| Information2

菜单

Real-time Data Replication Solution

Real-time, Byte-level, Graphical

Download Materials

Byte-level replication and efficient transmission;

Flexible deployment without network architecture modification;

Data consistency for applications and databases;

Centralized management via a graphical interface

Solution Background

In today's highly informatized era, data has emerged as a new production factor, on par with land, labor, capital, and technology. However, as an intangible asset, data is also the most susceptible to various accidents. Once lost, it can have severe consequences.

Data replication serves as a crucial application in data circulation, exchange, sharing, protection, integration, computational analysis, management, and other fields, playing a significant role in maximizing data value. Data that can be replicated, circulated, and exploited is wealth; otherwise, it may become a burden for enterprises.

Enterprises and organizations need to ensure the security of their critical data. Data replication ensures information security and availability, so it's always there when you need it. Data replication technology constantly prepares for disaster recovery, ensuring data is never lost.

Solution Overview

This solution provides enterprise-level continuous data replication and on-demand recovery services, enabling you to easily backup data from various branch nodes to local or remote disaster recovery center nodes in real-time and quickly restore required data as needed. It strictly guarantees the consistency and integrity of data between the production system and the disaster recovery center. Based on byte-level real-time replication technology, it is widely applicable to real-time disaster recovery and backup protection for file systems, database systems, mail systems, etc. By deploying a lightweight client on the production server, it captures byte-level increments in real-time and transmits them to the disaster recovery server in real-time.

Solution Architecture

Control Machine

The host providing the web interface console service; it provides a unified data management platform service, enabling users to manage and utilize node resources and rule tasks.

working machine

The source host for data replication; the Worker program reads data from the user's production system and sends it to the target backup server over an IP network. Simultaneously, it captures IO changes in the source file system and sends them to the target system in serialized transmission over an IP network.

Disaster Recovery Machine

The target host for data replication; the Disaster Recovery program receives production data over the IP network and saves it to the local file system in the same file IO write order as the source end.

Solution Description

Data Replication Workflow

Multiple-to-One Data Replication

One-to-Multiple Data Replication

Cascading Data Replication

The console configures replication rules and distributes them to the working machine. A data transmission channel is established between the Worker and Disaster Recovery machines. Data real-time replication consists of mirroring and replication stages:

Data Mirroring Stage

Before data synchronization begins, the Worker traverses the directories to be protected. After traversal, it transmits the data content and file attributes on the local disk to the disaster recovery system over an IP network and saves them to the target local disk. During full synchronization, incremental data is not transmitted but only captured. Incremental data is transmitted to the disaster recovery machine after full synchronization is complete.

The data mirroring process supports verification, improving file synchronization efficiency during re-mirroring to quickly enter the replication state.

It supports sending some files to the transmission queue for data transmission and verification during traversal.

It supports multi-threaded concurrent traversal/mirroring, improving synchronization performance in scenarios with a large number of small files.

Data Replication Stage

The Worker continuously captures IO changes at the file system level and uses a unique data serialization transmission technology (Data Order Transfer) to write the data to the backup server's disk in the original IO order after transmission, strictly ensuring data consistency and integrity between the production system and the disaster recovery center. If the current replication rule is in the mirroring stage, the real-time captured IO data will be cached first and then transmitted to the target disk of the disaster recovery machine after the mirroring stage ends.

Both during data mirroring and replication, it supports advanced functions such as data verification, encrypted transmission, bandwidth limiting, data comparison, broken point resumption, orphan file processing, etc.

By configuring two replication rules:

From working machine A to Disaster Recovery Machine

From working machine B to Disaster Recovery Machine

Centralized backup of data is possible, and data from multiple platforms, such as Windows and Linux, can be backed up to a single server for unified backup management.

By configuring two replication rules:

From working machine to Disaster Recovery Machine A

From working machine to Disaster Recovery Machine B

This allows the working machine to publish the latest data in real-time to different disaster recovery hosts, enabling real-time data sharing or disaster recovery protection between branches at any distance.

By configuring two replication rules:

From the local working machine to the local Disaster Recovery Machine in the same city

From the local Disaster Recovery Machine to the remote Disaster Recovery Machine in a different city

This realizes a typical "two locations, three centers" deployment approach, with data automatically replicated in real-time between the three centers and ensures data security during transmission, achieving rapid cascading synchronization of data.

Solution Advantages

Byte-level Replication and Efficient Transmission

With bytes as the smallest unit for data capture, the amount of data that needs to be replicated is greatly reduced. The serialized transmission method is much more efficient than traditional backup transmission in scenarios with narrow bandwidth and remote transmission. Network speed limits can be set to ensure that bandwidth priority satisfies production systems and business applications.

Data Consistency for Applications and Databases

Apart from supporting real-time protection for unstructured data, it also supports real-time protection for structured data such as Oracle, SQL Server, DB2, and ensures data consistency through byte-level replication technology and serialized transmission mechanisms.

Flexible Deployment without Network Architecture Modification

Supports one-to-one, multiple-to-one, one-to-multiple, cascading, and other data replication methods, with simple network structure requirements and no need to modify the existing network mode of the production environment.

Centralized Management via a Graphical Interface

Intelligent installation wizards and flexible backup strategies; rich statistical information for easy viewing of backup results; supports custom alert settings with notifications via email for abnormalities; supports automatic verification to check the recoverability and integrity of backup data; supports user permission management functions. Only one console is needed for centralized management of different databases.

Application Scenarios

Mass Unstructured Data Disaster Recovery

Supports massive unstructured data synchronization across heterogeneous platforms and storage scenarios, supports multi-threaded concurrent traversal/transmission, improves transmission efficiency; based on byte replication technology, it achieves real-time replication of incremental data; simultaneously supports multiple data verification methods to ensure data consistency between the source and backup ends.

Shared Storage Cluster Disaster Recovery Support

Supports real-time protection of shared storage data in cluster environments such as MSCS and VCS.

Taking MSCS cluster protection as an example, simply install the Node node software on each cluster node, then create a Node resource in the MSCS cluster management for scheduling replication rules, and you can perform real-time protection on files on the shared disk of the Cluster, replicating data from the shared disk to the specified disaster recovery machine in real-time.

When an MSCS cluster fails over, the shared storage transfers to another available cluster node. At the same time, the Node resource detects the failover and switches the related real-time replication rules to run on the failover cluster node.

Real-time Replication of Database Files

Supports real-time replication of data files for various standalone applications & databases (such as Oracle, MySQL, SQL Server, etc.) and ensures the consistency of synchronized data through data verification functions.

Local and Remote Data Disaster Recovery

Based on low data replication latency and low bandwidth requirements, it can support replication to both local and remote data centers, enabling a two-location, three-center deployment scheme.

Customer Case Study

Case Name: CITIC Securities

Case Description

Real-time multiple-to-one replication of securities production and transaction logs to a local log aggregation server, distinguishing data sources by appending the source machine IP suffix to the replicated data, and real-time replication of data from the local log aggregation server to the remote disaster recovery center;

The aggregation server can retain expired transaction logs deleted from the original end through the replication data filtering and deletion function.snamejin

Case Value

Implements the replication and aggregation of massive transaction logs in a multiple-to-one data replication scenario;

Supports filtering out log deletion operations and remote disaster recovery, fulfilling log audit and security compliance requirements.