Github user ehiggs commented on the pull request:
https://github.com/apache/spark/pull/4204#issuecomment-71448742
Well, I could submit a patch to hadoop-common to do the sort in
`o.a.h.fs.FileSystem`, and that would fix it basically everywhere. However, I
had assumed that hadoop doesn't always care about ordering, hence this low
footprint patch.
>You could imagine this is an issue for something backed by, say, HBase.
I haven't used HBase in anger or looked the code, so I can't really say if
there is a bug there. FWIW, [according to the mailing
list](http://apache-spark-developers-list.1001551.n3.nabble.com/RDD-order-guarantees-tp10142p10147.html),
Spark relies on HDFS returning file parts in an order that appears to work.
Perhaps HBase gets away with something similar and HBase on Lustre would have
some difficulty. It's an interesting question.
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
