Before learning DUBBO, let's introduce a few conceptual things
1. What is a distributed system
2. What is a cluster
3. How to achieve distributed
4. Why use DUBBO
Let's say one by one:
1. First of all, what is a distributed system: that is, multiple different systems call and communicate with each other, which can be called a distributed system. Distributed systems are mostly generated because of the complexity of the system, so the functions are split and then deployed. On different machines, they communicate with each other, and as a user, they cannot feel any changes in the system.
2. Talking about the cluster, the cluster is actually a group of machines with the same application, such as the mysql cluster, redis cluster, etc. we are talking about... Deploy the same application on different machines.
3. How to achieve distributed, in fact, the realization of distributed system mainly lies in the mutual communication between applications. Now there are many communication methods between systems, such as tcp, http, etc., as well as some framework webservice, as well as hession, netty, etc. , can communicate between systems. For example, to take a simple example, we bought a piece of clothing on an e-commerce platform. We want to buy, place an order, generate an order, pay, ship, sign, etc. This is a A very typical example of distributed system work, so many steps are not completed in one service, there may be commodity services, erp/order services, financial services, logistics platform services, so many services work together work to form a large e-commerce system.
4. Why do we use DUBBO? First, DUBBO is a distributed framework for distributed development. Second, it provides a wealth of outgoing protocols for us to choose from, which enables us to use a unified communication standard between services. For mutual calls, it also supports soft load balancing, mature functions and strategies (these are explained on the official website, the official website address dubbo.io), and the Alibaba team has given good advice on disaster recovery and performance tuning. . .
The above is the reason why DUBBO is used for distributed development, but DUBBO also has shortcomings, such as it is not suitable for the transmission of large amounts of data, etc.