Service Foundry
summary
Argues for making services as easy to deploy as apps: cf push mysql and give it a disk. Identifies CAP tradeoffs, BOSH as slow for lifecycle work, and v1 service brokers as nonproduction. Proposes two tracks: change data software (Cassandra, RabbitMQ, Crate-style replication) and add platform primitives: Diego/xcd, persistent disks, SDN (Calico, Weave, Flannel), volume orchestration (Flocker), discovery/monitoring and provisioning (Brooklyn/Terraform) to enable micro data services and automated scaling and upgrades.
AI-written summary · about
chapters (6)
- 0:00Introduction and speakers
- 2:12Statefulness in PaaS explained
- 11:17Challenges solving stateful problems
- 19:15Changing data services for cloud
- 22:27Enabling platform for service foundry
- 25:16Call for community collaboration
related talks
30mOn-Demand Data Services with Cloud Foundry - Maria Ntalla & Simon Jones, Pivotal
31mCloud Foundry and Microservices: A Mutualistic Symbiotic Relationship
39mManaged CF-as-a-Service: An Operator's Cookbook
06mMake Workloads, Not Infrastructure
28mKeep Calm and Ship Services
43mHow Platforms Work
30mHell Freezes Over: The New Reality of Open Cloud Services and the OSBAPI - Josh McKenty, Pivotal
35mLearn to Develop a Service Broker in No Time
