Gateways & multi-WAN

BGP →

{{#DRIFT_BADGE}}{{#B_UNSAVED}}Unsaved edits{{/B_UNSAVED}}{{#B_OK}}In sync{{/B_OK}}{{/DRIFT_BADGE}}

UPLINKS Groups {{GROUPS_LABEL}} Health checks {{MONITORS_LABEL}} Probing {{LIVENESS_LABEL}}

Live status requires JavaScript.

⚠ Every uplink in a group is down. The last-known uplink stays installed so the box keeps a default route; it recovers the moment any uplink comes back.

⚠ A group steers your management path. A failover re-homes this admin session like any other traffic. Acknowledge it in the footer before applying.

⚠ IPv4 forwarding is off — this box will not forward traffic between its WANs and LANs. Enable it under Setup → Network tuning.

Awaiting confirmation by {{PENDING_ACTOR}}{{^PENDING_REVERT_FAILED}} · reverts in {{PENDING_REMAINING}} s unless kept — if the change broke your connectivity, do nothing{{/PENDING_REVERT_FAILED}}

Automatic revert failed — the gateway configuration may be inconsistent. The revert keeps retrying in the background; if it persists, resolve it from the console.

Uplink groups

A group is the set of WAN uplinks that share one default route. Tier 0 carries traffic; a higher tier waits as backup and takes over when every uplink below it is down. Uplinks in the same tier share the load by weight. When a health check marks an uplink down it leaves the pool and traffic moves to a survivor.

Read-only gateways.yaml rendered from the saved configuration (re-rendered after each save; unsaved edits are not reflected).

{{GW_RENDERED_PREVIEW}}

Preview truncated for display — validation and drift always use the full render.

Nothing to preview yet — add a group with an uplink and save to render gateways.yaml.

{{#MGMT_CONFIRM}} This change steers my management path. A failover can re-home this admin session and a state flush drops it; the auto-revert window below restores the previous configuration if I lose access. {{/MGMT_CONFIRM}} {{#WRITE}}

Auto-revert unless I confirm within s

The new configuration goes live at once but reverts itself unless you keep it in time. Recommended whenever a group steers the management path.

HA pair: skip the state flush while this node is CARP backup

pfsync replicates the master's states onto the backup; flushing them there would destroy them. Probing and pool updates keep working. Enable on both peers.