GitHub user SaintBacchus opened a pull request:
https://github.com/apache/spark/pull/3509
[Shuflle][SPARK-4641]file not found excepton in hash shuffle
It's very hard to replay this problem but actually I have meet this.
Assuming that, some hash key bucket have no elems in it and in the code of
HashShuffleWriter.write:
for (elem <- iter) {
val bucketId = dep.partitioner.getPartition(elem._1)
shuffle.writers(bucketId).write(elem)
}
The bucket will never call the 'write' funciton and will never create the
file(create file in first writer and file.length never check file is exist or
not).
But the shuffle fetch will also fetch the unexist file, and causes the
problem.
I have tested the problem described in SPARK-4641.
Sort Shuffle will also throw "snappy (parsing_error(2))" but can not be
resolved by this modify.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/SaintBacchus/spark FileNotFound
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/spark/pull/3509.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #3509
----
commit 5846a40f3cbc7a31be9c5cf210236bcd5e799b36
Author: carlmartin <[email protected]>
Date: 2014-11-28T12:45:23Z
file not found in hash shuffle
----
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
