Custom Checkpoint, Performing check pointing in hadoop hadoop, Edit log size hadoop, edit log size more hoadoop

Some time it may happen that checkpoint will not happen, and edit size becomes more, of if check point is not happening the is a chance of data loss, if something bad happens, for this

Verify Secondry namenode is up and running fine

• check for services ps -ef | grep secondary\
• check if log rotation of secondary name node is happening properly

 Solution :

If secondary name node is running fine the, you need to checkpoint yourself which you can do as follows

Check if cluster is in safe mode if not send it to safe mode use following command to make hadoop in safe mode
 

• hadoop dfsadmin -safemode enter

    And use following command

• hadoop dfsadmin -saveNamespace

After that issue following command for normalize the cluster

• hadoop dfsadmin -safemode enter

This will do the check point and the edit size will reduce for sure.  Happy Hadooping :)