[SPARK-28699][Core] Fix a corner case for aborting indeterminate stage #25498
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
cloud-fan
reviewed
Aug 19, 2019
core/src/test/scala/org/apache/spark/scheduler/DAGSchedulerSuite.scala
Outdated
Show resolved
Hide resolved
|
good catch! LGTM |
viirya
reviewed
Aug 19, 2019
|
Test build #109349 has finished for PR 25498 at commit
|
viirya
reviewed
Aug 19, 2019
| @@ -2741,27 +2741,8 @@ class DAGSchedulerSuite extends SparkFunSuite with LocalSparkContext with TimeLi | |||
| FetchFailed(makeBlockManagerId("hostC"), shuffleId2, 0, 0, "ignored"), | |||
| null)) | |||
|
|
|||
| val failedStages = scheduler.failedStages.toSeq | |||
| assert(failedStages.length == 2) | |||
There was a problem hiding this comment.
not big deal, but I think this assert still applies?
There was a problem hiding this comment.
After this change, failedStages.length == 0 because we do the cleanup work in failJobAndIndependentStages by cleanupStateForJobAndIndependentStages.
viirya
approved these changes
Aug 19, 2019
viirya
reviewed
Aug 19, 2019
|
Test build #109378 has finished for PR 25498 at commit
|
|
thanks, merging to master! @xuanyuanking can you send PRs for branch 2.3 and 2.4? the code conflicts. |
|
Sure, I'm doing the backport now. |
xuanyuanking
added a commit
to xuanyuanking/spark
that referenced
this pull request
Aug 20, 2019
Change the logic of collecting the indeterminate stage, we should look at stages from mapStage, not failedStage during handle FetchFailed.
In the fetch failed error handle logic, the original logic of collecting indeterminate stage from the fetch failed stage. And in the scenario of the fetch failed happened in the first task of this stage, this logic will cause the indeterminate stage to resubmit partially. Eventually, we are capable of getting correctness bug.
It makes the corner case of indeterminate stage abort as expected.
New UT in DAGSchedulerSuite.
Run below integrated test with `local-cluster[5, 2, 5120]`, and set `spark.sql.execution.sortBeforeRepartition`=false, it will abort the indeterminate stage as expected:
```
import scala.sys.process._
import org.apache.spark.TaskContext
val res = spark.range(0, 10000 * 10000, 1).map{ x => (x % 1000, x)}
// kill an executor in the stage that performs repartition(239)
val df = res.repartition(113).map{ x => (x._1 + 1, x._2)}.repartition(239).map { x =>
if (TaskContext.get.attemptNumber == 0 && TaskContext.get.partitionId < 1 && TaskContext.get.stageAttemptNumber == 0) {
throw new Exception("pkill -f -n java".!!)
}
x
}
val r2 = df.distinct.count()
```
Closes apache#25498 from xuanyuanking/SPARK-28699-followup.
Authored-by: Yuanjian Li <[email protected]>
Signed-off-by: Wenchen Fan <[email protected]>
(cherry picked from commit 0d3a783)
Signed-off-by: Yuanjian Li <[email protected]>
xuanyuanking
added a commit
to xuanyuanking/spark
that referenced
this pull request
Aug 20, 2019
Change the logic of collecting the indeterminate stage, we should look at stages from mapStage, not failedStage during handle FetchFailed.
In the fetch failed error handle logic, the original logic of collecting indeterminate stage from the fetch failed stage. And in the scenario of the fetch failed happened in the first task of this stage, this logic will cause the indeterminate stage to resubmit partially. Eventually, we are capable of getting correctness bug.
It makes the corner case of indeterminate stage abort as expected.
New UT in DAGSchedulerSuite.
Run below integrated test with `local-cluster[5, 2, 5120]`, and set `spark.sql.execution.sortBeforeRepartition`=false, it will abort the indeterminate stage as expected:
```
import scala.sys.process._
import org.apache.spark.TaskContext
val res = spark.range(0, 10000 * 10000, 1).map{ x => (x % 1000, x)}
// kill an executor in the stage that performs repartition(239)
val df = res.repartition(113).map{ x => (x._1 + 1, x._2)}.repartition(239).map { x =>
if (TaskContext.get.attemptNumber == 0 && TaskContext.get.partitionId < 1 && TaskContext.get.stageAttemptNumber == 0) {
throw new Exception("pkill -f -n java".!!)
}
x
}
val r2 = df.distinct.count()
```
Closes apache#25498 from xuanyuanking/SPARK-28699-followup.
Authored-by: Yuanjian Li <[email protected]>
Signed-off-by: Wenchen Fan <[email protected]>
(cherry picked from commit 0d3a783)
Signed-off-by: Yuanjian Li <[email protected]>
|
Thank you, @xuanyuanking , @cloud-fan and @viirya ! Also, cc @kiszk for |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What changes were proposed in this pull request?
When collecting the indeterminate stages for handling FetchFailed, we should look at stages from mapStage, instead of failedStage.
Why are the changes needed?
In the fetch failed error handle logic, the original logic of collecting indeterminate stage from the fetch failed stage. And in the scenario of the fetch failed happened in the first task of this stage, this logic will cause the indeterminate stage to resubmit partially. Eventually, we are capable of getting correctness bug.
Does this PR introduce any user-facing change?
It makes the corner case of indeterminate stage abort as expected.
How was this patch tested?
New UT in DAGSchedulerSuite.
Run below integrated test with
local-cluster[5, 2, 5120], and setspark.sql.execution.sortBeforeRepartition=false, it will abort the indeterminate stage as expected: