Project background:
Most modern large enterprises have a large number of servers that provide the users of Internet and Intranet with a variety of different services, such as data base or Email system. Maintaining the business continuity is an important aspect to consider when current enterprise users store data. System failure may cause stoppages in production. Customer satisfaction decreases, so companies may even lose customers, which will result in the decline in the competitiveness of enterprises. Therefore, maintaining the business continuity is an important index when choosing computer system. This is because an important way of maintaining the business continuity is to improve the high reliability of the computer system and minimize the loss of data at the same time.
Now the Maternal and Child Health Hospital has a machine room in both Fanyu and downtown of Guangzhou. The business operated in these machine room are the core of hospital business system, and the backup operation of these key applications and data base have become a part of daily operation process to ensure the important data can be recovered in time. Traditional solutions, such as cold standby or backing up with tape driver has many obvious defects and can not recover data and services in real time. With the upgrading of industry needs, the development of technology and the consideration of information security, the management and construction goal of disaster recovery and business continuity should at least include the following aspects:
Guarantee of business continuity: recovery time objective~0, let the system rapidly recover business system and provide services after a failure.
Anti - calamity: It is required to implement remote disaster recovery, resist large-scale disasters of various regions, and be convenient to do disaster drills.
Safe and reliable: Disaster recovery system will get all of the user’s data completely. For those data and units with confidentiality requirements, data security is particularly important.
Simple and easy to use: Simple and easy to use is the necessary condition of disaster recovery construction. Otherwise, the companies spend much money, but no one can use it successfully.
Reasonable TOC: easy to maintain and upgrade. It should be the system which commonly used in China.
Therefore, Maternal and Child Health Hospital of Guangdong Province backup these key application/data through 10M line to the five-star computer room of Guangzhou Unicom in real time, and implement the real time backup protection for data and application.
1. Requirement description
As the importance and strict requirement of related systems, it’s good to use disaster recovery software to protect the related data of Maternal and Child Health Hospital. That means using the backup software on the server to replicate data in distance through IP network. The data and application are required to be backed up to ensure the data protection in real time and HA of application.
Backup software should support unified Chinese interface of monitoring/configuration in WEB mode, and it should also support for multiple operating system platforms (contain at least windowns, Linux) data file backup and protection, a variety of database backup and protection, application of real-time protection and abnormal takeover , remote disaster recovery and so on. Backup software should support byte level incremental data bypass capture mode, and log functions through software backup, formulate corresponding strategies and process for backup data, including start-up time, storage path, flow control, backup mode, state management and so on, then the number of according to the update action through the network faithfully copied to local or remote backup system. When the local backup system failure, you can recover the data through the remote disaster recovery system. And you can also be used for data CDP (continuous data protection) method of protection, abnormal circumstances, the data can be recovered to any historical point of time.
I2Soft using advanced asynchronous real-time data replication technology, immediately on all servers for disk system fine-grained byte level change through network transmission to the standby server, rather than the entire file or disk mirroring, so impact on server performance and network bandwidth can reduced to the minimum. Once the backup center took over the service, users need not recover data, can in the shortest time to activate the system, prepared by a machine to provide foreign service. Customers will minimize the cost,achieve real time data protection,and maximize the protection of data real-time, integrity and consistency if they use i2soft. Business data is one of the most valuable assets of users, the loss of data is the loss of corporate assets, so the protection of business data is one of the main functions of enterprise computing system. The disaster recovery scheme of I2Soft implementation can let user data loss to a minimum or even zero.
In order to achieve for all system disaster recovery under the limited bandwidth , combined with the customer's specific requirements with reliability, efficiency and safety as the top priority by using I2Soft software. The detailed description of the program design and implementation of the specific details of the program.
2. Program Description
2.1 Architecture Review
In the view of actual conditions and above requirements, i2Soft provide the following situation. As shown in the following topological graph:
As the above programs
I2 software is installed on the corresponding master servers and standby servers to protect related data in real time. The data is backed up in real time, and related data can be rapidly recovered when an exception occurs. The production servers include the business servers in the data center of Fanyu and Guangzhou. The disaster recovery system deployed in Guangzhou Unicom can provide rapid data recovery service in exceptional circumstances, and the recovered data will be quickly sent to a selected service(the original server or a new server). At the same time, the data can be used normally.
As follows, when a server break down and need data recovery, the related data can be rapidly and directly recovered from the computer room of Guangzhou Unicom.
When the related core system of Maternal and Child Health Hospital of Guangdong needs application level high availability protection, we can also deploy one to one high availability mode. As follows, when the system is abnormal or goes down, its corresponding disaster recovery server in computer room of Unicom will take over its application, and the IP address of the source server can be chosen whether to follow the drift to the disaster recovery server, which implement business continuity, that is the high availability at application level:
Therefore, the above program can implement multilevel disaster tolerant protection:
Centralized data protection: when any exception occurs at the source side, the related data will not lost and can be rapidly recovered from the Unicom computer room.
Application level disaster recovery protection for core business system(more advanced one): It can implement the business continuity, when the business of source side can not be accessed, the corresponding disaster recovery server in Unicom computer room will take over the business and provide service to the public.
The two level of disaster recovery protection can coexist, the protection of a server can be upgraded smoothly from data level to application level without any effect on other systems. Therefore, the total disaster recovery topology is as follows, there are both data backup protection for common server and application level HA protection for key system.