GroupByKey stalls with GroupingShuffleReader split refusals

With info messages with something like:
> `Refused to split GroupingShuffleReader at ShufflePosition(base64:AAAAA1CWNvkAAQ)`
>
> e lull messages look like this:
> `Processing lull for PT300.006S in state read-shuffle of Write Vectors2/GroupIntoShards/Read at com.google.cloud.dataflow.worker.ApplianceShuffleReader.readIncludingPosition(Native Method) at com.google.cloud.dataflow.worker.ChunkingShuffleBatchReader.read(ChunkingShuffleBatchReader.java:62) at com.google.cloud.dataflow.worker.util.common.worker.CachingShuffleBatchReader$1.load(CachingShuffleBatchReader.java:57) at com.google.cloud.dataflow.worker.util.common.worker.CachingShuffleBatchReader$1.load(CachingShuffleBatchReader.java:53) at com.google.cloud.dataflow.worker.repackaged.com.google.common.cache.LocalCache$LoadingValueReference.loadFuture(LocalCache.java:3628) at com.google.cloud.dataflow.worker.repackaged.com.google.common.cache.LocalCache$Segment.loadSync(LocalCache.java:2336) at com.google.cloud.dataflow.worker.repackaged.com.google.common.cache.LocalCache$Segment.lockedGetOrLoad(LocalCache.java:2295) at com.google.cloud.dataflow.worker.repackaged.com.google.common.cache.LocalCache$Segment.get(LocalCache.java:2208) at com.google.cloud.dataflow.worker.repackaged.com.google.common.cache.LocalCache.get(LocalCache.java:4053) at com.google.cloud.dataflow.worker.repackaged.com.google.common.cache.LocalCache.getOrLoad(LocalCache.java:4057) at com.google.cloud.dataflow.worker.repackaged.com.google.common.cache.LocalCache$LocalLoadingCache.get(LocalCache.java:4986) at com.google.cloud.dataflow.worker.util.common.worker.CachingShuffleBatchReader.read(CachingShuffleBatchReader.java:76) at com.google.cloud.dataflow.worker.util.common.worker.BatchingShuffleEntryReader$ShuffleReadIterator.fillEntries(BatchingShuffleEntryReader.java:133) at com.google.cloud.dataflow.worker.util.common.worker.BatchingShuffleEntryReader$ShuffleReadIterator.fillEntriesIfNeeded(BatchingShuffleEntryReader.java:126) at com.google.cloud.dataflow.worker.util.common.worker.BatchingShuffleEntryReader$ShuffleReadIterator.hasNext(BatchingShuffleEntryReader.java:90) at com.google.cloud.dataflow.worker.util.common.ForwardingReiterator.hasNext(ForwardingReiterator.java:62) at com.google.cloud.dataflow.worker.util.common.worker.GroupingShuffleEntryIterator.advance(GroupingShuffleEntryIterator.java:118) at com.google.cloud.dataflow.worker.GroupingShuffleReader$GroupingShuffleReaderIterator.advance(GroupingShuffleReader.java:230) at com.google.cloud.dataflow.worker.GroupingShuffleReader$GroupingShuffleReaderIterator.start(GroupingShuffleReader.java:224) at com.google.cloud.dataflow.worker.util.common.worker.ReadOperation$SynchronizedReaderIterator.start(ReadOperation.java:347) at com.google.cloud.dataflow.worker.util.common.worker.ReadOperation.runReadLoop(ReadOperation.java:183) at com.google.cloud.dataflow.worker.util.common.worker.ReadOperation.start(ReadOperation.java:148) at com.google.cloud.dataflow.worker.util.common.worker.MapTaskExecutor.execute(MapTaskExecutor.java:68) at com.google.cloud.dataflow.worker.DataflowWorker.executeWork(DataflowWorker.java:330) at com.google.cloud.dataflow.worker.DataflowWorker.doWork(DataflowWorker.java:302) at com.google.cloud.dataflow.worker.DataflowWorker.getAndPerformWork(DataflowWorker.java:251) at com.google.cloud.dataflow.worker.DataflowBatchWorkerHarness$WorkerThread.doWork(DataflowBatchWorkerHarness.java:135) at com.google.cloud.dataflow.worker.DataflowBatchWorkerHarness$WorkerThread.call(DataflowBatchWorkerHarness.java:115) at com.google.cloud.dataflow.worker.DataflowBatchWorkerHarness$WorkerThread.call(DataflowBatchWorkerHarness.java:102) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617) at java.lang.Thread.run(Thread.java:745)`
>

Imported from Jira [BEAM-3487](https://issues.apache.org/jira/browse/BEAM-3487). Original Jira may contain additional context.
Reported by: ikarienator.

Inizia dai frame GroupingShuffleReader e GroupingShuffleEntryIterator nello stack trace Java riportato, quindi esamina il percorso circostante di lettura dello shuffle e il messaggio di rifiuto dello split. Riproduci il blocco di GroupByKey, se possibile, e determina perché read-shuffle entra in una pausa prolungata. Il lavoro è completato quando l’elaborazione GroupByKey interessata non si blocca più a causa di ripetuti rifiuti dello split da parte di GroupingShuffleReader.

Scritto dal modello di indicizzazione a partire dal testo della issue.