacity.root.capacity" : "100",
              "yarn.scheduler.capacity.root.default.acl_submit_applications" : "*",
              "yarn.scheduler.capacity.root.default.maximum-capacity" : "100",
              "yarn.scheduler.capacity.node-locality-delay" : "40",
              "yarn.scheduler.capacity.maximum-applications" : "10000",
              "yarn.scheduler.capacity.root.default.state" : "RUNNING"
            }
          },
  

• subsequent invocations gets capacity-schdeuler as:

  "capacity-scheduler": {
            "properties": {
              "capacity-scheduler": "yarn.scheduler.capacity.root.queues=default\n"
                                    "yarn.scheduler.capacity.root.default.user-limit-factor=1\n"
                                    "yarn.scheduler.capacity.root.default.state=RUNNING\n"
                                    "yarn.scheduler.capacity.root.default.maximum-capacity=100\n"
                                    "yarn.scheduler.capacity.root.default.capacity=100\n"
                                    "yarn.scheduler.capacity.root.default.acl_submit_applications=*\n"
                                    "yarn.scheduler.capacity.root.capacity=100\n"
                                    "yarn.scheduler.capacity.root.acl_administer_queue=*\n"
                                    "yarn.scheduler.capacity.root.accessible-node-labels=*\n"
                                    "yarn.scheduler.capacity.node-locality-delay=40\n"
                                    "yarn.scheduler.capacity.maximum-applications=10000\n"
                                    "yarn.scheduler.capacity.maximum-am-resource-percent=1\n"
                                    "yarn.scheduler.capacity.queue-mappings-override.enable=false\n"
            }
          },
  

• Therefore, SA fails to create the 'llap' queue for Hive Server Interactive, as SA knows to handle only the 2nd case.
• This specific issue was seen with while deploying cluster with Blueprints.

Issue Links

• links to: Reviewboard Link