Getting to a Multi-Cloud Deployment - A Case Study of
summary
Adobe Sign moves from a single ISP rack to automated multi‑cloud runs: a Python stack builder creates CloudFormation/troposphere templates for AWS and ARM templates for Azure, runs Chef to converge generic AMIs, and cuts build time from ~3 hours to under an hour by using a dependency tree. It runs ~11 data centers on four continents with ~50 services each, achieves four nines uptime, uses push-button deployment via Jenkins pipelines, release candidates, and scorched‑earth DB rebuilds for repeatable testing.
AI-written summary · about
chapters (17)
- 0:00Start
- 13:37Partnership with Microsoft
- 14:26Aws Stack Builder
- 18:14Open Source Tools
- 19:39Where We Are Today
- 20:54Stag Builder
- 22:32Data Center Switch
- 24:22Abstraction of Cloud Operation
- 26:18The Dependency Tree
- 27:33Summary of How the Tier Is Deployed in Multi Cloud
- 28:08Chef
- 29:41Dependency Tree for Chef Recipes
- 31:00Release Build Pipeline
- 33:40Final Steps
- 34:06Knowing Your Artifacts
- 34:22Release Candidates
- 38:50Build Pipelines
related talks
15mJourney to Hybrid Multi-Cloud
44mMulti Cloud Mesos
49mAutomation in Deployment on Hybrid Hosting and Private Cloud Environments
34mLessons Learned from Five Years of Multi-Cloud
27mFrom Supercomputing to Serving: A Case Study Delivering Cloud Native Foundation Models
22mMulticloud by Design
26mCase Study of Success
35mScaling Innovation: ADB's Cloud Journey