Resilience
- A failing PVE API call no longer aborts the whole analysis. Each call now degrades gracefully: the affected check is skipped and a Warning (
WG0042, sub-contextApiError) is reported so it is clear the picture is incomplete. Endpoints not implemented on the running PVE version (HTTP 501) are skipped silently.
New checks
Cluster:
CC0005— HA resource in error state (manual recovery required).WC0009— replication job disabled (guest data no longer replicated).WC0010— enabled replication job without a schedule (it will never run).WC0011— nodes run different Proxmox VE versions.WC0012— nodes run different kernel versions.IC0007— user without an email (will not receive notifications).IC0008— empty group.IC0009— unused custom role (not assigned in any ACL).
Node:
WN0023— certificate expiring within 30 days.WN0034— bond with fewer than two slaves (no link redundancy).IN0004— self-signed certificate.
Storage:
WS0008— storage disabled.
Fixes
CS0001(Storage unavailable) no longer fires for storages disabled on purpose — those are reported asWS0008(Warning) instead of a false Critical.
What's Changed
- feat: add cluster/node/storage diagnostic checks by @franklupo in #43
- feat: resilient API calls — a failing fetch no longer aborts the analysis by @franklupo in #44
- chore: release 2.3.0 by @franklupo in #45
Full Changelog: v2.2.4...v2.3.0