Fail fast on terminal AKS machine failures

[karenychen]

Summary

• add ListMachines polling during machine readiness waits to fail early when expected machines terminally fail
• keep BatchPutMachine request sizing within the 50-machine service limit
• record successful_machines as a count instead of uploading machine names
• add focused unit coverage for timeout propagation, ListMachines pagination, terminal failure detection, and batch sizing

Validation

• python3 -m pytest modules/python/tests/test_aks_machine_client.py modules/python/tests/test_machine_crud.py modules/python/tests/test_crud_main.py -q
• python3 -m py_compile modules/python/clients/aks_machine_client.py modules/python/tests/test_aks_machine_client.py
• PYTHONPATH=modules/python python3 -m pylint modules/python/tests/test_aks_machine_client.py modules/python/clients/aks_machine_client.py modules/python/crud/azure/machine_crud.py modules/python/crud/main.py
• git diff --check

[Copilot]

Pull request overview

This PR improves AKS Machine-mode scale-out observability and failure behavior by failing readiness waits early when the AKS Machine resources reach terminal failure states, while also enforcing the 50-machine BatchPutMachine request limit and tightening recorded operation metadata.

Changes:

• Add Machine API ListMachines polling during node readiness waits to fail fast when expected machines are terminal and failed.
• Enforce BatchPutMachine sizing constraints (≤ 50 machines per request), including worker-based batch sizing validation.
• Record successful_machines as a count (not machine names) and add targeted unit tests for the new behaviors.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
modules/python/clients/aks_machine_client.py	Adds ListMachines pagination + terminal failure detection hooked into readiness polling; enforces 50-machine BatchPutMachine sizing; stores successful machine count metadata.
modules/python/tests/test_aks_machine_client.py	Updates metadata assertions and adds focused unit coverage for timeout propagation, ListMachines pagination, terminal failure detection, and batch sizing limits.