Stale routes on new Primary after failover

[scachi]

Hi!

As I understand, after failover, system ha's "route-ttl" timer dictates validity period of FIB routes inherited from former Primary. Default is 10s - low value.

However, BGP Graceful Restart timer "graceful-update-delay" default is 120s - a much higher value.

How does Fortigate enforce these two values or is it up to me to align them? If so, given that "route-ttl" affects all routes, not just BGP routes, how would I align - recommendations?

Thanks!

[Toshi_Esumi]

Those two timer should work independently. No direct correlation betwen them. As described in the diagram below:
https://community.fortinet.com/t5/FortiGate/Technical-Tip-FortiGate-Viewing-FIB-RIB-routing-informat...
RIB (and then FIB) takes routes from those all upstream protocol's route as candidates then select the best routes.
When the new primary looked for new routes if a BGP route, regardless if it's an active route or a held up route, is there, it would take it as a candidate. If it's not there in BGP, it of course can't take it.
In other words, those add up. If BGP holds it 120 sec, then route-ttl hold 10 sec, the total hold time is 120+10=130 sec.

Toshi 

[Toshi_Esumi]

Forgot to add KB link you must have referred to.
https://community.fortinet.com/t5/FortiGate/Technical-Tip-Controlling-how-HA-synchronizes-routing-ta...