Centralized Backup Solution

Time: 2017-06-08 11:03:37
Source: myself
Click: 4976

The System Status and Problems

1. Incomprehensive data protection, unable to cover all the important business systems

2. Lack data backup mechanism, causing high risks of system maintenance

3.Inflexible informatization infrastructure and low utilization rate of resources

4. Applications must be attached to the old version of operating system

5. Short life span of equipment and management and maintenance difficulties caused by decentralized business system deployment

Project Construction Goals

The goal is to build a mode featuring horizontal data distribution and centralization and two-dimension centralized backup and storage.

1. Reconstruct file servers through clustering and mirrored storage, and achieve centralized storage in order to improve data availability and reliability;

2.Backup the data of automatic production cluster, to-be-released correspondent business systems and workshop servers for self use in the current project, and store the data in headquarters in future projects so as to check valid data intuitively such as open the file directly through backup data recovery in headquarters;

3. Provide data basis by making profound data analysis in order to build big data platform in future;

4. Due to the limited network between plant and headquarter, it is necessary to delete repeated local data in each plant by deduplication and rebuild a new backup network in future data guard for data remote backup;

5.Achieve unified management to meet the need to construct integrated and centralized data backup and storage;

Project Scheme

1. Topological Graph of the Scheme


2. Introduction to the Scheme Design

ORACLE RAC for dual-cluster system and Symantec Storage Foundation HA are used to establish a local fault-tolerant mechanism. Storage devices provide different levels of data backup solutions for major industries by backup methods based on high-level FC RAID (Fabric Redundant Array of Independent Disks) and EMC Data domain + NETWORKER. Featuring unified management, flexible expansion and cost control function, such solutions effectively solve the problems of large enterprises' demands on data security operation and maintenance.

Based on ORACLE RAC cluster, Xiamen Node tests the state of the other side through heartbeat detection and periodically send detected information of the main server to the backup server. Once the information transfer stops, the backup server immediately takes over the role of the main server to provide services, enabling rapid switch of database server, ensuring continuous application and reducing downtime losses.

The database server uses LAN network to connect the user. Clusters are transparent to network users. ORACLE RAC cluster provide a virtual IP address, so that users only need to visit that IP address to access the database server. When failure occurs to the backend server, another server will automatically replace the IP address of its network card with the virtual one, so that the user can continue to visit the website in spite of one server's failure.

Storage devices turn volumes striding across the array into images through Symantec Storage Foundation HA by creating the same disk group and new volumes on it. Related data are send to both the production file server and the backup server. Through such fault-tolerant mechanism, if one LUN fails, another LUN can continue to work and provide services continuously. Even if failure occurs to the disk group which one of the LUNs belongs to, the data can be rapidly recovered by increment after replacing the disk.

3. Advantages of the Scheme

An Integrated Platform

The major plant is the core of the whole platform. By setting up backup management platform there, demands such as platform sharing, hierarchical management and node self-service can be fulfilled flexibly so as to meet different management needs of distributed organizations.

·           Construct and manage the centralized backup platform of node of multi-branch agencies such as unified strategy, authorization and failure tolerance

·           Backup node is supported by backup software and backup storage products and connected with the backup management platform through relevant technology.

Backup Nodes of Multi-level Service

·           Data backup node

·           Data backup node is supported by backup software or backup storage products, supporting centralized backup of operating systems including Windows, Linux and Unix as well as file, mail and mainstream database application backup;

·           Enable unified management, authorization, hosting and remote data disaster recovery of the backup node through the management platform.

A More Secure and More Protective Platform

Greatly improve the security needs of distributed organizations' backup disaster recovery through a unique security backup technology.

·           Ensure the safety of remote backup disaster recovery through safe transmission and storage, and reduce data transmission amount of remote disaster recovery to save the band width through intra-node data deduplication technology;

·           Ensure and improve the management and operational security of the system by secure access and verification techniques;

·           Enhance the data protection level of distributed organizations and save the cost of unified planning and deploying by nodes' remote data and application disaster recovery.

Improve Management Flexibility

·           Make distribution and conduct unified management on centralized strategy according to the system center, which can simplify the management complexity of backup disaster recovery of multi-branch agencies;

·           Cope with the situation where branch does not have or need a separate backup disaster recovery administrator and improve management flexibility through the management of the entrusted platform of branches' backup system.

Flexible Scalability

·           Scale-based authorization model helps fulfill the needs to protect the changing environment, servers and applications. With the increase in capacity and scale, authorization of different scale can be smoothly upgraded to help reduce the risk of upgrading the platform;

·           Through multi-node connection, data backup nodes can be accessed. Different nodes can be switched according to certain demands and help administrators fulfill the needs of increasing node number and changing protected objects and data.

Complete and Forward-looking Backup System Transformation Program

·           The backup system transforming scheme proposed by EMC cannot only provide reliable and effective data protection for existing business systems, but also meet the backup needs after IT planning.

Deduplication Technology

·           solve the problem of the explosive growth of backup data;

·           Data Domain has the deduplication rate of 20 times more than the total data number, solving the problem of the explosive growth of backup data in the information system. Its smart cloning integration with Networker easily achieves remote disaster recovery, solving the conflict between backup and cloning operations, so that the business data can be protected in time.

Advanced Designing Concept and Reliable Structure and Hardware Device

·           The essence of Data Domain is a set of storage devices whose structure and performance determine the stability of the entire backup system and data security. The backend storage of Data Domain adopts the disk array of RAID6 technology.

More Secure Data

·           Data Domain has unique lossless data architecture including RAID-6 protective mechanism: each disk group has two spare disks as protection to protect the completeness and usability of the backup data through NVRAM power-off protection technology, continuous fault detection and self-repair technology of the file system.

Better Scalability

        ·           Data Domain storage device can achieve capacity expansion easily.