1. Distributed
A large system is divided into several business modules, and the business modules are deployed to different machines, and each business module interacts with the data through the interface. The difference between distributed methods is based on different machines and different businesses.
above: Service A, B, C, D are business components, business access through API Geteway, respectively.
Note: Distributed needs to do transaction management.
Distributed transactions can be consulted: Distributed transaction solutions for MicroServices architectures
2. Cluster mode
Cluster mode is the deployment of different servers to the same set of services to external access, to achieve load balancing services. The way to differentiate clusters is based on whether the deployment of multiple server services is the same.
Note: The cluster mode needs to do a good session sharing, to ensure that the process of switching between different servers will not abort the service because the session is not obtained.
General configuration nginx* Load container implementation: The static resource cache, session sharing can be implemented, NGINX support 5,000 concurrent volume. 3. Whether the distribution belongs to MicroServices.
The answer is yes. MicroServices mean that the module is split into a separate service unit through the interface to achieve data interaction. 4. Micro-Service Architecture
MicroServices are designed to not affect existing system operations because of upgrades and bugs in one module. The small difference between microservices and distribution is that microservices are not necessarily distributed across multiple servers, but can also be the same server.
Distributed and tiny architectures are similar, but they are not deployed in the same way.