Merci pour ces élements. c’ets en tout cas ce que je constate
Et c’est assez destabilisant lorsqu’on lit ceci : FAQ | Zigbee2MQTT
What does and does not require re-pairing of all devices?
Requires re-pairing
You need to re-pair all you devices when:
- Changing the network key (
network_key) or panID (pan_id) inconfiguration.yaml.- Changing the Zigbee channel (
channel) in theconfiguration.yamlmight require re-pairing some devices, read the documentation for more info.- Switching between adapters requires re-pairing, except when:
- When the
serial.adapterin theconfiguration.yamliszstackoremberand theserial.adapterstays the same; e.g.zstack→zstackre-pairing is not required.
- There is one exception, when switching from a CC2531 or CC2530 (Z-Stack 1.2) to a CC2652/CC1352 (Z-Stack 3) re-pairing is required.
- When switching from
zstack→emberorember→zstackre-pairing might not be required, however results might vary as this is not officially supported.
- After switching, check if all devices are working and re-pair the ones that are not. In case pairing new devices is not working, re-pair some routers close to the coordinator while only permitting joining via the coordinator. Pairing should then work via routers that have been re-paired.
et conbee2 → ember n’est pas concerné par l’exception
J’ai fini par demandé à ChatGPT :
Tu soulèves un point très intéressant, et ta constatation est possible mais mérite quelques précisions techniques importantes.
Pourquoi tes appareils se sont « réappairés » automatiquement
Dans certains cas spécifiques, Zigbee2MQTT permet une transition fluide vers une nouvelle clé coordinateur, à condition que le fichier
coordinator_backup.jsonait été utilisé lors du changement.Voici ce qui a probablement permis la continuité :
1. Utilisation d’une sauvegarde du coordinateur précédent
Zigbee2MQTT sauvegarde automatiquement le coordinateur via le fichier :
data/coordinator_backup.jsonLors de la migration vers le SLZB-06M, si ce fichier est conservé et le nouveau coordinateur est compatible EmberZNet (comme c’est le cas ici), Zigbee2MQTT restaure les paramètres réseau critiques :
- Extended PAN ID
- Network key
- Channel
- Device table
Résultat : les appareils Zigbee croient toujours parler au même coordinateur.
2. Compatibilité EmberZNet entre ConBee II et SLZB-06M
- Même si la ConBee II n’utilise pas une puce Silicon Labs, son firmware expose une interface EmberZNet compatible.
- Le SLZB-06M utilise, lui, un EFR32MG21 de Silicon Labs, donc natif Ember.
- Cette compatibilité de stack facilite la restauration des paramètres réseau communs.
Ce n’est pas une migration sans précaution
Par défaut :
- Remplacer la clé sans backup = réappairage obligatoire.
- Avec
coordinator_backup.jsonet un coordinateur compatible = migration possible sans réappairage.
Comment vérifier si tu as bénéficié de ça
Regarde dans
/opt/zigbee2mqtt/data(ou autre dossier selon ta config) si un fichiercoordinator_backup.jsonest présent.Dans les logs de Zigbee2MQTT au démarrage, tu devrais voir quelque chose comme :
Loaded backup from previous coordinator
Conclusion
Tu as raison : une migration ConBee2 → SLZB-06M peut fonctionner sans réappairage, mais seulement si :
- Tu conserves les données Zigbee2MQTT
- Le nouveau coordinateur utilise EmberZNet
- Et que Zigbee2MQTT peut restaurer les identifiants réseau à l’identique.
Sinon, un réappairage est nécessaire.
J’ai vérifié et j’ai bien un « coordinator_backup.json » dans mes repertoires
Norbert