updated 1y ago
Stop using Integer ID’s in your Database
The problem is that most companies don’t apply the required forethought to creating the right primary keys to allow data from different channels to be merged.
from Web Analytics 2.0: The Art of Online Accountability and Science of Customer Centricity by Avinash Kaushik
- Who is creating those identifiers? And who is organizing data around them? Today, it’s nearly every site you visit, product you use, or company you encounter. Identifiers are listed in a database they have created, mostly private and siloed from every other company’s. Data is put there, and linked there, as well. Often this organized around a user ... See more
from Identity as Information
Alex Wittenberg added
Having a unique ID number for every object can come in handy when you’re trying to determine whether two objects are the same as each other.
from The Well-Grounded Rubyist by Joe Leo
- 7 must-know strategies to scale your database
Indexing:
Check the query patterns of your application and create the right indexes.
Materialized Views:
Pre-compute complex query results and store them for faster access.
Denormalization:
Reduce complex joins to improve query performance.
Vertical Scaling
Boost your database server by adding more CPU, RAM, or... See morefrom Shortwave — rajhesh.panchanadhan@gmail.com [Gmail alternative]
Nicolay Gerold added
The current generation of designers have spent their careers learning how to work faster and faster and faster. And while there’s certainly something to be said for speed, excessive speed tends to blur one’s purpose. To get products through that gate before anyone noticed what they were and how foul they smelled. Because we broke some things.
... See morefrom Design’s Lost Generation by Mike Monteiro
sari added
- To manage data, we need a protocol that makes it easy to store, discover, and route information about an identifier. For Web3 to live up to its promise, that routing table should (a) be unified rather than siloed by application or any other boundary, and (b) be sovereign, give control of data directly to each identifier.This suggests a simple desig... See more
from Identity as Information
Alex Wittenberg added
A common workaround is to de-normalize the database so that queries that previously require joins can be performed on a single table.
from System Design Interview - An insider's guide by Alex
This approach has typically lead to expensive, monster-sized databases that are not designed for a specific purpose and do not perform well.
from The Profit Impact of Business Intelligence by Steve Williams