Autoscaling Apps without Supersizing Costs
summary
Cloud migration saves upfront maintenance but wastes money from misallocated resources; the talk gives concrete fixes: enforce comprehensive tagging and audit rules, right‑size and autoscale with switches to spot or reserved instances when appropriate, and fix developer inefficiencies like chatty logs or cache misuse — one case moves vault secrets local and drops pods from six to two, turning an opaque bill into measurable savings.
AI-written summary · about
chapters (22)
- 0:00Intro
- 0:37About Lara Newsom
- 1:45How did we get here
- 3:42Onprem vs Cloud
- 5:32Cloud Pricing
- 7:39What is the smallest service
- 8:38Understanding pricing structures
- 16:00Visibility
- 16:37Builtin tools
- 17:10External tools
- 17:30Dashboards
- 19:07Tagging
- 19:42Why Tags
- 20:41Tagging Plan
- 21:14Tag Hygiene
- 22:27Avoid Tag Sprawl
- 23:46Identify Problems
- 24:27Autoscaling
- 25:28Shutting down unused resources
- 26:37Reevaluating service options
- 28:25Refactoring
- 30:59Resources
related talks
45mAutoscale — Pay When You Earn
05mContinuous cost optimization as a never-ending process
25mPreparing Apps for Dynamic Scaling
23mFinOps and Cloud Cost Optimization
14mCut Costs, Not Performance: Implement Kubernetes Cost Optimization Strategies
10mReducing infrastructure cost during development and in production
26mCloud Agnostic Webscale
20mManage Scaling, While Keeping Costs Under Control. Challenge Accepted.