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
pve_memory_available_bytes is required because pve_memory_size_bytes shows the total memory size, and pve_memory_usage_bytes shows ALL used bytes, even the ones that are cached (which is a normal linux behavior), so this metric becomes useless for monitoring of free memory available
In pve_memory_available_bytes, I'd expect to see this:
As you can see it's only reporting 334M as free Memory, which is triggering all the alerts i have, because it thinkns all my RAM is used (while in reality 18G out of 31 are cached and will automatically release if some parts of the system require it)
The text was updated successfully, but these errors were encountered:
Metrics for nodes, VMs and containers are gathered via the PVE API. See the docs for all the values which are available.
You can also check what is available from the API with the pvesh command line tool.
Thus, if you find the value in the API, then we can maybe add it as a metric. Also I'd recommend to try node_exporter to scrape platform specific metrics.
pve_memory_available_bytes is required because pve_memory_size_bytes shows the total memory size, and pve_memory_usage_bytes shows ALL used bytes, even the ones that are cached (which is a normal linux behavior), so this metric becomes useless for monitoring of free memory available
In pve_memory_available_bytes, I'd expect to see this:
As you can see it's only reporting 334M as free Memory, which is triggering all the alerts i have, because it thinkns all my RAM is used (while in reality 18G out of 31 are cached and will automatically release if some parts of the system require it)
The text was updated successfully, but these errors were encountered: