ARCHIVED This repository is being archived, to prevent any future confusion: All development on the Kubernetes scheduler back-end for Apache Spark is now upstream at https://spark.apache.org/ and https://github.com/apache/spark/
DEPRECATED. Work on this fork is discontinued. Further development is continuing on the mainline implementation of Apache Spark: https://github.com/apache/spark.
You can run Spark on Kubernetes using Spark 2.3. Some features from this work need to be ported to mainline. If a feature is missing, please check https://issues.apache.org/jira/projects/SPARK/issues to see if we're tracking that work, and if we are not, please file a JIRA ticket indicating the missing behavior.
All other bugs and feature requests should either be proposed through JIRA or sent to [email protected] or [email protected].
This repository, located at https://github.com/apache-spark-on-k8s/spark, contains a fork of Apache Spark that enables running Spark jobs natively on a Kubernetes cluster.
This is a collaboratively maintained project working on SPARK-18278. The goal is to bring native support for Spark to use Kubernetes as a cluster manager, in a fully supported way on par with the Spark Standalone, Mesos, and Apache YARN cluster managers.
- Usage guide shows how to run the code
- Development docs shows how to get set up for development
- Architecture docs shows the high level architecture of Spark on Kubernetes
- Code is primarily located in the resource-managers/kubernetes folder
Adding native integration for a new cluster manager is a large undertaking. If poorly executed, it could introduce bugs into Spark when run on other cluster managers, cause release blockers slowing down the overall Spark project, or require hotfixes which divert attention away from development towards managing additional releases. Any work this deep inside Spark needs to be done carefully to minimize the risk of those negative externalities.
At the same time, an increasing number of people from various companies and organizations desire to work together to natively run Spark on Kubernetes. The group needs a code repository, communication forum, issue tracking, and continuous integration, all in order to work together effectively on an open source product.
We've been asked by an Apache Spark Committer to work outside of the Apache infrastructure for a short period of time to allow this feature to be hardened and improved without creating risk for Apache Spark. The aim is to rapidly bring it to the point where it can be brought into the mainline Apache Spark repository for continued development within the Apache umbrella. If all goes well, this should be a short-lived fork rather than a long-lived one.
This is a collaborative effort by several folks from different companies who are interested in seeing this feature be successful. Companies active in this project include (alphabetically):
- Bloomberg
- Haiwen
- Hyperpilot
- Intel
- Palantir
- Pepperdata
- Red Hat
(original README below)
Spark is a fast and general cluster computing system for Big Data. It provides high-level APIs in Scala, Java, Python, and R, and an optimized engine that supports general computation graphs for data analysis. It also supports a rich set of higher-level tools including Spark SQL for SQL and DataFrames, MLlib for machine learning, GraphX for graph processing, and Spark Streaming for stream processing.
You can find the latest Spark documentation, including a programming guide, on the project web page. This README file only contains basic setup instructions.
Spark is built using Apache Maven. To build Spark and its example programs, run:
build/mvn -DskipTests clean package
(You do not need to do this if you downloaded a pre-built package.)
You can build Spark using more than one thread by using the -T option with Maven, see "Parallel builds in Maven 3". More detailed documentation is available from the project site, at "Building Spark".
For general development tips, including info on developing Spark using an IDE, see "Useful Developer Tools".
The easiest way to start using Spark is through the Scala shell:
./bin/spark-shell
Try the following command, which should return 1000:
scala> sc.parallelize(1 to 1000).count()
Alternatively, if you prefer Python, you can use the Python shell:
./bin/pyspark
And run the following command, which should also return 1000:
>>> sc.parallelize(range(1000)).count()
Spark also comes with several sample programs in the examples
directory.
To run one of them, use ./bin/run-example <class> [params]
. For example:
./bin/run-example SparkPi
will run the Pi example locally.
You can set the MASTER environment variable when running examples to submit
examples to a cluster. This can be a mesos:// or spark:// URL,
"yarn" to run on YARN, and "local" to run
locally with one thread, or "local[N]" to run locally with N threads. You
can also use an abbreviated class name if the class is in the examples
package. For instance:
MASTER=spark://host:7077 ./bin/run-example SparkPi
Many of the example programs print usage help if no params are given.
Testing first requires building Spark. Once Spark is built, tests can be run using:
./dev/run-tests
Please see the guidance on how to run tests for a module, or individual tests.
Spark uses the Hadoop core library to talk to HDFS and other Hadoop-supported storage systems. Because the protocols have changed in different versions of Hadoop, you must build Spark against the same version that your cluster runs.
Please refer to the build documentation at "Specifying the Hadoop Version" for detailed guidance on building for a particular distribution of Hadoop, including building for particular Hive and Hive Thriftserver distributions.
Please refer to the Configuration Guide in the online documentation for an overview on how to configure Spark.
Please review the Contribution to Spark guide for information on how to get started contributing to the project.