Obliquely related to this. I’m graphing Mimosa’s TX and RX throughput. The values they return via SNMP are kbps * 100. So we get bps / 10 and truncated. Useful? Not so much.
If I just divide by 100, I get “kbps” so I set the label to that. When we’re passing 29 Mbps of traffic, with auto scale on the graph shows “29 Kkbps”. If I disable autoscale I get “29000 kbps”. Which is accurate, but takes another half second to parse for someone who knows what to expect.
For the best results with Mimosa’s design choice, I probably need to multiply by 10, set the label to “bps” and let the graph autoscale logic do it’s thing.
BTW, I now have a template for backhauls which graph the throughput, all 4 TX and RX PHY chains, PER streams, and EVM streams as well as getting the version, GPS, last reboot, and device name information. Sonar, feel free to snarf it and provide it to everyone if you don’t already have a better template built.