Hi All, I had a MR job that processed 2000 small (<3MB ea.) files and it took 40 minutes on 8 nodes. Since the files are small it triggerred 2000 tasks. I packed my 2000 files into a single 445MB sequence file (K,V == Text,Text == <filename>,<file-content>). The new MR job triggers 7 map tasks (approx 64MB each) but it takes even longer (49 minutes) so I'm trying to figure out why.
I noticed these errors and I'm hoping someone can shed some light on why? Before I ran the job I ran hadoop fsck / and everything was healthy. e.g.: no under-replicated, no corrupt blocks etc. ...... 2010-07-13 03:24:20,807 INFO org.apache.hadoop.mapred.ReduceTask: GetMapEventsThread exiting 2010-07-13 03:24:20,807 INFO org.apache.hadoop.mapred.ReduceTask: getMapsEventsThread joined. 2010-07-13 03:24:20,808 INFO org.apache.hadoop.mapred.ReduceTask: Closed ram manager 2010-07-13 03:24:20,808 INFO org.apache.hadoop.mapred.ReduceTask: Interleaved on-disk merge complete: 7 files left. 2010-07-13 03:24:20,808 INFO org.apache.hadoop.mapred.ReduceTask: In-memory merge complete: 0 files left. 2010-07-13 03:24:20,814 INFO org.apache.hadoop.mapred.ReduceTask: Merging 7 files, 2401573706 bytes from disk 2010-07-13 03:24:20,815 INFO org.apache.hadoop.mapred.ReduceTask: Merging 0 segments, 0 bytes from memory into reduce 2010-07-13 03:24:20,818 INFO org.apache.hadoop.mapred.Merger: Merging 7 sorted segments 2010-07-13 03:24:20,827 INFO org.apache.hadoop.mapred.Merger: Down to the last merge-pass, with 7 segments left of total size: 2401573678 bytes 2010-07-13 03:30:42,329 INFO org.apache.hadoop.hdfs.DFSClient: Exception in createBlockOutputStream java.io.IOException: Bad connect ack with firstBadLink 10.15.46.73:50010 2010-07-13 03:30:42,329 INFO org.apache.hadoop.hdfs.DFSClient: Abandoning block blk_4304053493083580280_260714 2010-07-13 03:31:03,846 INFO org.apache.hadoop.hdfs.DFSClient: Exception in createBlockOutputStream java.io.IOException: Bad connect ack with firstBadLink 10.15.46.35:50010 2010-07-13 03:31:03,846 INFO org.apache.hadoop.hdfs.DFSClient: Abandoning block blk_3680469905814989852_260716 2010-07-13 03:31:08,233 INFO org.apache.hadoop.hdfs.DFSClient: Exception in createBlockOutputStream java.io.IOException: Bad connect ack with firstBadLink 10.15.46.35:50010 2010-07-13 03:31:08,233 INFO org.apache.hadoop.hdfs.DFSClient: Abandoning block blk_-673505196560500372_260717 2010-07-13 03:31:14,243 INFO org.apache.hadoop.hdfs.DFSClient: Exception in createBlockOutputStream java.io.IOException: Bad connect ack with firstBadLink 10.15.46.73:50010 2010-07-13 03:31:14,243 INFO org.apache.hadoop.hdfs.DFSClient: Abandoning block blk_-7054031797345836167_260717 ......
