Hi All, I've just posted a preview of the Spark 1.2.0. release for community regression testing.
Issues reported now will get close attention, so please help us test! You can help by running an existing Spark 1.X workload on this and reporting any regressions. As we start voting, etc, the bar for reported issues to hold the release will get higher and higher, so test early! The tag to be is v1.2.0-snapshot1 (commit 38c1fbd96) The release files, including signatures, digests, etc can be found at: http://people.apache.org/~pwendell/spark-1.2.0-snapshot1 The staging repository for this release can be found at: https://repository.apache.org/content/repositories/orgapachespark-1038/ The documentation corresponding to this release can be found at: http://people.apache.org/~pwendell/spark-1.2.0-snapshot1-docs/ == Notes == - Maven artifacts are published for both Scala 2.10 and 2.11. Binary distributions are not posted for Scala 2.11 yet, but will be posted soon. - There are two significant config default changes that users may want to revert if doing A:B testing against older versions. "spark.shuffle.manager" default has changed to "sort" (was "hash") "spark.shuffle.blockTransferService" default has changed to "netty" (was "nio") - This release contains a shuffle service for YARN. This jar is present in all Hadoop 2.X binary packages in "lib/spark-1.2.0-yarn-shuffle.jar" Cheers, Patrick --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@spark.apache.org For additional commands, e-mail: dev-h...@spark.apache.org
