Building a Production Grade PostgreSQL Cloud Foundry Service
summary
Built a highly available, self‑provisioning PostgreSQL service for Cloud Foundry that runs in production and has been used for over a year, developed by a six‑person effort over two years. Key choices: dedicated clusters per instance instead of shared databases, hybrid pre‑provision pool plus on‑demand provisioning, automatic failover via repmgr + Consul, orchestration and scaling with BOSH deployments. Solution survives infra moves (VMware → OpenStack → AWS) and is in active use.
AI-written summary · about
chapters (11)
- 0:00Introduction and context
- 1:48Production grade requirements
- 3:14Service broker API design
- 4:44Postgres clustering challenges
- 9:00Cluster management with repmgr
- 11:53Shared vs dedicated instances
- 17:12Provisioning strategies
- 20:48Automation with BOSH
- 25:23Architecture overview
- 29:37Key lessons learned
- 31:41Q&A discussion
related talks
49mHave you ever wondered, how a cloud platform works?
33mPostgreSQL Service in Oracle OCI
13mBuilding reliable infrastructure with Postgres
05mFive Minutes by CloudNativePG River
34mRun PostgreSQL:The Kubernetes way
41mPostgres Plus Cloud Database
46mBig databases, many servers, data center or cloud - it's all here!
43mPostgreSQL in the cloud: Theory and Practice