BloxOne Service Edge supports Layer 3 (L3) high availability (HA) across two independent on-prem hosts. By providing redundancy for branch network traffic, Service Edge HA satisfies basic network redundancy requirements of the branch being accessible across links. Service Edge HA also helps mitigate failures with devices and with the on-prem control plane. L3 links are physical links.
Service Edge HA supports the following networks:
- LAN networks: HA is supported across multiple interfaces.
- WAN networks: HA is supported across multiple on-prem hosts for a single ISP.
A deployment of Service Edge HA comprises two running instances of Service Edge: one instance is configured as the active node, and the other as the passive node. Service Edge monitors the status of the LAN and WAN links through a health probe agent and uses the status to determine a failover. In the event of a failover, the active node moves into a failed state, the traffic is routed to the passive node (which now functions as the active node), and service continues without a disruption. After the failover in the active node is fixed, it takes over as the active node.
For an HA implementation to be successful, the traffic flow must be symmetric and HA on a WAN interface requires the interface to have a unique IP address for each on-prem host.
Deploy HA in the following order:
- Deploy two Service Edge on-prem hosts. For more information, see Deploying On-Prem Hosts for Service Edge.
- Enable the Service Edge router on both on-prem hosts. For more information, see Enabling Services for BloxOne Service Edge.
- Create health probe rules for both on-prem hosts. For more information, see Creating Health Probe Rules.
- Create HA rules for both on-prem hosts. For more information, see Creating HA Rules.
- Create routing protocol rules for both on-prem hosts. This step is required to get the routing protocol up and running. For more information, see Creating Routing Protocols.
- Create health probe policies for both on-prem hosts. For more information, see Creating Health Probe Policies.
- Create HA policies for both on-prem hosts. For more information, see Creating HA Policies.
- Create routing protocol policies for both on-prem hosts. For more information, see Creating Routing Protocols.
- On both on-prem hosts, create three Service Edge profiles: for the health probe, HA, and routing protocol. For more information, see Creating Edge Profiles.
- Select both Service Edge on-prem hosts, and associate the profiles with the instances. For more information, see Configuring Edges and Connectivity.
When you associate the profiles with the on-prem hosts, the set of all the rules you configured for HA is pushed to both on-prem hosts. The on-prem host with the lower metric is selected as the active node, and the other on-prem host is selected as the passive node. The health probe agent starts sending probes through the WAN interface to the configured probe. For example, if the configured probe is www.google.com and you ping www.google.com, and if the WAN interface is up on both on-prem hosts, then both hosts will report the status as healthy. Whatever actions you perform, they are routed through the active node.
If the WAN interface on the active node goes down, the health probe agent cannot ping www.google.com and reports the active node's status as unhealthy and considers it as having failed. In addition, Service Edge reroutes the traffic through the passive node. Once the active node is up and running, traffic is once again routed through the active node.