master_get_active_worker_nodes doesn't include coordinator

We don't include coordinator if it is added as a worker to the cluster in master_get_active_worker_nodes:

```
[local] talha@talha:9700-5673=# select * from pg_dist_node;
nodeid | groupid | nodename | nodeport | noderack | hasmetadata | isactive | noderole | nodecluster | m
etadatasynced | shouldhaveshards
--------+---------+-----------+----------+----------+-------------+----------+----------+-------------+--
--------------+------------------
1 | 1 | localhost | 9701 | default | f | t | primary | default | f
| t
2 | 2 | localhost | 9702 | default | f | t | primary | default | f
| t
35 | 0 | localhost | 9700 | default | t | t | primary | default | t
| t
(3 rows)

Time: 4.534 ms
[local] talha@talha:9700-5673=# select master_get_active_worker_nodes();
master_get_active_worker_nodes
--------------------------------
(localhost,9701)
(localhost,9702)
(2 rows)
```

This is also used by `run_command_on_workers` which means that `run_command_on_workers` will not be useful in single node and it will be misleading in case coordinator is added to the cluster when there are other nodes.

The fix is to change this line: https://github.com/citusdata/citus/blob/93764a3782591ff0c680121a9ae4ee398541eef7/src/backend/distributed/operations/node_protocol.c#L466