Detailed Oracle Distributed System data replication technology

A medium-sized or large company is often comprised of geographically dispersed departments that typically require data sharing. For these shared data, it can be stored on a site that requires users to access from the site. The advantage of this scheme is that data consistency is easy to guarantee, but its shortcomings are very prominent, that is, the site's load, Network load, remote user data response is slow. Data replication technology can effectively solve this problem, it enables local access to data by replicating these shared data to multiple databases in different locations, reducing network load, and improving the performance of data access, and by periodically synchronizing the data in the database (usually every night), This ensures that all users use the same, most recent data. This technique is applicable to the application mode of large number of users, wide geographical distribution and the need to access the same data in real time.

The concept and characteristics of data replication

1, the concept of data replication and classification.

Data replication is the copying of data from the database to one or more different physical sites, keeping the source database consistent with the data specified in the target database.

According to the real-time data replication, data replication can be divided into synchronous data replication and asynchronous data replication. Synchronous data replication is the replication of local production data in a completely synchronized manner, with each local IO transaction waiting to be released by the complete party of the remote replication. Asynchronous data replication refers to the way in which local production data is replicated offsite, and every local IO transaction is released normally without waiting for remote replication to complete. Synchronous replication is strong in real time, and remote data is fully synchronized with local data. But this way is affected by the bandwidth greatly, the data transmission distance is short. Asynchronous replication does not affect local transactions, long distances, but its data is slightly delayed than local data. In an asynchronous replication environment, the key to all applications is to ensure data consistency.

According to the type of replication site, data replication can be divided into multiple master Site replication, materialized view replication, and mixed replication. Multi-Master Site replication is also known as Peer-to-peer Site replication, where each site is the master site, all need to communicate with other sites, the site is equal. Materialized view replication contains a master site, one or more materialized view sites where the contents of the materialized view can be all copies or partial copies of the target main object at a point in time, where the target main object can be either a table on the master site or a master materialized view on the materialized view site. Hybrid replication consists of multiple master sites and multiple materialized view sites, a combination of master Site replication and materialized view replication, and is suitable for complex business situations.

2, the characteristics of data replication.

Data replication improves the security of data by making backups on multiple sites, as well as increasing the availability of data, because if a site has a problem, the user can choose another site to continue, and the application can continue to operate, which provides a fault-tolerant protection mechanism for data replication.

However, the most basic function of data replication is to improve database performance. By copying the data from the remote database to the local area, the application can access the data nearby, thus reducing the network transmission load and increasing the efficiency. And in a data replication system, you can provide load balancing between multiple sites, allowing these users to use the server, and several other users can use other servers to avoid overloading some sites.

Materialized views also provide a subset of replication, so that each site can only copy the data they need, but also reduce the amount of traffic on the network.

How to implement data replication

Before the implementation of the specific, the first to do a good job of design and planning. This requires a detailed analysis of the specific business situation, the design of a set to meet the business needs of the program. Usually in the design process, you need to determine the database site to be established, the types of sites, the data objects need to be replicated, as well as synchronization methods, conflict solutions and so on.

After the design is complete, can be specific to achieve data replication, the data replication implementation process as shown in the following figure:

As you can see from the diagram, the implementation of data replication mainly includes the following steps:

(1) Create a replication site;

(2) Create group objects;

(3) Configure the conflict resolution.