enh: Default cluster zone_id should be 'auto'

[dmoore247]

Describe the issue

The create cluster command failed to create the cluster because the instance type wasn't available in zone us-east-1

The default zone should be auto (or the enhanced auto) so the command is more likely to succeed. (overall, there should be more 'good' defaults)

Steps to reproduce the behavior

Run

databricks clusters create "13.3.x-scala2.12" \
--data-security-mode LEGACY_TABLE_ACL \
--cluster-name platform-ucx-tacl2 \
--num-workers 1 \
--enable-elastic-disk \
--node-type-id "m5.xlarge"

The command failed because the instance type wasn't available in us-east-1. Note there are no command line options to specify zone, nothing in the --help to indicate how

Expected Behavior

The create cluster command should default to zone_id = auto

Actual Behavior

Finding instances for new nodes, acquiring more instances if necessaryError: failed to reach RUNNING, got TERMINATED: The specified request is not supported. aws_api_error_code: Unsupported, aws_error_message: Your requested instance type (m5.xlarge) is not supported in your requested Availability Zone (us-east-1...

OS and CLI version

Databricks CLI v0.208.2 MacOS

Is this a regression?

Don't think so

Debug Logs

This is a feature request

[dmoore247]

While were here... I was forced to include all those other options. It would be good that this command (databricks create cluster) and required no other options (maybe give an advisement) Creating cluster "Douglas Moore's cluster" as a Single node ....