a19af545a7
Docker host-ul era singurul guest fara replicare. Adaugate job-urile 102-0 (pve1, 21:20) si 102-1 (pveelite, 21:21), in sloturi libere. Sync initial rulat manual: ~88s fiecare, ambele OK. Prima incercare a esuat cu "No common base snapshot": ambele noduri tinta aveau volume orfane ramase de la o replicare stearsa candva (pve1 50 GB, pveelite 30 GB, din epoci diferite, nefolosite). Sterse prin API, apoi sync complet. Documentate: capcana volumelor orfane si faptul ca SSH intre noduri trece prin Tailscale — pentru inspectii pe alt nod se foloseste `pvesh get /nodes/<nod>/...`, nu ssh imbricat. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01MoFGxftQw92PX4Sdkn94pv