Migration to Scala 2.10
== Below description was written by Prashant Sharma ==
This PR migrates spark to scala 2.10.
Summary of changes apart from scala 2.10 migration:
(has no implications for user.)
1. Migrated Akka to 2.2.3.
Does not use remote death watch for it has a bug, where it tries to send message to dead node infinitely.
Uses an indestructible actorsystem which tolerates errors only on executors.
(Might be useful for user.)
4. New configuration settings introduced:
System.getProperty("spark.akka.heartbeat.pauses", "600")
System.getProperty("spark.akka.failure-detector.threshold", "300.0")
System.getProperty("spark.akka.heartbeat.interval", "1000")
Defaults for these are fairly large to only disable Failure detector that comes with akka. The reason for doing so is we have our own failure detector like mechanism in place and then this is just an overhead on top of that + it leads to a lot of false positives. But with these properties it is possible to enable them. A good use case for enabling it could be when someone wants spark to be sensitive (in a controllable manner ofc.) to GC pauses/Network lags and quickly evict executors that experienced it. More information is included in configuration.md
Once we have the SPARK-544 merged, I had like to deprecate atleast these akka properties and may be others too.
This PR is duplicate of #221(where all the discussion happened.) for that one pointed to master this one points to scala-2.10 branch.
Scala 2.10 migration
This PR migrates spark to scala 2.10.
Summary of changes apart from scala 2.10 migration:
(has no implications for user.)
1. Migrated Akka to 2.2.3.
Does not use remote death watch for it has a bug, where it tries to send message to dead node infinitely.
Uses an indestructible actorsystem which tolerates errors only on executors.
(Might be useful for user.)
4. New configuration settings introduced:
System.getProperty("spark.akka.heartbeat.pauses", "600")
System.getProperty("spark.akka.failure-detector.threshold", "300.0")
System.getProperty("spark.akka.heartbeat.interval", "1000")
Defaults for these are fairly large to only disable Failure detector that comes with akka. The reason for doing so is we have our own failure detector like mechanism in place and then this is just an overhead on top of that + it leads to a lot of false positives. But with these properties it is possible to enable them. A good use case for enabling it could be when someone wants spark to be sensitive (in a controllable manner ofc.) to GC pauses/Network lags and quickly evict executors that experienced it. More information is included in configuration.md
Once we have the SPARK-544 merged, I had like to deprecate atleast these akka properties and may be others too.
This PR is duplicate of #221(where all the discussion happened.) for that one pointed to master this one points to scala-2.10 branch.
README incorrectly suggests build sources spark-env.sh
This is misleading because the build doesn't source that file. IMO
it's better to force people to specify build environment variables
on the command line always, like we do in every example, so I'm
just removing this doc.
This is misleading because the build doesn't source that file. IMO
it's better to force people to specify build environment variables
on the command line always, like we do in every example.
stageId <--> jobId mapping in DAGScheduler
Okay, I think this one is ready to go -- or at least it's ready for review and discussion. It's a carry-over of https://github.com/mesos/spark/pull/842 with updates for the newer job cancellation functionality. The prior discussion still applies. I've actually changed the job cancellation flow a bit: Instead of ``cancelTasks`` going to the TaskScheduler and then ``taskSetFailed`` coming back to the DAGScheduler (resulting in ``abortStage`` there), the DAGScheduler now takes care of figuring out which stages should be cancelled, tells the TaskScheduler to cancel tasks for those stages, then does the cleanup within the DAGScheduler directly without the need for any further prompting by the TaskScheduler.
I know of three outstanding issues, each of which can and should, I believe, be handled in follow-up pull requests:
1) https://spark-project.atlassian.net/browse/SPARK-960
2) JobLogger should be re-factored to eliminate duplication
3) Related to 2), the WebUI should also become a consumer of the DAGScheduler's new understanding of the relationship between jobs and stages so that it can display progress indication and the like grouped by job. Right now, some of this information is just being sent out as part of ``SparkListenerJobStart`` messages, but more or different job <--> stage information may need to be exported from the DAGScheduler to meet listeners needs.
Except for the eventQueue -> Actor commit, the rest can be cherry-picked almost cleanly into branch-0.8. A little merging is needed in MapOutputTracker and the DAGScheduler. Merged versions of those files are in aba2b40ce0
Note that between the recent Actor change in the DAGScheduler and the cleaning up of DAGScheduler data structures on job completion in this PR, some races have been introduced into the DAGSchedulerSuite. Those tests usually pass, and I don't think that better-behaved code that doesn't directly inspect DAGScheduler data structures should be seeing any problems, but I'll work on fixing DAGSchedulerSuite as either an addition to this PR or as a separate request.
UPDATE: Fixed the race that I introduced. Created a JIRA issue (SPARK-965) for the one that was introduced with the switch to eventProcessorActor in the DAGScheduler.
Change the name of input argument in ClusterScheduler#initialize from context to backend.
The SchedulerBackend used to be called ClusterSchedulerContext so just want to make small
change of the input param in the ClusterScheduler#initialize to reflect this.
Added logging of scheduler delays to UI
This commit adds two metrics to the UI:
1) The time to get task results, if they're fetched remotely
2) The scheduler delay. When the scheduler starts getting overwhelmed (because it can't keep up with the rate at which tasks are being submitted), the result is that tasks get delayed on the tail-end: the message from the worker saying that the task has completed ends up in a long queue and takes a while to be processed by the scheduler. This commit records that delay in the UI so that users can tell when the scheduler is becoming the bottleneck.
