AGIC will give 1-2 minutes of downtime on backend application access when AGIC pod got restarted

[Aprilllllll]

When the AGIC pod has to be recreated for some reasons, like node reimage, the access to backend pods will be impacted. The new AGIC pod would take 1-2 minutes to sync the configuration, and backend application is unreachable during this 1-2 minutes. Is there anyway to improve the availability of AGIC deployment? Thank you!