Server data recovery engineer to explain the raid storage structure principle

Before all official content begins to ask a question, what is raid? The concept is described raid abound on the Internet, are a bit too professional. I want to describe the most simple principle, it is not allowed to affect all the data in the case of the missing part of the data when defining storage, similar to the error correction code field of communications. Different redundancy modes form different RAID levels. We need to describe RAID models only have one disk redundancy (thinking with RAID3, RAID4, RAID5). Suppose there are three blank sheet of paper, to record some figures, in order to more clearly record, we draw a blank first page of some form of the same size. Suppose there is a possibility: there will be three pages under specific circumstances where a page is missing. In order to ensure that the set of these numbers can complete safety record, we have to design some redundancy relationship with each other implicated. If we are to be recorded sequence of numbers: 3,14,28,4,98,88. We can turn on page 1 and page 2 write the number to be recorded, on page 3 and write on them. So no matter which of the three missing pages, can be calculated by the data another two another page. Obviously, the situation is even more than three, according to the above design recording mode, it can also support any one loss record. This is the principle of data storage of the raid.
Well, let's get to the point, if a raid array while three hard drives dropped, and how we should deal with in order to avoid data loss it? Here is the North server data recovery data recovery center a real case raid6 array data recovery engineers of recovery: the
failure of the need for data recovery is a set of six 750G disk RAID6 disk arrays, servers in the normal course of operation, there have been two disks offline, but maintenance personnel in this case is still no replacement disk, hard drive offline after the third raid array on natural collapse, the server was paralyzed, and therefore business stagnation. Replacing the hard disk you want to rebuild the array is not a problem, but all this data set in the array will be lost, there is the resulting loss will be enormous.
Server administrator for data analysis and data recovery after a portion of the back, but still most of the data loss can not be recovered, so the customer contact to the Beijing North data recovery center data recovery operations. Client server is a web server, mount the hard disk 6, when three hard disk is offline. Run the MySQL database and some of the more messy other types of files. Server data recovery engineer first hard disk, including all the normal hard disk in the server, including a complete mirror, and then the client server returned, the remaining work is to analyze the image and restore the data.
For the hard disk server has dropped the pieces, the hard drive must be dropped one after the other, after all, while there are three hard disks dropped the chance is negligible. So for an array of servers have multiple hard disk data recovery principle is dropped off-line analysis of the hard drive in the order, repair hard drive failure last off-line, off-line and then extract the last of the hard drive data, analyze some of the basic parameters restructuring storage structure according to the raid raid It can be. Raid patterns that occur in this case is a data recovery raid array, RAID6 array using the dual-mode check, a check is first ordinary oxr i.e. XOR operation, the second check is a reed-solmon algorithm. Data recovery is usually adopted when the first check method to reply, but after analysis found an array of server data recovery engineers in front of two hard drives have long been dropped, do not have the conditions for data recovery, so the first school test mode apparently has no use. The second check mode and more complex, there are also no mature software or tools you can use, so a large extent missing customer data administrator of recovery, and the database can not be used.
Fortunately, however, data recovery center in North Asia have some core technologies reed-solomon algorithm, also had a corresponding experience in data recovery, server data recovery after more than a common analysis engineers finally acquired the necessary parameters customer raid array and extract the full It mirrored data. After the customer administrator to verify that all the data is successfully restored, the database can use this data recovery a success.
Important data to do data backup, virus protection do the work, found that data loss must protect the hard disk failure. Looking for data recovery company best to find some professional data recovery company that can maximize the protection of data recovery success rate.