Multi-Region Replication & DR for Distributed Postgres in Azure Cosmos DB

on YouTube ↗09m

" summary

Explains how a single-node native Postgres becomes a distributed cluster by adding worker nodes and moving shards via online shard rebalancing with zero downtime, using coordinator and worker roles; describes synchronous cross-region read replicas (up to five replicas per region) that each have separate connection strings and configurable compute; details checking replication lag with Azure Monitor and that promoting a read replica to read/write is irreversible and takes minutes, enabling cross-region DR and region migration.

AI-written summary · about

" chapters (11)

1. 0:00Introduction
2. 0:32Distributed Postgres
3. 1:54Single Node Postgres
4. 2:11Single Node Postgres Features
5. 3:35Transition from Single Node to Multi Node
6. 4:30How Transition Works
7. 5:45Real Replicas
8. 6:58Nested Replicas
9. 7:19Read Replicas
10. 8:02Recap
11. 8:22Outro

" related talks

• 59mLearn Live: Model data for Azure Cosmos DB for PostgreSQLCédric Derue, Gary Hope · Microsoft Build 2023
• 28mHA and DR at a glance with Azure Database for PostgreSQLSilvano Coriani · POSETTE: An Event for Postgres 2024
• 22mCloud-Native Apps, recovery ready!Yan Borowski, Machado Andre Racz · Azure Cosmos DB Conf 2022
• 22mCloud-Native Apps using Azure Cosmos DB, ready for recovery!Yan Borowski Machado, Andre Racz · Azure Cosmos DB Conf 2022
• 52mBuilding Distributed OLTP with Azure Cosmos DB for PostgreSQLCharles Feddersen · SQLBits 2023
• 32mAuto scaling Azure Cosmos DB for PostgreSQL with Citus, Grafana & Azure ServerlessLucas Borges · Citus Con: An Event for Postgres 2023
• 44mMulti-tenant SaaS apps made simple on Azure Cosmos DB for PostgreSQLAdam Wolk, Alicja Kucharczyk · SQLBits 2023
• 49mPostgreSQL’s Journey to Distributed DatabasesSameer M, Steve A · Postgres Conference Orlando 2025