[ http://issues.apache.org/jira/browse/HADOOP-95?page=comments#action_12371188 ]
eric baldeschwieler commented on HADOOP-95: ------------------------------------------- I vote for a dedicated thread on each name node using 1% of disk IO to validate blocks continuously. Low cost and uncovering bad blocks is high value. This would also let you collect disk stats reliability to determine if the disks were bad. > dfs validation > -------------- > > Key: HADOOP-95 > URL: http://issues.apache.org/jira/browse/HADOOP-95 > Project: Hadoop > Type: Improvement > Components: dfs > Versions: 0.1 > Reporter: Hairong Kuang > > Dfs needs a validation operation similiar to fsck, so that we get to know the > files that are corrupted and which data blocks are missing. > Dfs namenode also needs to log more specific information such as which block > is replication or is deleted. So when something goes wrong, we have a clue > what has happened. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira
