knowledge_base: add bind_existing_resources example

knowledge_base/bind_existing_resources/.gitignore

@@ -0,0 +1 @@
+.databricks

knowledge_base/bind_existing_resources/README.md

@@ -0,0 +1,38 @@
+# Bind existing resources
+
+This Declarative Automation Bundles example puts an existing job and dashboard under bundle management.
+It generates the configuration, binds it to the workspace resources, deploys it, and then unbinds it.
+
+For more information, see the [bind documentation](https://docs.databricks.com/dev-tools/bundles/migrate-resources.html).
+
+## Prerequisites
+
+* Databricks CLI v1.20.0 or above
+* A workspace with serverless jobs and a SQL warehouse named `Serverless Starter Warehouse`.
+* An existing job and an existing AI/BI dashboard. Copy their IDs from the workspace URLs.
+
+## Usage
+
+1. Set the folder of the existing dashboard: `export BUNDLE_VAR_dashboard_parent_path=/Workspace/Users/<user>/<folder>`
+2. Generate the job configuration. This replaces the sample files in `resources/` and `src/`:
+   `databricks bundle generate job --existing-job-id <job-id> --key existing_job --force --profile <profile>`
+3. Generate the dashboard configuration:
+   `databricks bundle generate dashboard --existing-id <dashboard-id> --key existing_dashboard --force --profile <profile>`
+4. In `resources/existing_dashboard.dashboard.yml`, replace the literal `warehouse_id` with `${var.warehouse_id}`.
+5. In the same file, set `parent_path: ${var.dashboard_parent_path}`.
+6. Bind the job: `databricks bundle deployment bind existing_job <job-id> -t prod --profile <profile>`
+7. Bind the dashboard: `databricks bundle deployment bind existing_dashboard <dashboard-id> -t prod --profile <profile>`
+8. Show the changes that the deploy makes: `databricks bundle plan -t prod --profile <profile>`
+9. Deploy the bundle: `databricks bundle deploy -t prod --profile <profile>`
+10. Run the job: `databricks bundle run existing_job -t prod --profile <profile>`
+11. Unbind the job: `databricks bundle deployment unbind existing_job -t prod --profile <profile>`
+12. Unbind the dashboard: `databricks bundle deployment unbind existing_dashboard -t prod --profile <profile>`
+13. Destroy the bundle deployment: `databricks bundle destroy -t prod --profile <profile>`
+
+| Step | Effect on the workspace resource |
+| --- | --- |
+| `bind` | No change. The bundle state records the resource ID. |
+| `plan` | Shows drift. Changes made in the UI appear as `update`. |
+| `deploy` | Overwrites the resource with the bundle configuration. The ID stays the same. |
+| `unbind` | No change. The bundle state forgets the resource ID. |
+| `destroy` | Deletes 0 resources and the `root_path` folder. The unbound job and dashboard stay. |

knowledge_base/bind_existing_resources/databricks.yml

@@ -0,0 +1,31 @@
+# This is a Declarative Automation Bundle definition for bind_existing_resources.
+# See https://docs.databricks.com/dev-tools/bundles/migrate-resources.html for documentation.
+bundle:
+  name: bind_existing_resources
+
+include:
+  - resources/*.yml
+
+variables:
+  warehouse_id:
+    description: SQL warehouse that runs the dashboard queries
+    lookup:
+      warehouse: Serverless Starter Warehouse
+  dashboard_parent_path:
+    description: Workspace folder of the existing dashboard, for example /Workspace/Users/<user>/<folder>
+
+# The workspace host is not set here. Use a profile (--profile / -p) or the
+# DATABRICKS_HOST environment variable.
+
+targets:
+  dev:
+    # Development mode adds a "[dev <user>]" prefix to the bound job name and pauses its schedules.
+    mode: development
+    default: true
+    workspace:
+      root_path: /Workspace/bundle-examples/${bundle.name}/${bundle.target}/${workspace.current_user.short_name}
+
+  prod:
+    mode: production
+    workspace:
+      root_path: /Workspace/bundle-examples/${bundle.name}/${bundle.target}

knowledge_base/bind_existing_resources/resources/existing_dashboard.dashboard.yml

@@ -0,0 +1,9 @@
+resources:
+  dashboards:
+    existing_dashboard:
+      display_name: "Existing dashboard"
+      # Generated as a literal warehouse ID. Replaced with a variable so the bundle works in any workspace.
+      warehouse_id: ${var.warehouse_id}
+      # The parent path cannot change. Keep the folder of the existing dashboard, or deploy recreates it with a new ID.
+      parent_path: ${var.dashboard_parent_path}
+      file_path: ../src/existing_dashboard.lvdash.json

knowledge_base/bind_existing_resources/resources/existing_job.job.yml

@@ -0,0 +1,24 @@
+resources:
+  jobs:
+    existing_job:
+      name: Existing job
+      tasks:
+        - task_key: hello
+          disabled: false
+          email_notifications: {}
+          environment_key: default
+          notebook_task:
+            notebook_path: ../src/hello.py
+            source: WORKSPACE
+          run_if: ALL_SUCCESS
+          timeout_seconds: 0
+      email_notifications: {}
+      environments:
+        - environment_key: default
+          spec:
+            environment_version: "6"
+      max_concurrent_runs: 1
+      queue:
+        enabled: true
+      timeout_seconds: 0
+      webhook_notifications: {}

knowledge_base/bind_existing_resources/src/existing_dashboard.lvdash.json

@@ -0,0 +1,18 @@
+{
+  "datasets": [
+    {
+      "displayName": "trips",
+      "name": "d1",
+      "queryLines": [
+        "SELECT * FROM samples.nyctaxi.trips LIMIT 10"
+      ]
+    }
+  ],
+  "pages": [
+    {
+      "displayName": "Trips",
+      "name": "p1",
+      "pageType": "PAGE_TYPE_CANVAS"
+    }
+  ]
+}

knowledge_base/bind_existing_resources/src/hello.py

@@ -0,0 +1,2 @@
+# Databricks notebook source
+print("Hello from an existing job!")