Synchronize data from multiple different site servers to the total Server

Actual case: a famous company is headquartered in Beijing and is also distributed across all sites across China, such as Changsha, Guangzhou, and Shanghai. they communicate through the company's intranet, now they have developed a software that keeps pace with the times to use on each site. Each site has corresponding server support, and the data stored in the software is stored in the oracle database on the server.

Actual case: a famous company is headquartered in Beijing and is also distributed across all sites across China, such as Changsha, Guangzhou, and Shanghai. they communicate through the company's intranet, now they have developed a software that keeps pace with the times to use on each site. Each site has corresponding server support, and the data stored in the software is stored in the oracle database on the server.

Actual case: a famous company is headquartered in Beijing and is also distributed across all sites across China, such as Changsha, Guangzhou, and Shanghai. they communicate through the company's intranet, now they have developed a software that keeps pace with the times to use on each site. Each site has corresponding server support, and the data stored in the software is stored in the oracle database on the server, it is now required to synchronize some of the relatively large and important data tables on each site server to the Headquarters server (Beijing server) for a better solution;

The simulated synchronization tables are item_rec and user_rec, both of which are record tables and produce a large amount of data every day. The headquarters may need to view the data in these two tables on each site from time to time, because you can directly connect to each site to see if there are sometimes network limitations, you need to synchronize it to the Headquarters server, as shown below:

Principal + teO3/Principal + 3b/i1tDOqs/Principal + Principal/Principal + t/weight + 0 vMfCvMrHt/weight + uavLvtfcsr + 3/s7xxvejutTauavLvtfcsr + 3 /s7xxvfK/b7dv + examples/examples + teO3/examples + examples/jQxc + examples/fI1da + se3W0LXEz + examples/examples + CjxwPsq + examples + pgltzybzcm9" http://www.2cto.com/uploadfile/Collfiles/20140929/201409290921142.jpg "Alt =" ">

In fact, there are many problems in using this synchronization solution:

1. The network is unstable;

2. Many tables need to be synchronized to each site. Each table corresponds to one trigger. Too many triggers may affect the normal use of the software;

3. When the head office server executes the operation record data of each site, the operation of the head office server may be affected due to the large data records in each site at a time, in fact, there is not much impact after tests;

4. When synchronizing the corresponding single record information (successfully executing a single SQL statement) on the head office server, you need to modify the corresponding synchronization completion mark of the operation log table in each site, due to the large number of records and the need for remote access, there will also be some impact;

During actual testing, many problems are found, so we hope that professionals can give good suggestions. If there is better technology, we hope to share it with others, this makes data synchronization with such problems more stable. Thank you !!