[
https://issues.apache.org/jira/browse/HDFS-14809?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16920322#comment-16920322
]
KenCao commented on HDFS-14809:
-------------------------------
[~jojochuang] i will have a try. But i don't think i am a good coder, so it may
take me a long time to contribute an implementation. And i also need
confirmation of my analysis above from some hdfs committers. :)
> Make a new BlockReader for hdfs client lib
> -------------------------------------------
>
> Key: HDFS-14809
> URL: https://issues.apache.org/jira/browse/HDFS-14809
> Project: Hadoop HDFS
> Issue Type: New Feature
> Affects Versions: 2.6.0
> Reporter: KenCao
> Priority: Major
>
> as we known, the hdfs client java lib uses BlockReaderLocal for short circuit
> read by default, which allocate shared memory first, and make a slot within
> it. After all these steps, it will request the fds from the DataNode.
> However, the slot and shared memory sturcture is only used by DataNode when
> uncaching replicas, the client process can work well just with the fds asked
> later and it is nearly impossible to cache replicas in product environment.
> The api to release fds is called by client only with the slot given, the fds
> is close in the client process finally.
> so i think we can make a new BlockReader implementation which just requests
> the fds, and it will reduce the rpc calls from 3(allocate shm, request fds,
> release fds) to 1(request fds).
--
This message was sent by Atlassian Jira
(v8.3.2#803003)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
