Yesterday our root diectory(/) has no space left in namenode server. After that namenode terminated. After restart namenode all the files in HDFS were corrupted.
If i run hdfs hdfs fsck / and search one block that listed corrupted in datanode, that block is reachable and healthy.
How can i found the problem and how can i recover that problem.
also it says under replicated but nothing happened since yesterday.
Are you sure that datanodes are running fine? How many datanodes shown in the namenode web UI?