mirror of
https://github.com/rcourtman/Pulse.git
synced 2026-05-02 13:30:13 +00:00
7c895df1f3
Proxmox VE 9.x removed support for the "full" parameter in the
/nodes/{node}/qemu/{vmid}/status/current endpoint. When Pulse sent
GetVMStatus() requests with ?full=1, Proxmox responded with:
API error 400: {"errors":{"full":"property is not defined in schema..."}}
This caused the cluster client to mark ALL endpoints as unhealthy, which
cascaded into multiple failures:
- VM status checks failed
- Guest agent queries were blocked
- Filesystem data collection stopped working
- All Windows VMs showed disk:-1 (unknown) instead of actual disk usage
The fix removes the ?full=1 parameter since Proxmox 9.x returns all data
by default without needing this parameter. This maintains backward
compatibility with older Proxmox versions while fixing the issue in 9.x.
After this fix:
- Cluster endpoints are correctly marked as healthy
- Guest agent queries work properly
- Windows VMs report actual disk usage (e.g., 26% on C:\ drive)
- VM monitoring functions normally on Proxmox 9.x
|
||
|---|---|---|
| .. | ||
| ceph.go | ||
| client.go | ||
| client_test.go | ||
| cluster_client.go | ||
| cluster_client_test.go | ||
| replication.go | ||
| zfs.go | ||
| zfs_test.go | ||