You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
VDO is a kernel subsystem that provides compression and deduplication to a Linux device. This project provides a Python2 based Prometheus exporter
for VDO volume statistics. Once exported, you may visualize the data and define alerts within Prometheus itself, or
define dashboards and alerts in systems like Grafana.
Background
Writing an exporter for Prometheus would normally entail using the prometheus-client Python package, but at the time of writing
the rpm for this wasn't available for CentOS or RHEL. So this project just uses standard Python features to implement
the exporter (a Prometheus client is just a HTTP endpoint anyway!)
I've tested against vm's with the following configurations;
NB.
if you're interested in bringing the VDO stats together with Ceph, you should use DNS names instead of IP addresses within the scrape definition. This will allow your promQL to 'match' across metrics based on "exported_instance" and "instance".
Grafana Configuration
I've included a dashboard to show some of the stats the exporter provides. Just import the
the 'VDO_Information.json' dashboard from the grafana subdirectory into Grafana, then open the dashboard.