Do not try to install the jupyter scala kernel if it's already present.

tasks/main.yml

@@ -64,8 +64,14 @@
   block:
     - name: Download the coursier executable
       get_url: url=https://git.io/coursier-cli-linux dest=/usr/local/bin/coursier owner=root group=root mode='0555'
+
+    - name: Check if the scala support for jupyter is already installed
+      stat: path="/root/.local/share/jupyter/kernels/scala"
+      register: jupyter_scala
+
     - name: Integrate Spark and Scala
       command: /usr/local/bin/coursier launch --fork almond:0.5.0 --scala {{ jupyter_spark_scala_version }} -- --install
+      when: not jupyter_scala.stat.exists
 
   when:
     - jupyter_is_hadoop_node