workflows

Lab: Workflows

Ensure that the LAB_ROOT environment variable points to the exercise's root directory by executing:

export LAB_ROOT=~/cloudify-training-labs/workflows/exercise

Part I: execute_operation

$LAB_ROOT/blueprint/blueprint.yaml is a very simple blueprint, defining a node called lab. Your task consists of:

1. Adding an operation to the lab node
2. Binding the operation to the included script, scripts/my-logging-operation.sh
3. Using the execute_operation workflow to invoke the script

Step 1: Replace placeholders

Grep the blueprint for REPLACE_WITH. Replace that string with an operation definition.

Step 2: Test your changes

Install the blueprint on the manager:

cfy install $LAB_ROOT/blueprint/blueprint.yaml -b workflows -d workflows -i ip=<your-app-vm-ip>

Prepare a file, ~/execution-parameters.yaml, containing parameters to pass to the execute_operation workflow, such as:

• operation
• node_ids
• etc.

Run the execute_operation workflow, providing the ~/execution-parameters.yaml file as an input:

cfy executions start execute_operation -d workflows -p ~/execution-parameters.yaml

Tip: Use the execute_operation workflow documentation.

Part II: heal

In this part, we will demonstrate the heal workflow.

First, we need to find the instance ID of the node we would like to heal. Remember: the heal workflow uninstalls, and then reinstalls, the entire Compute node containing the node we wish to heal; therefore, you may either look for the instance ID of the Compute node itself, or of any node which is contained in (directly or indirectly) that Compute node.

To find the node instance, execute the following command (from the CLI VM):

cfy node-instances list -d workflows

The output shows all node instances in the topology. For each node instance, both the instance ID and the node ID are shown. The node ID is the name of the node in the blueprint; the node instance ID is the ID of the particular instance of that node.

Then, execute the heal workflow. For example:

cfy executions start heal -d workflows -p node_instance_id=lab_vm_a644a

Part III: Uninstall

Use the commands learned in previous labs to uninstall the lab application.