[ http://issues.apache.org/jira/browse/HADOOP-829?page=all ]

dhruba borthakur updated HADOOP-829:
------------------------------------

        Summary: Separate the datanode contents that is written to the fsimage 
vs the contents used in over-the-wire communication  (was: Seperate the 
datanode contents that is written to the fsimage vs the contents used in 
over-the-wire communication)
    Description: 
In the existing implementation, the Namenode has a data structure called 
DatanodeDescriptor. It is used to serialize contents of the Datanode while 
writing it to the fsimage. The same serialization method is used to send a 
Datanode object to the Datanode and/or the Client. This introduces the 
shortcoming that one cannot introduce non-persistent fields in the 
DatanodeDescriptor.

One solution is to separate out the following two functionality into two 
separate classes:
1. The fields from a Datanode that are part of the ClientProtocol and/or 
DatanodeProcotol.
2. The fields from a Datanode that are stored in the fsImage.

  was:
In the existing implementation, the Namenode has a data structure called 
DatanodeDescriptor. It is used to serialize contents of the Datanode while 
writing it to the fsimage. The same serialization method is used to send a 
Datanode object to the Datanode and/or the Client. This introduces the 
shortcoming that one cannot introduce non-persistent fields in the 
DatanodeDescriptor.

One solution is to seperate out the following two functionality into two 
seperate classes:
1. The fields from a Datanode that are part of the ClientProtocol and/or 
DatanodeProcotol.
2. The fields from a Datanode that are stored in the fsImage.


> Separate the datanode contents that is written to the fsimage vs the contents 
> used in over-the-wire communication
> -----------------------------------------------------------------------------------------------------------------
>
>                 Key: HADOOP-829
>                 URL: http://issues.apache.org/jira/browse/HADOOP-829
>             Project: Hadoop
>          Issue Type: Bug
>          Components: dfs
>            Reporter: dhruba borthakur
>         Assigned To: dhruba borthakur
>
> In the existing implementation, the Namenode has a data structure called 
> DatanodeDescriptor. It is used to serialize contents of the Datanode while 
> writing it to the fsimage. The same serialization method is used to send a 
> Datanode object to the Datanode and/or the Client. This introduces the 
> shortcoming that one cannot introduce non-persistent fields in the 
> DatanodeDescriptor.
> One solution is to separate out the following two functionality into two 
> separate classes:
> 1. The fields from a Datanode that are part of the ClientProtocol and/or 
> DatanodeProcotol.
> 2. The fields from a Datanode that are stored in the fsImage.

-- 
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

        

Reply via email to