About the Blog Perspectives focuses on high-scale services, data center design and operations, server hardware design and optimization, high-scale storage software and hardware systems, flash memory, service design principles, power efficiency and power management. About the Author: /jrh/work Most Recent Perspectives Posting: http//perspectives.mvdirona.com

Talks and Presentations

• Best Practices in Data Center Efficiency
• Grand Challenges in Database Self-Management
• Cloud Computing Economies of Scale
• Service Design Best Practices
• AFCOM Western Washington Chapter Meeting
• Slides from Conference on Innovative Data Systems Research
• CACM Interview with Pat Selinger
• Two Presentations at University of Washington
• Hotnets 2008 Paper
• Embarrassingly Distributed Cloud Services
• Internet-Scale Service Efficiency
• USENIX LISA
• High Performance Transaction Systems
• Most of my talks and papers are posted at: /jrh/work

Data Center Architecture and Efficiency

• Next Point of Server Differentiation: Efficiency at Very High Temperature
• Costs of Higher Temperature Data Centers?
• McKinsey Speculates that Cloud Computing May Be More Expensive than Internal IT
• Google Data Center Efficiency Summit (#5, #4, #3, #2, #1)
• 32C (90F) in the Data Center
• Google's Will Power and Data Center Efficiency
• Annual Fully Burdened Cost of Power
• Cost of Power in Large-Scale Data Centers
• A Small Window into Google's Data Centers
• The Box: How the Shipping Container Made the World Smaller

Service Architectures

• FriendFeed use of MySQL
• FAST 2009 Keynote: AWS S3
• Building Scalable Web Apps with Google App Engine
• Facebook Cassandra Architecture and Design
• Wikipedia Architecture
• MySpace Architecture and .Net
• Mesh Services Architecture and Concepts
• Google Megastore
• Flickr DB Architecture
• Geo-Replication at Facebook
• Scaling at LucasFilms
• Facebook: Needle in a Haystack: Efficient Storage of Billions of Photos
• Scaling LinkedIn
• 1,800 MySQL Servers with Two DBAs
• Berkeley Above the Clouds
• Rules of Thumb
• Scaling Web Sites

Storage

• Solid State Disks
  • SSD versus Enterprise SATA and SAS disks
  • Where SSDs Don't Make Sense in Server Applications
  • Intel's Solid State Drives
  • When SSDs Make Sense in Client Applications
  • When SSDs Make Sense in Server Applications
  • 1,000,000 IOPS
  • Laptop SSD Performance Degradation Problems
  • Flash SSD in 38% of Laptops by 2011
  • 100,000 IOPS
  • Flash Memory SSD Performance
• Hard Disk Drives
  • Disks, Lies, and Damn Disks
  • Why Disks Speeds aren't Increasing
• Software & Services
  • Under the Covers of Google App Engine Datastore
  • Key Value Stores (Project Voldemort, Ringo, Scalaris, Kai, Dynomite, MemcacheDB, ThruDB, CouchDB, Cassandra, HBase and Hypertable)
  • The Cost of Bulk Cold Storage
  • Measurement and Analysis of Large-Scale Network File System Workloads
  • Facebook Releases Cassandra as Open Source
  • Petascale SQL DB at Yahoo
  • Architecture of a Database System
  • Amazon SimpleDB Announced

Server Hardware

• Low Power Amdahl Blades for Data Intensive Computing
• Microslice Servers
• The Case For Low-Cost, Low-Power Servers
• Why Blade Servers Aren't the Answer to All Questions
• Client-Side Power Consumption
• Service Free Systems

High-Scale Service Optimizations, Techniques, & Random Observations

• Resource Consumption Shaping
• Should we Shut Off Servers?
• Monitoring at Scale
• Degraded Operations Mode
• Erlang and High-Scale System Software
• Why I love Full Text Search
• Interviewing Program Managers
• The Landscape of Parallel Computing Research: A View from Berkeley
• Jeff Dean on Google Infrastructure (with early Google server pictures)
• Argument for Degraded Operations Mode
• The Next Big Thing
• Server-Side Process Models
• Interviewing with insight at Microsoft
• Professional Engineering

Page last changed: 2009.05.17