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

    https://github.com/apache/spark/pull/8007#discussion_r38494894
  
    --- Diff: 
core/src/main/scala/org/apache/spark/scheduler/cluster/YarnSchedulerBackend.scala
 ---
    @@ -91,6 +96,43 @@ private[spark] abstract class YarnSchedulerBackend(
       }
     
       /**
    +   * Override the DriverEndpoint to add extra logic for the case when an 
executor is disconnected.
    +   * We should check the cluster manager and find if the loss of the 
executor was caused by YARN
    +   * force killing it due to preemption.
    +   */
    +  private class YarnDriverEndpoint(rpcEnv: RpcEnv, sparkProperties: 
ArrayBuffer[(String, String)])
    +      extends DriverEndpoint(rpcEnv, sparkProperties) {
    +
    +    /**
    +     * When onDisconnected is received at the driver endpoint, the 
superclass DriverEndpoint
    +     * handles it by assuming the Executor was lost for a bad reason and 
removes the executor
    +     * immediately.
    +     *
    +     * In YARN's case however it is crucial to talk to the application 
master and ask why the
    +     * executor had exited. In particular, the executor may have exited 
due to the executor
    +     * having been preempted. If the executor "exited normally" according 
to the application
    +     * master then we pass that information down to the TaskSetManager to 
inform the
    +     * TaskSetManager that tasks on that lost executor should not count 
towards a job failure.
    +     */
    +    override def onDisconnected(rpcAddress: RpcAddress): Unit = {
    +      addressToExecutorId.get(rpcAddress).foreach { executorId =>
    --- End diff --
    
    Man this is pretty hard to read. I wonder if something like this would 
compile:
    
          addressToExecutorId.get(rpcAddress).foreach { executorId =>
            yarnSchedulerEndpoint.askForExecutorLossReason(executorId, 
              reason => 
driverEndpoint.askWithRetry[Boolean](RemoveExecutor(executorId, reason),
              e => {
                logWarning(s"Attempted to get executor loss reason" +
                  s" for $rpcAddress but got no response. Marking as slave 
lost.", e)
                driverEndpoint.askWithRetry[Boolean](RemoveExecutor(executorId, 
SlaveLost()))
              })
          }
    
    Since this is the only call site for `askForExecutorLossReason` that I can 
see, it might be better to just inline the code. Should make it more readable.


---
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