Resolution for mqtt.message.pub_to_sub.latency.*

[KrlosWd]

Hello,

After working with mzbench + vmq_mzbench for a while, I noticed that the resolution for mqtt.message.pub_to_sub.latency.50, mqtt.message.pub_to_sub.latency.90, mqtt.message.pub_to_sub.latency.99, mqtt.message.pub_to_sub.latency.999 and mqtt.message.pub_to_sub.latency.mean is reduced due to the fact of using an exponential notation to save the data (e.g. 1.23e4). Is that something that we can configure? otherwise I think it would be a good improvement to save those values with as much resolution as possible.

Thanks in advance for your attention/help!