10 Ways to Ruin Your Cloud Experience (and How to Mitigate Them)
summary
Ten common pitfalls and concrete fixes: start with a tight spec, stop handcrafting VMs (one client had ~100 manually built instances), adopt Puppet/Chef or another CM, store configs in git-backed configs, avoid homegrown orchestration, measure then act with monitoring and elastic scheduling (example: shifting memory between dev and web clusters for a TV site), mirror production across environments, pair dev and ops, and expect provisioning SLAs to be slow (two day SLA often takes three).
AI-written summary · about
chapters (13)
- 0:00Introduction
- 1:43What is cloud computing
- 5:15Define a specification
- 7:40Environments
- 11:45Service Level Agreements
- 14:52Developers vs Operations
- 20:02Standardization
- 32:17Scalability
- 33:37Nontechnical problems
- 35:51Conclusion
- 38:57Cloud Computing
- 41:01Advantages
- 43:09Problems
related talks
35mHow to Cloud Ops Like a Boss
44mNine Ways To Fail at Cloud Native
40mAvoiding Pain When Operating in the Cloud
42mLessons Learned as a DevOps Consultant
13mTen Ways to Doom Your DevSecOps
07mTen Ways to Doom Your DevOps
22mCloudy Operations
46mDeploying and maintaining software for today's infrastructure