HDDS-15628. [DiskBalancer] Preserve report output order for datanodes with the same density.

[slfan1989]

What changes were proposed in this pull request?

This PR makes DiskBalancer report output stable when multiple datanodes have the same currentVolumeDensitySum.

Previously, the report command collected successful datanode reports from ConcurrentHashMap.values() before sorting by density. Since ConcurrentHashMap does not preserve insertion order, datanodes with the same density could be displayed in an order different from the user-provided datanode order.

This patch builds the report list according to the successful datanode order instead. The existing density sort is stable, so datanodes with different density values are still sorted by currentVolumeDensitySum in descending order, while datanodes with the same density keep the input order.

What is the link to the Apache JIRA

JIRA: HDDS-15628. [DiskBalancer] Preserve report output order for datanodes with the same density.

How was this patch tested?

Add Junit Test.

For example, when running:

ozone admin datanode diskbalancer report host-2 host-1

and both datanodes have the same density, the output should keep host-2 before host-1.

Before the fix:

Report result:
Datanode: host-1 (127.0.0.1:19864)
Aggregate VolumeDataDensity: 14.09%

-------

Datanode: host-2 (127.0.0.1:19864)
Aggregate VolumeDataDensity: 14.09%

After the fix:

Report result:
Datanode: host-2 (127.0.0.1:19864)
Aggregate VolumeDataDensity: 14.09%

-------

Datanode: host-1 (127.0.0.1:19864)
Aggregate VolumeDataDensity: 14.09%

[adoroszlai]

Thanks @slfan1989 for the patch.

[slfan1989]

Thanks @slfan1989 for the patch.

@adoroszlai Thank you very much for helping review the code!