Verified Recovery and Safe Degraded Operation in Distributed Cloud Services: An Independent Recovery-Plane Approach

16 Nov

Authors: Research Scholar Raghvendra Rao, Dr. Yashpal Joshi

Abstract: Cloud recovery plans often treat a successful backup job or a restarted service as proof of recovery, although neither shows that a trustworthy service has been restored. This paper presents a recovery approach based on an independent recovery plane—separate identities, trusted artefacts, protected backups and keys—together with a six-step acceptance procedure and explicit degraded-operation states. A simple model shows why restore testing must be regular: if backups fail silently on average once every 180 days, the probability that the latest backup is unusable at a random incident is 1.9% with weekly restore tests, 7.7% with monthly tests and 20% with quarterly tests. Degraded modes are defined so that a service can remain partly useful while dependencies are lost, without hiding changes in meaning from users. The approach turns recovery from an assumed capability into one that is tested and accepted against stated criteria.

DOI: https://doi.org/10.5281/zenodo.24275542