OS : Ubuntu 16.04
Spark : 2.0.2
JDK : java 8
After run this command " ./gradlew -Dspark.version=2.0.2 shadowJar" ,
Error message ::
fatal: Not a git repository (or any of the parent directories): .git
FAILURE: Build failed with an exception.
Process ‘command ‘git’’ finished with non-zero exit value 128
Run with --stacktrace option to get the stack trace. Run with --info or --debug option to get more log output.
We’d recommend using the precompiled binaries in the distributions linked here: https://hail.is/docs/stable/getting_started.html
This will eliminate the need for you to compile Hail at all! If you definitely need to compile yourself, we can work through the error.
I’m also getting this same git install error when I try to compile. The pre-built version works for me but I’m looking to run Hail on my own spark cluster and the Getting Started guide states:
“For all other Spark clusters, you will need to build Hail from the source code.”
So can I use the prebuilt package on my own Spark Cluster? Sorry I’m just a little confused, please explain…
That sentence is misleading! You can totally use the distribution to run with a cluster, as long as the Spark version matches. Just start following the directions below the line:
You can then open an IPython shell which can run Hail backed by the cluster with the ipython command.