The Hoof & Paw
DocsCategoriesTagsView the current conditions from the WolfspyreLabs WeatherstationToggle Dark/Light/Auto modeToggle Dark/Light/Auto modeToggle Dark/Light/Auto modeBack to homepage

Ceph Device metrics in Proxmox

OOH! I can get this thing to visually display SMART data? neat!

There’s some interesting shit one can do with ceph’s dashboard… Not that there aren’t multiple ways to represent this data, but I wanted to see how easy it’d be to get it represented… Fortunately, the Ceph Docs1 are pretty good thus far.

There were a few things I had to install to get the monitoring components rolling… and it’s entirely possible there are more still; nevertheless, it appears that there’s two “simple” ways to get devices’ smart data polled a few times

Herd-of-ants style

with an open session on each ceph host; you can parallelize the inspection and have it run a bit faster:

collecting ceph device metrics from each hostLang: shell
for D in `ceph device ls-by-host \`hostname\` \
  |awk '{print $3}'|egrep -v DAEMONS`; do
  echo "`date` ${D}";
  ceph device scrape-daemon-health-metrics ${D};
done

One-Liner to get them all

collecting ceph device metrics from each hostLang: shell
ceph device scrape-health-metrics