I'm using Spark 1.5.2 and my RDD has 512 equally sized partitions and is 100% cached in memory across 512 executors.
I have a filter-map-collect job with 512 tasks. Sometimes this job completes sub-second. On other occasions when I run it 50% of the tasks complete sub-second, 45% of the tasks take 10 seconds and 5% of the tasks take 20 seconds. Here is the log from an executor where the task took 20 seconds: 15/12/16 09:44:37 INFO executor.CoarseGrainedExecutorBackend: Got assigned task 5312 15/12/16 09:44:37 INFO executor.Executor: Running task 215.0 in stage 17.0 (TID 5312) 15/12/16 09:44:37 INFO broadcast.TorrentBroadcast: Started reading broadcast variable 10 15/12/16 09:44:37 INFO storage.MemoryStore: ensureFreeSpace(1777) called with curMem=908793307, maxMem=5927684014 15/12/16 09:44:37 INFO storage.MemoryStore: Block broadcast_10_piece0 stored as bytes in memory (estimated size 1777.0 B, free 4.7 GB) 15/12/16 09:44:37 INFO broadcast.TorrentBroadcast: Reading broadcast variable 10 took 186 ms 15/12/16 09:44:37 INFO storage.MemoryStore: ensureFreeSpace(3272) called with curMem=908795084, maxMem=5927684014 15/12/16 *09:44:37* INFO storage.MemoryStore: Block broadcast_10 stored as values in memory (estimated size 3.2 KB, free 4.7 GB) 15/12/16 *09:44:57* INFO storage.BlockManager: Found block rdd_5_215 locally 15/12/16 09:44:57 INFO executor.Executor: Finished task 215.0 in stage 17.0 (TID 5312). 2074 bytes result sent to driver So it appears the 20 seconds is spent finding the local block. Since the lag is always either exactly 10 seconds or exactly 20 seconds I suspect it's due to a 10 second timeout on some listener, or something like that. If that is true then I guess my options are either find out why it's timing out and fix it or make the timeout shorter so it tries more frequently. Any advice welcome. Many thanks in advance, Pat -- View this message in context: http://apache-spark-user-list.1001560.n3.nabble.com/Some-tasks-take-a-long-time-to-find-local-block-tp25727.html Sent from the Apache Spark User List mailing list archive at Nabble.com. --------------------------------------------------------------------- To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org