Skip to content

Version 0.0.16

Summary

Version 0.0.16 fixes the operational cluster dashboard panel so it reports the correct node totals when Proxmox cluster members exceed what the live /proxmox/cluster/status payload returns and when sibling ProxmoxNode rows are only discoverable by cluster name. It pairs with backend proxbox-api 0.0.12 — the entire fix lives in the plugin's dashboard data layer; no migration, no model change, no new persisted state.

It fixes one issue:

  • Issue #455: the Cluster Dashboard panel on the Proxmox-endpoint detail page showed incorrect online / offline / total node counts. Two cooperating bugs were responsible:
  • cluster_summary_from_node_rows unconditionally overrode the API-reported nodes_total / nodes_online with the count of locally rendered rows, so any time the live node row list was a strict subset of the cluster's real membership the total collapsed to whatever was rendered.
  • build_local_node_rows only consulted sibling ProxmoxNode records when both the cluster name and the scoped sibling-cluster-names set were populated. Endpoints whose siblings could only be matched by name (the common case for freshly imported clusters that have not yet linked to a NetBox Cluster object) silently dropped every sibling row.
  • There was no signal in the rendered node list for cluster members that existed in the cluster status payload but had not been synced yet — they simply vanished from the panel.

#455 — Cluster dashboard panel reports correct node totals

The fix is split across three small dashboard helpers in netbox_proxbox/views/dashboard_data.py plus one wiring change in netbox_proxbox/views/dashboard.py:

  • cluster_summary_from_node_rows preserves API totals (B1). When the API-reported nodes_total is greater than the count of live node rows we rendered, the helper now keeps the larger API total and the API's nodes_online count instead of silently overriding both. The override path remains in place only when the live rows are richer than the API summary (e.g. the API is incomplete), so existing behavior on healthy clusters is unchanged.
  • build_local_node_rows broadens sibling lookup by cluster name (B2). The DB sibling query now fires whenever cluster_name is set; the scoped_cluster_names filter is applied only when present. Freshly imported clusters that have not yet linked to a NetBox Cluster object surface every sibling row instead of going blank.
  • append_unsynced_node_placeholders renders pending members (B3). A new helper appends {"name": <member>, "status": "unknown"} placeholder rows for every cluster member named in the API status payload that did not match a synced ProxmoxNode. The dashboard panel now visibly distinguishes "synced and offline" from "not yet discovered".
  • Dashboard view wires the placeholders (B3 cont.). dashboard.py calls append_unsynced_node_placeholders between merge_node_rows and cluster_summary_from_node_rows so the summary recount sees the full membership.

Regression tests

Three new tests in tests/test_dashboard.py pin the fix:

  1. API-reported totals survive when the live node rows are a strict subset.
  2. Sibling rows are returned when only the cluster name is set (no scoped_cluster_names, no endpoint linkage).
  3. Placeholder rows are appended for members named by the API that have no matching local ProxmoxNode.

Compatibility

NetBox netbox-proxbox proxbox-api netbox-sdk proxmox-sdk
>=4.5.8 v0.0.16 v0.0.12 v0.0.8.post1 v0.0.3.post1
>=4.5.8 v0.0.15 v0.0.11 v0.0.8.post1 v0.0.3.post1

NetBox compatibility range: 4.5.8 – 4.6.99 (unchanged). Certified simultaneously against NetBox v4.5.8, v4.5.9, and official v4.6.0.

Upgrade Notes

  • No database migration. No model change. No OVERWRITE_FIELDS change.
  • Restart the NetBox WSGI process so the patched dashboard module is loaded.
  • The plugin pairs with proxbox-api 0.0.12; the dashboard panel fix does not depend on backend changes.