Description

• The node could be in SchedulingDisabled status after reboot task if graceful_reboot: True in procedure.yaml

Fixes #654

Solution

• Changed "drain/uncordon" condition
  • Previously node was drained/uncordoned if (1) graceful reboot is required and (2) node is k8s node (not balancer)
  • Now node is drained/uncordoned if (1) graceful reboot is required and (2) node is k8s node (not balancer) and (3) node is NOT new (not from add_node) and (4) cluster already installed (not fresh install)
• Uncordon is now performed with retries, failure is not allowed

Test Cases

These changes affect following procedures:

• fresh install (on freshly rebuilt nodes)
• add_node (adding nodes of different roles balancer/worker/control-plane)
• upgrade (if packages are specified for install/upgrade/remove on upgrade)
• restore (at the end of restore)

Checklist

• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• There is no breaking changes, or migration patch is provided
• Integration CI passed
• Unit tests. If Yes list of new/changed tests with brief description
• There is no merge conflicts