Obsidians still seems to watch just for /dev/sd[x] drives for having metrics gathered in Grafana.
Example from one of my hosts where NVMe drives are totally ignored, while the metrics are delivered by the kernel for the drives:
Read more
Example from one of my hosts where NVMe drives are totally ignored, while the metrics are delivered by the kernel for the drives:
Bash:
# cat /proc/diskstats
259 0 nvme0n1 11244539 41601 1166709299 2534288 41946913 16638180 4583151873 83687872 0 3518900 86257601 38481 5 3539397224 35440 0 0
259 2 nvme0n1p1 179 31 6226 21 5 0 12 0 0 104 110 121 1 3808256 88 0 0
259 3...
Read more