Having problems with --query_network_latency_meters

I just deployed a new scaleio fabric on vmware, using 2.0.13000_211, over vmware esxi

Having trouble with the scli command --query_network_latency_meters

The output shows the latency to itself (see below), is the display just showing the wrong addresses, and is the data accurate.
I was expecting to run this command on SDS .18, and see the output for SDS .18 to both SDS .20 and .21.

Anymore insight into the output, is that a 30 second average latency? How it is calculated

ScaleIO-172-24-99-9:~ # scli --query_network_latency_meters --sds_ip 1.1.1.18

SDS with IP 1.1.1.18 (port 7072) returned information on 2 SDSs

SDS 1.1.1.18

  Average IO size: 13.0 KB (13312 Bytes)

  Average latency (micro seconds): 1304

SDS 1.1.1.18

  Average IO size: 3.0 KB (3072 Bytes)

  Average latency (micro seconds): 1071

ScaleIO-172-24-99-9:~ # scli --query_network_latency_meters --sds_ip 1.1.1.20

SDS with IP 1.1.1.20 (port 7072) returned information on 2 SDSs

SDS 1.1.1.20

  Average IO size: 7.0 KB (7168 Bytes)

  Average latency (micro seconds): 757

SDS 1.1.1.20

  Average IO size: 4.0 KB (4096 Bytes)

  Average latency (micro seconds): 2218

ScaleIO-172-24-99-9:~ # scli --query_network_latency_meters --sds_ip 1.1.1.21

SDS with IP 1.1.1.21 (port 7072) returned information on 2 SDSs

SDS 1.1.1.21

  Average IO size: 1.0 KB (1024 Bytes)

  Average latency (micro seconds): 932

SDS 1.1.1.21

  Average IO size: 4.0 KB (4096 Bytes)

  Average latency (micro seconds): 1631

Thanks Simon