Zero Data Loss at Scale

on YouTube ↗11m

" summary

Claims a production system that achieves zero data loss during a regional outage by combining message replication across regions, replay-based recovery and snapshot restores. Architecture uses a message queue with persisted transaction logs, quorum writes to multiple replicas, and peer‑assisted rebuilds for failed nodes. Case study reports successful recovery with end‑to‑end message tracking and monitoring; performance tradeoffs favor eventual consistency for throughput while prioritizing durability and measurable SLA outcomes.

AI-written summary · about

" chapters (12)

1. 0:00Introduction to Microservice Architecture
2. 0:25Understanding Asynchronous Systems
3. 1:54Challenges in Distributed Systems
4. 2:43Impact on Site Reliability Engineering (SRE)
5. 3:45Preventing Data Loss
6. 4:14Advanced Recovery Techniques
7. 4:56Distributed Acknowledgement Protocol Design
8. 6:06Standardizing Message Formats
9. 6:56Performance Optimization Strategies
10. 8:58Consistency vs. Performance Trade-offs
11. 10:18Case Study and Key Takeaways
12. 11:13Conclusion

" related talks

• 29mMulti Active-Active and disaster recovery with MongoDB database centerXixuan Xixuan · Percona Live Open Source Database Conference 2017
• 45mChasing unicorns : 0 dataloss and 99.999% availabilityLaetitia Avrot · PGConf NYC 2021
• 30mResilience strategies to level up your workloadsRada Stanic · AWS Summit ANZ 2023
• 47mDisaster Recovery Strategies for SQL ServerAndrew Pruski · SQLBits 2022
• 47mDatabase ResiliencyRavikumar Buragapu · Percona Community Live 2022
• 47mDisaster Recovery Strategies for SQL ServerAndrew Pruski · SQLBits 2024
• 24mFailure is Not an Option: Highly Available Distributed SQLEric Pratt · Distributed SQL Summit Asia 2021
• 32mActive Data Guard: Disaster recovery tips and tricksMARKUS MICHALEWICZ · Oracle DatabaseWorld 2021