viirya commented on a change in pull request #24699: [SPARK-27666][CORE] Do not
release lock while TaskContext already completed
URL: https://github.com/apache/spark/pull/24699#discussion_r294263527
##########
File path: core/src/test/scala/org/apache/spark/rdd/RDDSuite.scala
##########
@@ -1176,6 +1177,31 @@ class RDDSuite extends SparkFunSuite with
SharedSparkContext {
}.collect()
}
+ test("SPARK-27666: Do not release lock while TaskContext already completed")
{
+ val rdd = sc.parallelize(Range(0, 10), 1).cache()
+ val tid = sc.longAccumulator("threadId")
+ // validate cache
+ rdd.collect()
+ rdd.mapPartitions { iter =>
+ val t = new Thread(() => {
+ while (iter.hasNext) {
+ iter.next()
+ Thread.sleep(100)
+ }
+ })
+ t.setDaemon(false)
+ t.start()
+ tid.add(t.getId)
+ Iterator(0)
+ }.collect()
+ val tmx = ManagementFactory.getThreadMXBean
+ var t = tmx.getThreadInfo(tid.value)
+ // getThreadInfo() will return null after child thread `t` died
+ while (t != null && t.getThreadState != Thread.State.TERMINATED) {
+ t = tmx.getThreadInfo(tid.value)
+ }
+ }
Review comment:
Can we use eventually with timeout instead of a while loop?
```scala
eventually(timeout(10.seconds)) {
val t = tmx.getThreadInfo(tid.value)
assert(t == null || t.getThreadState == Thread.State.TERMINATED)
}
```
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
