Optimization for tracevirtcmdlat

[dzy176]

Hi, @Lyt99 , I have doubts about the metric calculation logic in tracevirtcmdlat.
The current logic is as follows：

if event.Latency > 100000000 {
	p.updateMetrics(VIRTCMD100MS)
} else {
	p.updateMetrics(VIRTCMD)
}

Would it be better to change it in this way? The VIRTCMD metric should contains the VIRTCMD100MS metric.

p.updateMetrics(VIRTCMD)

if event.Latency > 100000000 {
	p.updateMetrics(VIRTCMD100MS)
} 

[Lyt99]

Hi, thanks for your issue!

It's reasonable for me to make VIRTCMD contain VIRTCMD100MS, and so do other probes like socketlatency, kernellatnecy, etc.

If you are willing to, cloud you please take a look at these probes, and make a PR for it?

[dzy176]

PR #194, please review 😄😄😄😄