[
https://issues.apache.org/jira/browse/SPARK-1529?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16924940#comment-16924940
]
Zheng Shao commented on SPARK-1529:
-----------------------------------
For huge Spark jobs, we are seeing frequent failures due to the Shuffle phase:
local disk being full, unable to connect to External Shuffle Service.
The frustrating fact is that for these huge jobs, even a single task retry
sometimes takes a very long time.
It will be great to have the option to write the temporary shuffle data to a
reliable storage, e.g. HDFS but also others.
Maybe it should be a different Shuffle Manager implementation altogether.
Any thoughts?
> Support DFS based shuffle in addition to Netty shuffle
> ------------------------------------------------------
>
> Key: SPARK-1529
> URL: https://issues.apache.org/jira/browse/SPARK-1529
> Project: Spark
> Issue Type: Improvement
> Components: Spark Core
> Reporter: Patrick Wendell
> Assignee: Kannan Rajah
> Priority: Major
> Attachments: Spark Shuffle using HDFS.pdf
>
>
> In some environments, like with MapR, local volumes are accessed through the
> Hadoop filesystem interface. Shuffle is implemented by writing intermediate
> data to local disk and serving it to remote node using Netty as a transport
> mechanism. We want to provide an HDFS based shuffle such that data can be
> written to HDFS (instead of local disk) and served using HDFS API on the
> remote nodes. This could involve exposing a file system abstraction to Spark
> shuffle and have 2 modes of running it. In default mode, it will write to
> local disk and in the DFS mode, it will write to HDFS.
--
This message was sent by Atlassian Jira
(v8.3.2#803003)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
