• Open Cloud Shell
• Run the following to clone the repo and change directories

  cd ~
  git clone https://github.com/roitraining/gcp-demos.git
  cd gcp-demos

Simple demos used to illustrate various BigQuery techniques

• github_demo is useful for walking people through array functions and searching for rows that have a given value in an array column
• hackernews_demo is useful for walking people through array work and window functions
• scds shows some ways of updating nested tables when a dimension changes. This should be fleshed out with other techniques.
• wiki_queries is just the simple wiki demo query

Compised of the bq-do-it-nows folder and the docs folder, this is a collection of 20+ quick hands-on activities that can be done as demos of by students.

Students access the activities using https://roitraining.github.io/gcp-demos/#0

Many require setup using the bq-schema-demo resources.

This can be used to demonstrate the impact of schema design choices. The datasets generated by the code in this folder is also used in the Do-It-Nows. For more info, see the README in the bq-schema-demo folder.

This can be used to demonstrate how to write a streaming pipeline in Beam using the Python SDK, as well as how to run and test in Dataflow. Code is pretty clean and easy to follow. Found in the dflow-bq-stream-python folder.

1. Open application at https://roi-gcp-demos.appspot.com/dlp_demo
2. Enter text with no sensitive data into left pane; see results on right
3. Enter sensitive data in left pane; see results on right
4. Fiddle around with contextual info on left, see ratings change on right
5. Demonstrate different remediation tactics with blue buttons
6. Optional - show source code to class (found in dlp-demo-app folder)

• streamlist_chat_simple offers a PaLM-based chat bot application that’s easy to understand and easy to deploy

• Dataproc scaling demo
  1. Open dataproc_scale_demo.sh
  2. Use first command in Cloud Shell to create the cluster
  3. Use second set of commands in cloud Shell to submit the job
  4. Show the slow progress of the job
  5. Open a 2nd Cloud Shell session, and use the third command to add nodes to the cluster, changing the number of nodes to fit within your quotas.
  6. Show the new nodes. Show the improved rate of progress
  7. Run the last command to tear down the cluster