Hey all,

I encountered an issue where an HDFS block was lost because a single node
(dn-0) went down. The setup used dfs.replication=2. The data loss had two
bugs conspiring to have this happen.

I already filed both issues:
https://issues.apache.org/jira/browse/HDFS-17983
https://issues.apache.org/jira/browse/HDFS-17984
They include proposed patches.

The short summary: dn-1 discarded its block due a mismatch in disk and
in-memory state (https://issues.apache.org/jira/browse/HDFS-17983), dn-0
reported the block as non-existing during boot, due to a race condition (
https://issues.apache.org/jira/browse/HDFS-17984). dn-1 running the
recovery determined the block was lost.

I would appreciate a review and guidance on these issues, what i can do to
get these in a state they can be accepted by HDFS trunk.

Kind regards,

Bram Schuur

Reply via email to