SDWAN SLA

Hi @HS08 

- The Single Performance SLA with Two Servers functions like an "AND" gate: both servers must be unreachable before a failure is triggered, making it suitable for mirrored or load-balanced services where simultaneous downtime indicates a significant issue.

- In contrast, the Dual Performance SLA, with individual SLAs for each server, operates independently, enabling more granular control and faster failover.

- Each server's status is monitored separately, allowing traffic to be rerouted promptly if one server fails, thereby enhancing high availability and resilience.

Best regards,

Erlin

• FortiGate sends probes to both servers.

It chooses the best responding server to calculate:

latency

jitter

packet loss

• The SLA is considered UP if at least one server meets the SLA thresholds.

IMHO the most significant difference is that you can only use 1 SLA in an SD-WAN rule if you select the "Best quality". This is fine for detecting loss of connectivity. The second server target will prevent false positives, e.g. when one server reboots/is in maintenance mode.

But still, you only test one behavior, e.g. using ping.

For "Lowest cost", you can let it decide on multiple SLAs, for instance DNS response and ping response. Thus, in this case you have finer control, which might be desireable if you want to ensure best WAN quality based on multiple criteria.