mixin: update units for disk and networking panels

[v-zhuravlev]

1. Update units of network and disk graphs

https://prometheus.io/docs/prometheus/latest/querying/functions/#rate
rate() calculates per-second average rate, therefore Bps units should be used for disks.
In networking bandwidth throughput is usually measured in bits/s so network graphs' units are changed accordingly.

2. Change io time units from seconds to %util

   When applying rate() to seconds we have 'seconds per second' or fractions of the second, so actually it actually can be from 0 to 1.

3. Update intervalFactor to 1 for better rates

[rgeyer]

LGTM.

Couple things to consider.

1. If you have access to the grafana fork, feel free to just make a feature branch here, rather than forking again to your own repo. If you don't have access, we can get that sorted.
2. Has this already been changed in the upstream mixin? If not, can you also offer this as a PR upstream (if you haven't already, I didn't check).

Thanks!

[v-zhuravlev]

Thanks Ryan for reviewing,

1. it is actually branch not fork, just prefixed by my name
2. it is opened but no activity there for some time: mixin: update units for disk and networking panels prometheus/node_exporter#2375
   therefore I want to have at least here so we can start using it in our integrations.

[rgeyer]

1. it is actually branch not fork, just prefixed by my name

Yup, sorry.. Too little coffee this morning, thanks!

2. it is opened but no activity there for some time: mixin: update units for disk and networking panels prometheus/node_exporter#2375
   therefore I want to have at least here so we can start using it in our integrations.

Perfectly reasonable, thanks for being on top of it!