Building a Robust Cloud Foundry (HA, Security and DR)
summary
Compares two HA patterns: dual foundations (active/active or active/passive) versus one foundation split across AZs, with concrete tradeoffs — F5 GTM/LTM + HAProxy + NSX introduces three load‑balancer layers and requires DNS/gtm logic; dual‑dc services force sub‑5ms or use GemFire/Cassandra instead of ugly dual‑Redis writes. Uses BOSH and Ops Manager export (or BOSH manifests) for recovery, prefers external BOSH DB + blob store, relies on S3 deny S3 bucket deletion + versioning, MFA, security groups and CloudTrail for security and fast VM resurrection.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:07High availability basics
- 2:24Multi-site deployment
- 7:17Single site HA
- 13:23Deployment trade-offs
- 14:33Disaster recovery
- 19:41Security hardening
- 31:00Key takeaways
related talks
43mHow Platforms Work
24mExpand Cloud Foundry for the Enterprise - Tim Leong, Comcast
36mCatch Me If You Can – Leveraging Cloud Foundry, SDN, BOSH and Predictive Analytics
30mCloud Foundry: The Gaps
20mAdvancing the Future of Cloud Foundry
41mSeven patterns to increase application availability
39mManaged CF-as-a-Service: An Operator's Cookbook
23mThe Genesis of Cloud Foundry: Using Open Source Tools to Easily Deploy...