Is your feature request related to a problem? Please describe.
Currently, when the Panel restarts or becomes unavailable, all Node configurations are cleared. This leads to a complete service interruption because the nodes cannot maintain their last known functional state without a live connection to the panel.
Describe the solution you'd like
I would like to suggest an "Offline Mode" or "Local Configuration Caching" mechanism:
Persistence: Nodes should save the received configuration to a local file (e.g., config.cache.json or a local SQLite db).
Graceful Failure: If a Node cannot connect to the Panel on startup or loses connection during runtime, it should continue to serve traffic using the last cached configuration instead of clearing it.
Recovery: Once the connection to the Panel is restored, the Node can then sync and apply any new configuration updates.
Describe alternatives you've considered
None. Currently, the system has a single point of failure (the Panel) regarding configuration state.
Additional context
This is critical for high availability. In a production environment, the data plane (traffic forwarding) should ideally be decoupled from the control plane (management panel) to ensure that maintenance on the panel doesn't bring down the entire network.
Is your feature request related to a problem? Please describe.
Currently, when the Panel restarts or becomes unavailable, all Node configurations are cleared. This leads to a complete service interruption because the nodes cannot maintain their last known functional state without a live connection to the panel.
Describe the solution you'd like
I would like to suggest an "Offline Mode" or "Local Configuration Caching" mechanism:
Persistence: Nodes should save the received configuration to a local file (e.g., config.cache.json or a local SQLite db).
Graceful Failure: If a Node cannot connect to the Panel on startup or loses connection during runtime, it should continue to serve traffic using the last cached configuration instead of clearing it.
Recovery: Once the connection to the Panel is restored, the Node can then sync and apply any new configuration updates.
Describe alternatives you've considered
None. Currently, the system has a single point of failure (the Panel) regarding configuration state.
Additional context
This is critical for high availability. In a production environment, the data plane (traffic forwarding) should ideally be decoupled from the control plane (management panel) to ensure that maintenance on the panel doesn't bring down the entire network.