Preparing Apps for Dynamic Scaling
summary
Claim: VMs can be live resized—CPU and RAM added or removed without redeploy—so vertical scaling avoids load‑balancer complexity and session sharing. Demonstration: a WordPress test reports 3× throughput vs AWS autoscaling. Works with Apache, Gunicorn, Python/PHP, MySQL/Postgres; fails or needs care for preallocating systems like Cassandra and LibreOffice crashes. Tooling: LXC/docker, mentions Kubernetes for comparison. Check apps for single‑process CPU limits and heap/memory assumptions before adopting.
AI-written summary · about
chapters (10)
- 0:00Intro
- 2:52Outline
- 6:17Dynamic Scaling
- 7:28Planning Infrastructure
- 9:32Are You Ready
- 14:00What To Look Out For
- 15:37Compatibility
- 17:27Questions
- 18:53Scaling Vertically
- 19:56Conclusion
related talks
31mAutoscaling Apps without Supersizing Costs
44mIntroduction to application scaling
20mDrive higher utilization of apps in Azure Spring Cloud with Autoscale
01mScaling Java Applications Vertically & Horizontally with Jelastic PaaS
22mInstant provisioning and hot migration
45mAutoscale — Pay When You Earn
20mManage Scaling, While Keeping Costs Under Control. Challenge Accepted.
37mScaling LAMP doesn't have to suck