We have seen couple of Block missing issues in stress testing. ( mostly all were with default configurations )
1) HDFS-3584 Blocks are getting marked as corrupt with append operation under high load.
2) HDFS-3585 Blocks are getting marked as corrupt when commitblocksyncronization(Since recovery called by another client) is success before BR send
3) HDFS-3122 Block recovery with closeFile flag true can race with blockReport. Due to this blocks are getting marked as corrupt.
4) HDFS-3605 Block mistakenly marked corrupt during edit log catchup phase of failover ===== Already Fixed
I would like your involvement in the discussion and comments on this issues for the further progress.