What is blockchain technology

1.

From the perspective of technology and architecture, I will tell you my understanding of blockchain in common language

what is blockchain? In a word, blockchain is a storage system. To be more specific, blockchain is a distributed storage system without an administrator and each node has all the data

What are the common storage systems like

first, how to ensure high availability

the common storage system usually uses "rendancy" to solve the problem of high availability. As shown in the figure above, if the data can be copied into several copies and rendant to multiple places, high availability can be guaranteed. The data in one place is hung, and there is data in other places. For example, the master-slave cluster of MySQL is the same principle, and the raid of disk is also the same principle

two points need to be emphasized in this place are: data rendancy often leads to consistency problems

1. For example, in the master-slave cluster of MySQL, there is actually a delay in reading and writing, which means there is a inconsistency in reading and writing in a short period of time. This is a side effect of data rendancy

The second point is that data rendancy often reces the efficiency of writing, because data synchronization also consumes resources. If you add two slave libraries, the write efficiency will be affected. The common storage system is to use rendancy to ensure the high availability of data

so the second question, ordinary storage system, can write more

the answer is yes, for example, take this graph as an example:

in fact, MySQL can do a master-slave synchronization of al masters, master-slave synchronization of al masters, two nodes can be written at the same time. If you want to do a multi room multi live data center, in fact, multi room multi live data synchronization. What we should emphasize here is that multi-point writing often leads to the consistency problem of writing conflicts. Take MySQL as an example, suppose that the attribute of a table is self incrementing ID, then the data in the database is 1234 now. If one of the nodes writes and inserts a piece of data, it may become 5, and then these 5 pieces of data are synchronized to another master node, Before synchronization, if another write node inserts a piece of data, a piece of data with self incrementing ID of 5 will be generated. Then, after the generation, synchronize to another node, and the synchronized data will conflict with the two local 5's after it arrives, which will lead to synchronization failure and write consistency conflict. This problem will occur in the case of multi-point writing

how to ensure consistency in multi-point writing

the reform "Swan class" gives you more technical work