HDDS-14195. Do not send DiskBalancer start command to DN with active run

[Gargi-jais11]

What changes were proposed in this pull request?

There are few issues with usability and readability of diskbalancer cli output:

Firstly, used set instead of list to store dnHostName for cli output so that it doesn't display duplicate hostnames if user give any diskbalancer command for same hostname twice like below:

bash-5.1$ ozone admin datanode diskbalancer status ozone-datanode-1 ozone-datanode-3 ozone-datanode-1 --json
[ {
  "datanode" : "ozone-datanode-1.ozone_default",
  "action" : "status",
  "status" : "success",
  "serviceStatus" : "RUNNING",
  "threshold" : 10.0,
  "bandwidthInMB" : 10,
  "threads" : 5,
  "stopAfterDiskEven" : false,
  "successMove" : 0,
  "failureMove" : 0,
  "bytesMovedMB" : 0,
  "estBytesToMoveMB" : 0,
  "estTimeLeftMin" : 0
}, {
  "datanode" : "ozone-datanode-3.ozone_default",
  "action" : "status",
  "status" : "success",
  "serviceStatus" : "RUNNING",
  "threshold" : 10.0,
  "bandwidthInMB" : 10,
  "threads" : 5,
  "stopAfterDiskEven" : false,
  "successMove" : 0,
  "failureMove" : 0,
  "bytesMovedMB" : 0,
  "estBytesToMoveMB" : 0,
  "estTimeLeftMin" : 0
}, {
"datanode" : "ozone-datanode-1.ozone_default",
  "action" : "status",
  "status" : "success",
  "serviceStatus" : "RUNNING",
  "threshold" : 10.0,
  "bandwidthInMB" : 10,
  "threads" : 5,
  "stopAfterDiskEven" : false,
  "successMove" : 0,
  "failureMove" : 0,
  "bytesMovedMB" : 0,
  "estBytesToMoveMB" : 0,
  "estTimeLeftMin" : 0
} ]

Secondly, Instead of showing ip-address in json cli output show dnHostName for better clarity.

bash-5.1$ ozone admin datanode diskbalancer stop --in-service-datanodes --json
[ {
  "datanode" : "172.18.0.5:19864",
  "action" : "stop",
  "status" : "success"
}, {
  "datanode" : "172.18.0.9:19864",
  "action" : "stop",
  "status" : "success"
}, {
  "datanode" : "172.18.0.10:19864",
  "action" : "stop",
  "status" : "success"
}, {
  "datanode" : "172.18.0.8:19864",
  "action" : "stop",
  "status" : "success"
}]

What is the link to the Apache JIRA

https://issues.apache.org/jira/browse/HDDS-14195

How was this patch tested?

Added new test cases to TestDiskBalancerProtocolServer and update TestDiskBalancerSubCommands, TestDiskBalancer.
Tested locally:

[ChenSammi]

Please do not do the filtering.

It's ok to send the start command again to a DN which has disk balancer running, and doesn't response with "*** is already running". It's called API idempotent.

Further more, filter DN at client side is inaccurate. There is chance that DN changes to STOPPED state after return the status.

[Gargi-jais11]

Closing this PR, will open a new one as it is a kind of messed up.