Experience Report: Cloud Foundry Open Source Operations
summary
anynines runs Cloud Foundry on OpenStack for 3+ years and turns a default OSS install into a production PaaS by removing single points of failure: they build a 3-node Postgres BOSH release for DB HA, swap the brittle NFS blobstore for an external Swift S3-compatible blobstore, replace single-box HAProxy with a clustered SSL gateway driven by a Virtual Host API and SSL gateway cluster delivered as BOSH releases, and introduce a 5-node console cluster with two DNS masters to avoid static IP issues; results are redundant, self-healing deployments that scale to ~80 VMs.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 2:38Agenda
- 3:54Requirements
- 9:07Analysis
- 15:09Solutions
- 15:28Databases
- 16:27Cloud Control
- 17:39SSL Certificates
- 22:56Conclusion
related talks
41mCloud Foundry Weekly - Live at Cloud Foundry Day NYC
02mCloud Foundry
08mBroadening the Base: Increasing Options for Deploying Cloud Foundry - Troy Topnik, SUSE
10mConversation with Julian Fischer, CEO, anynines
14mThe Makers of Marvels
47mOpen Source Ops Fast Track with Operate First
51mMaking Complex Open Source Safe for Operations
31mSponsored Session: Freedom to Build: Open Source Foundations for Sovereign Cloud Infrastructure