Scalability of software architecture

Two aspects

System performance can be extended

Algorithm

Scale up-compute intensive applications, using Multithread;io dense applications, async IO (reactive,for example,js node,c++ Ace,java netty Mina)

Scale Out-share no status,share little status (example Hadoop, inverse Oracle RAC), Distributed ADT ((probability) monoid etc.)

Hybrid-cache (tradeoff with scale out), batch

System functions can be extended

SOURCE level-Design pattern, oo (prefer composition to inheritance), function combinator

System level-decoupling, message middleware,pipeline Architecture,micro service

Cross-system-restful

Scalability of software architecture