Github user squito commented on a diff in the pull request:

    https://github.com/apache/spark/pull/6394#discussion_r32243839
  
    --- Diff: 
yarn/src/test/scala/org/apache/spark/deploy/yarn/YarnAllocatorSuite.scala ---
    @@ -242,4 +245,154 @@ class YarnAllocatorSuite extends SparkFunSuite with 
Matchers with BeforeAndAfter
         assert(pmemMsg.contains("2.1 MB of 2 GB physical memory used."))
       }
     
    +  test("allocate locality preferred containers with enough resource and no 
matched existed " +
    +    "containers") {
    +    // 1. All the locations of current containers cannot satisfy the new 
requirements
    +    // 2. Current requested container number can fully satisfy the pending 
tasks.
    +
    +    val handler = createAllocator(2)
    +    handler.updateResourceRequests()
    +    handler.handleAllocatedContainers(Array(
    +      createContainer("host1"),
    +      createContainer("host2")))
    +
    +    handler.requestTotalExecutorsWithPreferredLocalities(
    +      5, 15, Map("host3" -> 15, "host4" -> 15, "host5" -> 10))
    +    handler.updateResourceRequests()
    +    val nodesOfRequests = handler.containerRequests.toArray.map { request 
=>
    +      if (request.getNodes == null) {
    +        null
    +      } else {
    +        request.getNodes.toArray(new Array[String](request.getNodes.size))
    +      }
    +    }
    +
    +    assert(nodesOfRequests === Array(
    +      null, // container1 request
    +      null, // container2 request
    --- End diff --
    
    I am a little confused by the comments -- I thought that the `nodes` in the 
request will be null when you are requesting new containers, but there is 
absolutely no preference for which node the container ends up on?  The comment 
makes me think this is a request which is somehow related to the existing 
container1 on host1.
    
    (but again I'm kinda new to this code ... so this might just show my 
general level of confusion ...)


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to