forked from donnemartin/system-design-primer
-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Phuc Nguyen edited this page May 26, 2023
·
2 revisions
-
Performance: how system performs in a fixed environment, e.g. performance for 100 user and 2 servers
-
Scalability: how system performs in changing environments, e.g performance when users increasing from 1 to 10 to 100
-
Latency: how long it takes to see the output
-
Throughput: how many inputs can be processed per unit of time
-
CAP: have to support partition tolerance (system works during network failures), make trade-off between consistency (can work with latest data) and availability (can work with some data)
-
Object store: data + metadata + unique identifier, e.g. AWS S3. Good for storing large volume of data and retrieving part of it. Not good for operational data.