This read-create Integer32, from 5 to 12 (likely minutes, per typical APS wait-to-restore conventions), sets the latency time before the system restores traffic back to the working interface after it recovers from a failure. Admins tune this to avoid rapid flapping back to a working interface that may still be intermittently unstable. For example, an admin sets this to 10 to require the working interface to be stable for 10 minutes before APS automatically switches traffic back to it.