Spark in other clusters?
After experiencing some issues with GATK 3.2 on CentOS 7.4 (AVX changes broke everything), we've installed the newer versions - 3.8.
I thought I'd install 4 beta to test, but am confused by all this talk of SPARK.
We run a SLURM cluster. Do I need to install SPARK to run on top of that in order to get the advantages, or is SPARK not strictly necessary for running on a cluster?
That being the case - can we delete the gatk-package-4.beta.5-spark.jar?
(to my shame I found the answer in this tutorial after asking the question)