Shulou(Shulou.com)05/31 Report--

This article will explain in detail how to carry out the loss and disaster recovery of big data. The content of the article is of high quality, so the editor will share it with you for reference. I hope you will have a certain understanding of the relevant knowledge after reading this article.

As we have all seen over the years, natural disasters such as earthquakes, hurricanes, tsunamis, snowstorms or other natural disasters can destroy corporate data centers and other facilities in a short period of time. So a complete disaster recovery plan is a matter of life and death for the enterprise, because the loss of critical data will be a fatal blow to the enterprise. Disaster recovery is an important subset of business continuous availability, which is more and more concerned by IT senior managers. At the same time, it is also related to the development of major technologies in the IT industry, such as cloud disaster preparedness and security recovery.

I. hardware failure

Hardware failure is the number one culprit causing data loss; therefore, the neglect of hardware failure has also become the primary factor endangering business data. For example, if the hard disk is damaged, some people say that I have RAID5, but unfortunately, if two or more pieces of your hard disk are broken at once, RAID5 cannot guarantee that your data will not be lost; there are also corporate storage devices, which generally have A controller and B controller, and the two controllers are redundant, but once the B controller broke down and the storage really hung up. Although the B controller was later replaced and the storage was repaired, part of the data was lost. There is also the tape library that has been used. NBU backs up the data to the tape, and every time you find that the backup is successful, but when you restore, you will really encounter problems with some of the recovered files and need to restore this file many times; very often, when you change the power supply of the server, it will be restarted; when you do the stress test, the server will suddenly shut down. All these are not enumerated one by one. Operation and maintenance staff are really depressed when they encounter such a problem. The hardware is broken, resulting in data loss, but if you fail to use backup 100% recovery, you are also sad. DBA often says that backup is more important than everything, in fact, you should also tell him that backup is not omnipotent, of course, without backup is absolutely impossible.

Many enterprises use tape as backup media, because the tape itself has many congenital defects, especially the low reliability, there are certain hidden dangers, once a tape in the backup data is physically damaged, that will really make the data disappear and there is no way to recover it.

SAN or NAS storage devices are currently a good choice, but please note that in general, enterprises are a storage device, and not many people have backup storage or remote disaster recovery. And the construction of backup storage and remote disaster preparedness is also expensive. The NetApp storage solution of the US domain provides seamless storage management for the open network environment, which can be referenced, but it is also expensive.

II. Human error

Human error is the second leading cause of data loss. Human errors can take a wide range of forms, from accidentally deleting records to shutting down or restarting the system in the wrong way in defiance of management policies. People are born with inertia and often show a surprising tendency to degenerate in consciously abiding by policies. Although everyone knows the importance of information security management policies and data security, accidental deletion of documents and records is still inevitable.

The best weapon against human error is automation and backup. Under the creation and implementation of policies and procedures, automation technology avoids the contact between IT infrastructure and employees to the greatest extent, while backup becomes an important tool for data recovery. However, data backup is not the best way to achieve the high availability of enterprise IT systems, because recovery through backup usually requires some form of replication. High availability is effective in dealing with hardware failures, but it is difficult to recover logic errors caused by human factors.

It is indeed necessary to promote automation and standardization in operation and maintenance. similarly, in the era of cloud computing, the degree of automated operation and maintenance will be higher. Hardware, network and system services can be standardized and automated, and one person can manage thousands or even tens of thousands of servers in the cloud, which greatly reduces labor costs and human error. Both private and public clouds can achieve such standardized, automated and efficient operation and maintenance, and it is up to everyone to consider the pros and cons.

III. Software crash

It is said in the original text that software crash is the third largest cause of data loss. For example, everyone is no stranger to the blue screen crash of WINDOWS, and every friend who has used the Windows system is more or less troubled by this kind of accident. Of course, in addition to internal design defects, the causes of software crashes are often dragged down by system running errors.

A software crash, similar to human error, is a logical failure. For example, the software is not rigorous in design, the speed of development, there are many BUG, and it is put into use without strict testing, and it is usually found that the resulting information is damaged or lost after days, weeks, months or even years. Therefore, it is very important to strictly abide by the quality management policy and test acceptance before going online, and to build a reliable protection system for data by using automated testing and safety evaluation technology.

IV. Computer virus

Computer viruses not only seriously threaten the business system, but also have a bad impact on corporate reputation. We need to ensure that all devices in the business environment, including network devices, operating systems, databases, applications and storage, as well as personal computers and mobile terminals in the company, are installed with antivirus software and checked regularly, as well as backup servers. In this way, we can ensure that when the virus is rampant, it will not affect their own business environment and office environment.

Not to mention the external information security and * *, the office computers and mobile terminals of the company's internal employees should attract the attention of enterprise security engineers. Once the office computers and mobile terminals of internal employees run on the same system platform as business software, ruthless violence is bound to follow. Therefore, all kinds of data leakage and loss are also challenges faced by operation and maintenance engineers. In order to avoid risks, it is necessary to control and audit the information.

Fifth, the psychology and mentality in the face of disaster

It is difficult to count natural disasters as the main cause of data loss. Data loss caused by disaster events accounts for only 1% to 3% of all failures each year. Maybe within a year, there will not be a major data loss and disaster recovery scenario, but this does not mean that there is no unexpected disaster, so we can not always take chances with the disaster. Although major accidents will not occur frequently, once they occur, the consequences will be extremely serious. Even with a ratio of 1%, there is a 9/10 chance of data loss.

In addition, we should not take for granted the recovery after data loss. It can usually be said that this relaxed and lazy attitude, thinking that the plan will work, will hit us in the face at an important moment. For example, today's virtualization technology gives you more flexibility in the use and deployment of computing resources, but virtualization may sometimes give you a fake sense of security, and you may not want to make the right disaster recovery plan. I think everything can be done with virtualization. Be aware that virtualization does not completely replace the need for proper disaster recovery planning and testing.

In the face of the disaster of data loss, we must prepare in advance. No matter which recovery technique you use, the most important thing is to conduct recovery tests and drills on a regular basis. With backup, we are not always able to recover valuable data from backup files. The operation and maintenance staff are always faced with unexpected risks, and finally advise everyone that they must formulate detailed emergency plans in advance, test repeatedly, conduct more recovery drills, and consider a variety of different possibilities in order to gain insight into the nature of the problem.

On how to carry out the loss of big data and disaster recovery is shared here, I hope the above content can be of some help to you, can learn more knowledge. If you think the article is good, you can share it for more people to see.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.