What is the replication of more masters?

Multi-maaster replication is a technique used by database management software, so many users can use and update multiple copies of one database stored on different computers. When a change in a database that uses multiple masters replication is transmitted to all other computers on the network, so all the database copies used are updated. Some more replication settings include redundant data backup and decentralized architecture in which database copies can be updated, even if one of the computers on the network stops working. Without a single assigned main system, as is the case with Master-Slave architecture, changes and administrative database control orders can be issued from any of the terminals more masters in the network instead of relying on only one physical terminal. The largest complications that can occur when using a multi-mastreplika system ER involves updating all systems fast enough to make data bouTala always synchronized in the network.

The term "Multi-Master" comes from the basic method of replication of the main slave, in which one terminal is marked as the main. In the situation of Master-Slave he is able to update information about slave units. The replication system of multiple masters has several systems marked as Masters, and each master can only be responsible for himself or could be in charge of a number of computers known as a replication group. For multiple masters, any main system can initiate a change in all other main systems, which potentially leads to very complex network interactions in very large systems.

There are two forms of replication of multiple masters, the first is synchronous replication. Synchronous Replication Functions In Real Time, all copies of the database every time a change is made. This removes most situations that may occur when data is not synchronized over the entire networkI, but creates a huge amount of network traffic and may require a large amount of processing in a larger system. For these reasons, synchronous replications are primarily used in networks where there is only a handful of main computers.

The second type of replication multiple masters is asynchronous replication. Instead of updating the systems at each change, the changes in the database are stored as events in each system and these changes are used either when the event is started or during the planned update time. This means that a smaller bandwidth and processing of processing are used, but it increases the chances that two users could try to change the same information at some point in contradictory ways due to delay in updating or latency. Most distributed database management systems use asynchronous replications due to low amounts of resources consumed compared to synchronous replication of multiple masters.