clean up makefile

mango-jenkins/Dockerfile

@@ -28,9 +28,4 @@ ENV CHROME_BIN=/usr/bin/chromium
 # install node/npm
 RUN conda install -y nodejs
 
-
-# TODO: this was legacy code, when npm wasnt actually installed. does this do anything? is node even installed?
-# Maybe remove?
-# RUN ln -s /usr/bin/nodejs /usr/bin/node
-
 ENTRYPOINT ["${WORKSPACE}/scripts/jenkins-test"]

mango-jenkins/README.md

@@ -1,18 +1,31 @@
 This directory contains the docker images for
-[Mango](https://www.github.com/bigdatagenomics/mango). These containers build
-the mango codebase from HEAD, and then run the mango-submit scripts that submit
-the code/command to a running Spark cluster.
+[Mango](https://www.github.com/bigdatagenomics/mango-jenkins). This container contains
+all requirements for the [Mango](https://github.com/bigdatagenomics/mango) build
+for running with Jenkins.
 
 Running
 ===
 
-When the runtime container is run, a port should be opened for the Mango driver
-to communicate with the [Spark master](../apache-spark-master/README.md). The
-number of this port should be passed to mango as the Spark option
-`--conf spark.driver.port <portnumber>`.
+When the runtime container is run, the following environment variables should be
+specified:
+
+- SPARK_VERSION: Spark Version
+- HADOOP_VERSION: Hadoop version
+- COVERALLS_REPO_TOKEN: Token for Coveralls
+- WORKSPACE: Workspace where the current mango repository is pulled
+
 
 An example command to run mango Jenkins tests is:
 
 ```
-docker run bigdatagenomics/mango:latest
+
+docker run -v $WORKSPACE:${WORKSPACE} \
+	-e COVERALLS_REPO_TOKEN=${COVERALLS_REPO_TOKEN} \
+	-e WORKSPACE=${WORKSPACE} \
+	-e SPARK_VERSION=${SPARK_VERSION} \
+	-e HADOOP_VERSION=${HADOOP_VERSION} \
+	-e SCALAVER=${SCALAVER} \
+	--entrypoint=${WORKSPACE}/scripts/jenkins-test \
+	bigdatagenomics/mango-jenkins:latest
+
 ```