BOSH Splitter
summary
Tool that splits a single BOSH manifest into core + per-job manifests so large CF deployments are deployed as multiple BOSH deployments rather than one giant one. It is a post‑process that uses spruce and a Makefile and keeps all original releases, templates and networks intact while emitting core.yaml plus one split yaml per job. Used on Predix at ~600 VMs to turn multi‑day stemcell rollouts into per‑group ~8‑hour upgrades, preserves the original manifest and lives at the linked GitHub repo.
AI-written summary · about
chapters (18)
- 0:00Introduction and speakers
- 1:46Problem with large BOSH deployments
- 3:35BOSH Splitter use cases
- 5:00How BOSH Splitter works
- 8:59Demo setup and files
- 10:27Running the split command
- 12:16Deploying core and split manifests
- 13:42Verifying split deployments
- 14:47Accessing the tool and resources
- 15:31Q&A on use cases
- 17:57Change windows and capacity planning
- 18:41Q&A on split files and original manifest
- 19:50Role of Spruce in the process
- 20:53Networking considerations
- 21:49Q&A on production usage
- 23:00Managing deployment groups
- 24:09Q&A and future improvements
- 25:57Concourse pipeline challenges
related talks
32mIntroducing bosh-workspace: how we deploy all things BOSH
25mBOSH: The Simple Solution for Complex Systems
41mBOSH 101 - A Beginners Guide
30mStreamlining the BOSH Experience
25mBOSH Bootstrapping
22mThe New Way to BOSH Deploy - Urvashi Reddy, Pivotal
37mIs Bootstrapping BOSH Incredibly Simple Yet?
28mHow to Deploy CF and Other Bosh Releases on Bare-Metal Environment