Kubernetes Community Podcast: building disaster recovery and database resilience

A new episode of the Kubernetes Community Podcast features Vladimir Fedorkov, a database and resilience expert from FourNines and author of the Downtime Bar&Grill channel. The discussion focuses on Disaster Recovery (DR), architectural resilience, and how businesses can prepare for severe outages like data center failures or infrastructure breaches.

Qisqacha: - Guest speaker: Vladimir Fedorkov, database and resilience expert at FourNines. - Key topics include Disaster Recovery plans, active-active vs active-passive setups. - Deep dive into synchronous and asynchronous replication and midnight drills. - Available on YouTube, Apple Podcasts, Mave, Yandex Music, and Spotify.

Infrastructure redundancy and real-world risks A common architectural pitfall is deploying applications across two data centers that are not truly interchangeable, effectively doubling the risks. Furthermore, renting racks from different legal entities that happen to reside in the exact same physical facility provides a false sense of security that looks good only on paper.

Balancing business demands and technical reality Businesses often demand zero data loss and instantaneous failover without considering the financial implications relative to their size. Maintaining an active-active secondary site significantly increases hardware acquisition, maintenance, and labor costs. Pragmatism and strict adherence to common sense must dictate architectural decisions rather than chasing unrealistic perfection.

The importance of disaster recovery drills When unexpected failures occur, human panic often blocks logical thinking, making regular disaster recovery drills absolutely essential. The discussion warns against listening to overly optimistic evangelists who promise zero downtime and instant recovery for pennies, advising engineers to rely on rigorous planning and practical disaster recovery protocols instead.