hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Suresh Srinivas (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HDFS-4979) Implement retry cache on the namenode
Date Thu, 11 Jul 2013 01:12:47 GMT

     [ https://issues.apache.org/jira/browse/HDFS-4979?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Suresh Srinivas updated HDFS-4979:
----------------------------------

    Attachment: HDFS-4979.patch

Here is an early patch that adds retry cache to Namenode. Retry is supported for all the operations
except:
* saveNamespace() - if the implementation uses writeLock() instead of readLock() this can
be made idempotent
* create() and append() - want to first get startFileInternal() cleanup from HDFS-4912.
* updatePipeLine() and deleteSnapshot() can be made idempotent.

This patch currently uses a simple LightWeightGSet for implementing the retry cache. It currently
does not support limiting the number of cache entries or ejecting entries in the cache based
on timeout (design proposed keeping entries for 10 minutes). The RetryCache should also move
common, so that it can be used by other server that may want to use it.

updatePipeLine can be made idempotent. If the oldBlock and newBlock are same we could consider
it as retried. Checking for node list in the existing block to the newNodes seems unnecessary.
But could be added as an additional check.

{code}
  public void updatePipeline(String clientName, ExtendedBlock oldBlock, 
      ExtendedBlock newBlock, DatanodeID[] newNodes)
      throws IOException;
{code}


                
> Implement retry cache on the namenode
> -------------------------------------
>
>                 Key: HDFS-4979
>                 URL: https://issues.apache.org/jira/browse/HDFS-4979
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: namenode
>            Reporter: Suresh Srinivas
>         Attachments: HDFS-4979.patch
>
>


--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message