Demonstrate how to use a development cluster (#37)

Note: this requires databricks/cli#1698.

---------

Co-authored-by: Andrew Nester <[email protected]>

knowledge_base/development_cluster/README.md

@@ -0,0 +1,24 @@
+# Development cluster
+
+This example demonstrates how to define and use a development (all-purpose) cluster in a Databricks Asset Bundle.
+
+This bundle defines an `example_job` which is run on a job cluster in production mode.
+
+For the development mode (default `dev` target) the job is overriden to use a development cluster which is provisioned 
+as part of the bundle deployment as well.
+
+For more information, please refer to the [documentation](https://docs.databricks.com/en/dev-tools/bundles/settings.html#clusters).
+
+## Prerequisites
+
+* Databricks CLI v0.229.0 or above
+
+## Usage
+
+Update the `host` field under `workspace` in `databricks.yml` to the Databricks workspace you wish to deploy to.
+
+Run `databricks bundle deploy` to deploy the job. It's deployed to `dev` target with a defined `development_cluster` cluster.
+
+Run `databricks bundle deploy -t prod` to deploy the job to prod target. It's deployed with a job cluster instead of development one.
+
+Run `databricks bundle run example_job` to run the job.

knowledge_base/development_cluster/databricks.yml

@@ -0,0 +1,41 @@
+bundle:
+  name: development_cluster
+
+include:
+  - resources/*.yml
+
+workspace:
+  host: https://e2-dogfood.staging.cloud.databricks.com
+
+targets:
+  dev:
+    mode: development
+    default: true
+
+    # By configuring this field for the "dev" target, all jobs in this bundle
+    # are overridden to use the all-purpose cluster defined below.
+    #
+    # This can increase the speed of development when iterating on code and job definitions,
+    # as you don't have to wait for job clusters to start for every job run.
+    #
+    # Note: make sure that the cluster configuration below matches the job cluster
+    # definition that will be used when deploying the other targets.
+    cluster_id: ${resources.clusters.development_cluster.id}
+
+    resources:
+      clusters:
+        development_cluster:
+          cluster_name: Development cluster
+          spark_version: 15.4.x-scala2.12
+          node_type_id: i3.xlarge
+          num_workers: 0
+          autotermination_minutes: 30
+          spark_conf:
+              "spark.databricks.cluster.profile": "singleNode"
+              "spark.master": "local[*, 4]"
+          custom_tags:
+              "ResourceClass": "SingleNode"
+
+  prod: {
+    # No overrides
+  }

knowledge_base/development_cluster/resources/example_job.yml

@@ -0,0 +1,22 @@
+resources:
+  jobs:
+    example_job:
+      name: "Example job to demonstrate using an interactive cluster for development"
+
+      tasks:
+        - task_key: notebook
+          job_cluster_key: cluster
+          notebook_task:
+            notebook_path: ../src/hello.py
+
+      job_clusters:
+        - job_cluster_key: cluster
+          new_cluster:
+            spark_version: 15.4.x-scala2.12
+            node_type_id: i3.xlarge
+            num_workers: 0
+            spark_conf:
+                "spark.databricks.cluster.profile": "singleNode"
+                "spark.master": "local[*, 4]"
+            custom_tags:
+                "ResourceClass": "SingleNode"

knowledge_base/development_cluster/src/hello.py

@@ -0,0 +1,3 @@
+# Databricks notebook source
+
+print("Hello, World!")