[HDDS-10304] [DiskBalancer] Start command - Fix nodes not being processed and incorrect config values

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Component/s: Ozone CLI, SCM
• Labels: pull-request-available
• Assignee: Stephen O'Donnell
• Reporter: Stephen O'Donnell
• Created: 06/Feb/24 14:12
• Updated: 06/Feb/24 16:35
• Resolved: 06/Feb/24 16:35

Description

Fix various problems with the disk balancer start command:

1. When using -a for all datanodes, the hosts appear as an empty list due to the way the protobuf is deserialized, resulting in no hosts getting processed.

2. All the config values appear at zero, as if the proto value is not set for a double, it returns 0.0, so we need to check for has<field> and set an optional or empty

3. The isHealthy check needed negated, as it was excluding healthy nodes rather than unhealthy ones resulting in no nodes getting processed.

I also added a log message or two so we can have a record in the SCM log of the command being sent.

Issue Links

• links to: GitHub Pull Request #6181