Network data backup and recovery

Network data backup and recovery

I. Reasons for data failure
　
Fault classification of computer systems and analysis of faults
　
Class A: computer hardware and software faults
　
• Occurrence: for enterprises, the most likely and frequent occurrence may lead to abnormal work, and serious loss of important data may lead to huge economic losses. This is a type of fault that often occurs;
　
• Solution: Local dual-machine hot backup is recommended to achieve system redundancy and enhance the high availability of the business system.
　
Class B: manual operation fault
　
• Occurrence: In enterprises with strict management and high personnel quality, such accidents are not common, but they often occur in enterprises with loose management and insufficient personnel training. However, no matter what the situation, as long as someone fails to operate, it will lead to a pause in the work, bringing a lot of losses.
　
• Solution: Improve the automatic operation management level of the company's systems, make cold backup of local data, reduce human intervention on operations, and train staff or formulate strict management specifications, avoid misoperation as much as possible.
　
Class C: planned shutdown caused by insufficient resources
　
• Occurrence: As the business grows rapidly, more and more software, hardware upgrades, system resource expansion, and other events occur for some enterprises, occurs more frequently;
　
• Solution: the company plans to upgrade the system and implement local dual-host and system redundancy for the original data and system.
　
Category D: disasters at production sites
　
• Occurrence: Some irresistible factors (such as natural disasters and wars) often cause huge losses;
　
• Prevention Method: Establish a disaster recovery center.
　
Disadvantages of traditional backup methods
　
A few years ago, we used a built-in or external tape drive to perform cold backup on the data. This method is applicable when the data volume is small, the operating system type is single, and the number of servers is limited, it can be said that it is an economic and concise backup method. However, as the computer scale of modern enterprises expands, the explosive growth of data volume and the rise of the distributed network environment, enterprises distribute more and more businesses on different machines, operating platforms, and even different locations. This stand-alone manual cold backup method is increasingly unable to adapt to today's distributed network environment, there are the following drawbacks:
　
1. It is difficult for data management to become systematic and data loss is inevitable;
2. Scattered Data on different machines and applications, scattered management, and insecure security;
3. It is difficult to achieve efficient online backup of database data;
4. The running system makes it difficult for maintenance personnel to move away, reducing the efficiency of business personnel;
5. Storage media management is difficult. Nowadays, more and more media are used to store data. The storage of floppy disks, tapes, and CDs in different systems will bring great difficulties to management;
6. It is difficult to retain historical data;
7. Potential risks from non-computer system factors, such as fire, earthquake, and other disaster-related system reconstruction and business data operations.
　
Objective of the Network Backup System
　
The ideal backup system should be comprehensive and multi-level. First, hardware backup should be used to prevent hardware faults. If the data logic is damaged due to software faults or human error, the network storage backup system and hardware fault tolerance are used to recover the fault. This combination constitutes a multi-level protection for system hardware and software, which not only effectively prevents physical damages, but also effectively prevents logical damages.
An essential part of network system security construction is regular data backup and historical storage. Generally, there are two main purposes of local backup in production: First, after the business data of the production system is damaged or lost due to system or human error, data can be restored locally in production in a timely manner. Another objective is to achieve regional disaster (earthquake, fire, machine destruction, etc, data and disaster recovery of the entire system can be realized locally or remotely in a timely manner.
　
Considering the safety of the local production environment, regular data backup generally requires that at least two copies of one copy of data should be stored in the production center to ensure normal data recovery and Data Query recovery, the other copy should be moved to remote storage to ensure the minimum data recovery after a disaster occurs locally. In addition, a remote storage system for historical archiving data should be established to ensure reliable recovery and effective audit of historical business data.
　
To sum up, the ideal network backup system that modern enterprises can possess should have the following functions:
　
• Centralized Management: The network storage backup management system manages the data of the entire network. With the help of centralized management tools, the system administrator can manage the backup policies of the entire network in a unified manner. The backup service can monitor backup jobs of all computers, and modify the backup policies in a timely manner, you can view all directories in real time. All data can be backed up to any tape library connected to the backup server or application server, thus providing maximum manageability.
　
• Automatic backup: For most data center managers, backup is a heavy and boring task. Small Enterprises are doing well, and large enterprises have a huge data volume, this requires the related personnel to be careful every day, And do not dare to lose anything, for fear that an error will lead to irreparable losses. Therefore, once the network backup can achieve regular automatic backup, it can greatly reduce the pressure on administrators.
　
The backup system can reasonably define the data to be backed up based on your actual needs, and then set the backup schedule in a graphical interface. The backup system will automatically start the backup job without manual preparation. This automatic backup job is customizable, including a backup job, a specific day of a week, and the day of a month. After the plan is set, the backup job is automatically performed as planned.
　
• Database backup and recovery: Many people think that databases and files are still a concept. Of course, if your database system is based on a file system, you can use the backup file method to back up the database.
　
However, since its development, the database system has been quite complex and huge. It is no longer applicable to backing up the database using file backup methods. Whether the required data can be extracted from a large database file for backup is one of the signs of advanced network backup systems.
　
• Online indexes: the backup system should create online indexes on servers for daily backups. When users need to recover, they only need to retrieve files or data from the online indexes, the system will automatically restore the file.
　
• Archive Management: You can archive all data on a regular basis by project or time. A Unified Open tape format data storage format is provided to ensure that all application data is permanently stored by a unified data format, ensuring the permanent availability of data.
　
• Effective Media Management: the backup system automatically adds an electronic tag to each tape used for backup, and provides the tag identification function in the software, you only need to execute this function to quickly know the content of the tape.
　
• HSM Hierarchical Storage Management: for industries that are easy to produce large amounts of data, such as the publishing industry and manufacturing industry, most of the data is a space-consuming image, in addition, each design draft and document often need to be online at any time. Based on management and cost considerations, HSM (Hierarchical Storage Management, Hierarchical Storage Management) system is a suitable online backup solution. It uses hard disks, Erasable magnetic discs, and tapes for three-tier storage management. The so-called Hierarchical Storage Management System is an automated network storage management device that automatically determines the frequency of data use in the hard disk and automatically moves the infrequently used data to a slow CD, the least commonly used data is moved to the tape, which is set by the system administrator. Online data can be optimized after a period of moving.
　
• System disaster recovery: the ultimate goal of network backup is to ensure the smooth operation of the network system. Therefore, an excellent network backup solution should be able to back up key data of the system and quickly restore the network system in the case of network faults or even damages. From fault discovery to full recovery of the system, the ideal backup solution should not take more than half a working day. • Meeting the increasing demands of the system: backup software must support multiple platform systems. When the network is connected to other application servers, for the Network Storage Management System, you only need to install client software that supports such servers on it to back up data to the tape library or optical disk library.
　
Network backup storage management system
　
Network Data Storage Management system refers to the centralized management of network-wide data backup in a distributed network environment through professional data storage management software combined with the corresponding hardware and storage devices, this enables automated backup, file archiving, hierarchical data storage, and disaster recovery.
　
To achieve automatic data storage management in the entire network system, the organic combination of backup servers, backup management software and smart storage devices is the basis for achieving this goal. The Network Data Storage Management system works by selecting an application server on the Network (of course, you can also configure another server in the network as a dedicated Backup Server) as the network data storage management server, install Network Data Storage Management Server software as a backup server for the entire network. Connect a large storage device (tape library and optical disk library) to the backup server ). Install the backup client software on other servers in the network that require data backup management, and centrally back up data to the storage device connected to the backup server through the LAN.
　
The core of the network data storage management system is the backup management software. Through the backup software's planning function, a complete backup plan and strategy can be established for the entire enterprise and call functions can be used for backup, all server backups can be performed at the same time. Backup Software also provides comprehensive disaster recovery methods, which can fully utilize the excellent features of backup hardware, greatly shorten backup and disaster recovery time, and realize automatic and intelligent management of network data backup.
　
Backup policy and recovery plan
　
The daily backup system describes the method and media used for daily backup, which is the detailed implementation rules of the system backup solution. After preparation, routine backup should be performed in strict accordance with the system; otherwise, the backup solution will not be able to be achieved. There are multiple data backup methods: Full backup, Incremental backup, differential backup, and on-demand backup.
　
Full backup: backs up all the data in the system;
Incremental Backup: only backup of data that changes after the last backup;
Differential backup: only the data that changes after the last full backup is backed up.
On-demand backup: Select data backup as needed.
　
Full backup takes the longest time, but the shortest recovery time and the most convenient operation. When the data volume in the system is small, full backup is the most reliable. However, as the data volume increases, we will not be able to perform full backup every day, but we can only perform full backup on weekends. We will use Incremental backup with less time or differential backup between the two at other times. Different backups have different data volumes: Full backup> differential backup> Incremental backup. You must flexibly use the backup according to their features.
　
Setting Basic data storage policies
　
• Full database backup: it is automatically performed on Friday (or Saturday.
• Incremental database backup: It is performed by Unix or other host systems before and after each batch every night, and is triggered by batch processing personnel or automatically executed by the system.
• Full file backup: it backs up the data of the host system and other servers. It is automatically performed on Sunday.
• Incremental file backup: backup file increments between Monday and Thursday (or Friday.
• Full System: full system and database backup on Sunday at the beginning of the month.
• System increment: only the incremental configurations of the system and database are backed up for the rest of the time.
• Tracking backup: Real-time backup of system increments (transaction log backup ).
　
In combination with the preceding backup policies, from the perspective of ease of management and recovery, formulate policies for data groups and storage media pools, and classify data on tape groups with different numbers, and create different access permissions. We recommend that you:
　
• Database medium: stores database information.
• File medium: Files other than databases.
• Database logs and System Log media: Important data records for security auditing and system recovery must be kept for a long time. It is recommended that security management officers establish management on the NT side, form a running data record separated from the host system administrator. • System medium: backup system and system configuration changes to quickly restore the system.
Data backup process
• The automatic backup process is initiated by the backup server. Every night, incremental or full backup is automatically performed according to the pre-defined schedule. As daily backups are properly balanced, the peak backup data volume occurs on Friday (or Saturday) and Sunday.
• Pre-approval and post-approval backups are initiated on Unix or other hosts. The batch processing personnel typed the BACKUP command to trigger automatic backup of database-related content as required.
• Free backup of other files. Go to the software interaction menu and select the file to be backed up and back up.
• Track backups online. With the online database backup function of the data storage management software, you can define real-time or scheduled log backup.
• Cloning of Disaster Backup remote storage media. Automatically copy the data after completion every day to store data for disaster recovery in different regions.
　
Disaster recovery
　
Disaster recovery measures play an important role in the entire backup system. It is related to the ability of the system to quickly recover after a disaster. Disaster recovery operations can be divided into two types. The first type is full recovery, the second type is individual file recovery, and the other value is to mention redirection recovery.
　
• Full recovery: full recovery generally occurs when an unexpected disaster occurs on the server, resulting in full data loss, system crashes, planned system upgrades, and system restructuring. This is also called system recovery.
　
• Individual file recovery: due to the low level of the operator, individual file recovery may be much more common than full recovery, using the recovery function of the network backup system, we can easily recover damaged individual files. You only need to browse the backup database or directory, find the file, and touch the recovery function. The software will automatically drive the storage device, load the corresponding storage media, and restore the specified file.
　
• Redirection recovery: The redirection recovery refers to restoring the backup files to another different location or system, rather than the current location during the backup operation. Redirection recovery can be system-wide recovery or individual file recovery. You need to carefully consider the redirection recovery and ensure the availability of the system or file after recovery.
　
To prevent data loss, we need to prepare a detailed disaster recovery plan and conduct regular disaster drills. A disaster drill should be conducted every time. Disaster simulation can be performed using obsolete machines or redundant hard disks to familiarize yourself with the disaster recovery procedure and check whether the generated disaster recovery floppy disk and disaster recovery backup are reliable.