Data synchronization and replication issues for Sybase systems

The Sybase replication server (Sybase replication server) is designed to meet the needs of an increasingly evolving Enterprise Client/server computer environment. Since the replication server has been widely used, it has become the foundation of enterprise-wide client/server application model.

A practical replication system must take into account the following business requirements.

High reliability of data: Replication systems will make application systems more reliable and reduce the impact of computer system failures on business processes.

Consistency of information transmission: the consistency of data protection in each node of distributed system.

High system performance: Replication systems must allow each node to optimize access to local data without increasing the burden on the data source nodes and effectively using the network.

Convenient data centralized management: System administrators can easily manage the data distributed with each node.

Heterogeneous data source Access: Replication systems are capable of moving data between data sources from different vendors, such as data replication and synchronization between Sybase, Oracle, and MS SQL systems.

Local autonomy: Each node is able to determine what data it will accept, and to access and update this node's data.

How Replication Servers work

In this section we will describe the various parts of the Sybase replication server product and their capabilities. Replication data operations can be divided into four parts:

(a) Change the monitoring of data transactions for one or more data sources;

Transmission of transactions on the network;

Distributing transactions to each target data source;

Replication System Management.

In the Sybase replication server environment, modifications to the data in the source database are implemented by a process called Log transfer Management (Transfer Manager, the LTM). In the most typical structure, the LTM is placed on the same hardware platform as the source database.

The LTM transmits log to the replication server process, and the replication server process can be on another machine. In a replication system, the configuration of several replication servers can be distributed on different LANs, depending on the configuration of the replication system, the log of the changed data is transferred from one replication server to the other replication server, and the data changes are transmitted to the target database by these replication servers. The complete data replication operation process can be managed and monitored by a Replication server management interface based on the GUI environment. Because the stable queue (stable queue) is used in the replication system, data from the source database to the target database can be avoided in any part of the system that fails to occur. A stable queue is a security mechanism in which fault tolerant environments fail.

Log Transport Management

The Sybase replication server product is part of the Log Transfer Manager (LTM). The LTM is a low scheduling process, and the most typical method of configuration is to run the LTM on the same platform as the source database. If the source data for more than one node in the replication system is modified, you need one LTM per node.

The task of the LTM is to monitor and capture transactions with its associated database, and to transfer these transactions to the replication server so that the replication server passes transactions to other replication nodes.

The LTM has a join to SQL Server that reads the transaction log of the primary SQL Server and monitors whether it changes the primary node data. The transaction log is the best source of information used to observe changes in the primary node data, because it contains transactions that have been submitted to the server and can be used to recover transactions. In Figure VI, when an application modifies the master node data, the transaction is recorded in the transaction log to ensure data consistency, and data changes are written to disk when submitted. The replication process does not interfere with database functionality, except that the LTM monitors transactions in SQL Server and transmits it to the Sybase replication server when it finds that a transaction should be replicated.

The LTM process is an application written by the Sybase open Server/open client and has a join to the Sybase replication server. This process can reside on a separate machine. It is precisely because the Sybase replication server process can run on a different machine from the main data source that the user minimizes the burden that the replication system imposes on the primary database.

The LTM process can translate transactions that are independent of the main data source command to the replication server. For example, the completion of a transaction is "Update", and the LTM transmits the transaction to the replication server and interprets it as an update command completed by the master node. The mechanism of this source database transaction to the replication server translation or mapping of commands recognized is an important strategy for Sybase replication server technology. The interface from the LTM to the replication server is visible to the user. In other words, if you can get a transaction from a data source, you can map it to the Sybase replication system environment and replicate it. Unlike other competing manufacturer products, Sybase replication technology allows applications to replicate heterogeneous data sources.