Scale on: Upgrade on an existing server or switch directly to a new, higher performance server.
Scale out: Scaled-out, multiple servers are concurrently responding to requests from clients. Advantages: Low cost, simple extension architecture.
Cluster: A cluster, a redundant set of computers, that each computer implements the same functionality for load sharing and high availability.
Lb:load balancing/balancer Load balancer or load balancer.
Ha:high Availability High Availability
Hp:high Performance High Performance
Director: Dispatcher, which is the load balancer, is responsible for receiving client requests and distributing the request to a server that actually provides services in the background, based on an algorithm. Hardware such as F5, software such as Lvs,nginx
Lvs:linux Virtual Server
Realserver: is the server behind lb, and when LB receives the request, the request is forwarded to these servers by a certain load-balancing algorithm.
Basic Concepts in architecture