Performance Engineering: Slides on eBay Architecture Principles and High Performance Computing


I’ve got a couple of pdfs to share with you all in this post. Here are those:

1. eBay Marketplace Architecture – Architectural Strategies, Patterns and Forces – Randy Shoup, eBay Distinguished Architect
2. How to do 100K+ TPS at less than 1ms latency – Martin Thompson & Michael Barker

You can view these pdfs right here with slide notes for impatient souls:

eBay Architecture – Slide Notes

1. What eBay is up against? – Some Stats

1 Billion Page Views/Day
4.4 Billion API Calls/Month
44 Billion SQL Executions/Day

2. Architectural Strategies

Partition Everything (Patterns: Functional Segmentation and Horizontal Split)

Databases – No Database Transactions

Application Tier – No Session State
Search

Async Everywhere (Patterns: Message Dispatch and Periodic Batch)
Automate Everything (Patterns: Adaptive Configuration and Machine Learning)
Remember Everything Fails (Patterns: Failure Detection, Rollback and Graceful Degradation)

How to do 100K+ TPS at less than 1ms latency – Slide Notes

Tips for High Performance Computing

1. Show good Mechanical Sympathy (Memory/CPUs/Networks/Storage)
Keep the working set In-Memory
Write cache-friendly code
Write clean compact code
Invest in modelling your domain
Take the right approach to concurrency
2. What’s possible when you get this stuff right?
3. How to address the other non-functional concerns?
4. Concurrent access to Queues – The Issues
5. Disruptor

About these ads

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 27 other followers

%d bloggers like this: