Where Technology Becomes a Genre, kt cloud: Kechler Interview Series #2 — The Engineering Behind 'Multi-AZ' That Keeps Services Running Even During Outages

The existing DR (Disaster Recovery) structure worked by switching services to another region when a failure occurred, but it had limitations in operational burden and inefficiency. The main causes were the operational complexity of customers having to determine the recovery sequence themselves, and the cost issues arising from Active-Standby systems that go unused in normal times.

kt cloud shifted its perspective from "recovery after a failure" to "design that assumes failure will happen." To achieve this, it introduced a Multi-AZ (Multiple Availability Zones) architecture that distributes and operates services across multiple availability zones within a single region.

Through the Multi-AZ structure, service continuity can be maintained even if a failure occurs at a specific data center, with the goal of dramatically reducing the recovery operation burden on customers.