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
Resources overview in the dashboard page is wrong if arcane/agent deployed inside LXC container.
Steps To Reproduce
Create unprivileged LXC container with nesting feature enabled.
Limit container CPU/RAM resources
Bind-mount some additional external storage to the container not in the root (/) but somewhere else.
Make sure, that CPU binding inside LXC container have gaps/not in order (In my case it has CPU1 and CPU7 assigned, probably you'll have to recreate LXC several times in a row to achieve this)
Expected Behavior
It shows current CPU/RAM resources respecting LXC cgroups limits.
Disk usage show values based on system drive (with the / mountpoint).
Actual Behavior
It shows total host system memory value instead limited by cgroup.
It shows total CPU cores based on highest bound CPU number from the host, ignoring actual cgroup limits and core binding.
It shows disk usage of bind-mount instead of local root fs.
Screenshots
It currently show this:
When actual resources it have are 2 CPUs, 2GB of RAM and 10Gb of storage in the root fs.
By the way, at the same time, this window shows the correct one info
Activity
kmendell commentedon Apr 23, 2026
should be fixed in the next version .
Manriel commentedon Apr 23, 2026
@kmendell, It does not.
Still the same issue happened in 1.18.1
theoraim commentedon Jul 1, 2026
Hello, having the same issue here : LXC Container with 2CPUs, 4GB. Version is 2.3.0
The details windows display the correct info :