Errors when syncing virutal disks

[thel1988]

The same error from #157 seems to still be at large here, when syncing virtual-disks the following error occours:

proxbox [2026-05-26 08:21:12,762] [ERROR] virtual_disks: Error getting VM config for Ubuntu22.04Template: VM Config not found. | VM Config not found. Check if the 'node', 'type', and 'vmid' are correct. Session errors: Error fetching Proxmox VM config
proxbox [2026-05-26 08:21:12,762] [WARNING] virtual_disks: Could not get VM config for VM Ubuntu22.04Template (vmid: 2000)
proxbox [2026-05-26 08:21:16,715] [ERROR] virtual_disks: Error getting VM config for Ubuntu24.04Template: VM Config not found. | VM Config not found. Check if the 'node', 'type', and 'vmid' are correct. Session errors: Error fetching Proxmox VM config
proxbox [2026-05-26 08:21:16,715] [WARNING] virtual_disks: Could not get VM config for VM Ubuntu24.04Template (vmid: 2001)
proxbox [2026-05-26 08:21:17,290] [INFO] virtual_disks: Virtual disks sync complete: {'count': 132, 'created': 0, 'updated': 37, 'skipped': 95}

running versions:
proxbox: 0.0.15-nginx
netbox: 4.6.1
netbox-proxbox: 0.0.18

[thel1988]

Here to add more context from the more debug logs:

proxbox [2026-05-26 08:47:13,568] [DEBUG] exception: ProxboxException: Error fetching Proxmox VM config
 > Python Exception: HTTP 596 error:0A000086:SSL routines::certificate verify failed: None
proxbox [2026-05-26 08:47:13,568] [DEBUG] exception: ProxboxException: VM Config not found.
 > Detail: VM Config not found. Check if the 'node', 'type', and 'vmid' are correct. Session errors: Error fetching Proxmox VM config
proxbox [2026-05-26 08:47:13,568] [ERROR] virtual_disks: Error getting VM config for exam01.example.com: VM Config not found. | VM Config not found. Check if the 'node', 'type', and 'vmid' are correct. Session errors: Error fetching Proxmox VM config
proxbox [2026-05-26 08:47:13,569] [WARNING] virtual_disks: Could not get VM config for VM cbproxy01.synack.dk (vmid: 2009)

[thel1988]

Also this error also occours when debug is enabled:

proxbox [2026-05-26 08:51:47,330] [DEBUG] exception: ProxboxException: NetBox REST request failed
 > Detail: {"disk":["The specified disk size (3212288) must match the aggregate size of assigned virtual disks (1048576)."]}
proxbox [2026-05-26 08:51:47,338] [DEBUG] sync_vm: VM deps cluster=<proxbox_api.netbox_rest.RestRecord object at 0x707a3ab88b40> device=<proxbox_api.netbox_rest.RestRecord object at 0x707a3ab89240> role=<proxbox_api.netbox_rest.RestRecord object at 0x707a04d12350>

[thel1988]

Another error message:

proxbox [2026-05-26 08:58:18,772] [DEBUG] exception: ProxboxException: Error fetching Proxmox VM config
 > Python Exception: HTTP 500 proxy loop detected: proxy loop detected
proxbox [2026-05-26 08:58:18,772] [DEBUG] exception: ProxboxException: VM Config not found.
 > Detail: VM Config not found. Check if the 'node', 'type', and 'vmid' are correct. Session errors: Error fetching Proxmox VM config
proxbox [2026-05-26 08:58:18,772] [ERROR] virtual_disks: Error getting VM config for pritunl-vpnnode01.trifork.security: VM Config not found. | VM Config not found. Check if the 'node', 'type', and 'vmid' are correct. Session errors: Error fetching Proxmox VM config

[emersonfelipesp]

Completed

Implemented and merged the virtual disk sync fix for issue #211.

Merged PRs:

• proxbox-api: Fix virtual disk VM config node resolution #213 -> main, merge commit 1fba522
• netbox-proxbox: Forward Proxmox endpoint tuning to backend netbox-proxbox#539 -> develop, merge commit 0682870b2ce2f34252851b012cde0f7f5907cf19

What changed:

• proxbox-api now resolves the Proxmox VM config target for virtual disk sync from live cluster resource VMID/type data before falling back to NetBox-side placement fields. This avoids using stale or FQDN NetBox VM device names when calling the Proxmox VM config endpoint.
• The fallback order is now cluster resources -> custom_fields.proxmox_node -> NetBox VM device.name.
• VM config resolution now preserves the underlying session error detail, including Proxmox proxy-loop and TLS verification errors, instead of collapsing everything to the generic "Error fetching Proxmox VM config" message.
• Virtual disk sync still skips only the affected VM when config cannot be fetched, and now logs VM name, VMID, type, cluster, selected node, and node source.
• proxbox-api endpoint CRUD/public schemas now expose the persisted Proxmox timeout, max_retries, and retry_backoff fields.
• netbox-proxbox now forwards verify_ssl, timeout, max_retries, and retry_backoff when syncing Proxmox endpoint records into proxbox-api before live reads.
• Scoped agent docs were updated for the virtual disk target-resolution behavior and plugin backend endpoint payload behavior.

Validation:

• proxbox-api: uv sync --extra test --group dev
• proxbox-api: uv run python -m pytest tests/test_virtual_disks_sync.py tests/test_session_and_helpers.py tests/test_api_routes.py tests/test_endpoint_crud.py -> 115 passed in 517.85s
• proxbox-api: uv run python -m pytest tests/test_virtual_disks_sync.py -> 11 passed in 0.08s after the final helper refactor
• proxbox-api: uv run ruff check on all touched backend/test files -> passed
• proxbox-api: uv run ruff format --check on all touched backend/test files -> passed
• netbox-proxbox: uv run pytest tests/test_backend_sync_placement.py -> 1 passed
• netbox-proxbox: uv run ruff check netbox_proxbox/views/backend_sync.py tests/test_backend_sync_placement.py -> passed
• netbox-proxbox: uv run ruff format --check netbox_proxbox/views/backend_sync.py tests/test_backend_sync_placement.py -> passed

GitHub reported no status checks on either PR branch before merge; both PRs were merge-clean and were merged with gh.

Follow-up:

• If proxy-loop or certificate verification errors still appear after updating, the backend logs now include the exact upstream Proxmox/SDK detail and the selected VM config target node, which should make it clear whether the remaining issue is Proxmox node placement, endpoint TLS configuration, or an upstream Proxmox proxy condition.