Effective Resource Management: Through Systems Design
summary
Example system is an image server that serves cat photos and hits <10% CPU while maxing RAM so the team compresses in memory to trade CPU cycles for higher memory density, avoiding hardware upgrades for several years. Uses instrumentation and dashboards built on the U-S-E (utilization saturation errors) method, a capacity plan stored as a spreadsheet or YAML, and multi-level and domain-specific load balancers plus a manual throttling "big red button" to stop runaway autoscaling. Metrics confirm the change works.
AI-written summary · about
chapters (11)
- 0:00Resource management intro
- 1:51Hardware resource costs
- 3:18Business investment reasons
- 4:45Resource optimization trade-offs
- 6:54Systems design for efficiency
- 8:57Software efficiency start
- 10:23Instrumentation and data
- 12:10Capacity planning basics
- 13:53Prevent unbounded growth
- 15:20Advanced load balancing
- 17:07Next steps
related talks
59mMastering the System Design Methodology
01h 02mTools for simplified large scale systems
55mInteractive Model-based Resource Analysis for Systems Engineers
25mLessons From Running Ceph as a Platform at Scale
38mArchitecting for a Greener Future
56mMastering Cloud Scalability and System Designs Deep dive
18mMaking what you have enough
50mEnergy Efficient Design Principles of Software Engineering