Shulou(Shulou.com)06/02 Report--

Principle of server Raid 5 disk array algorithm

The independent disk structure of distributed parity (what we call raid 5) data recovery has a concept of parity that needs to be understood. We can simply think of it as an "XOR operation" in binary operations, and the commonly used logo is xor. The rule of operation is that if the values of the two are the same, the result is 0, and if the results are different, the result is 1.

What to do after a server disk array failure: (other raid arrays, including raid5 disk arrays, are similar)

Do not blindly force online

When prompted by the Offline indicator, the user is required to replace a good and identical hard disk and recheck the algorithm. Several of the multiple hard drives that RAID5 disk arrays are likely to purchase at the same time have experienced instability. The server's Offline prompt may not be the most accurate, often misreporting (there are also factors of disk instability) or omitting other hard drives that are also potentially dangerous. In the process of forced online, the hard disk needs a lot of read and write operations, and once other hard drives suddenly fall off, it is more difficult to recover data. Therefore, it is safer to check the stability of each hard drive before being forced online.

Protect the first scene.

For manufacturers, hardware equipment is the main purpose of after-sales service, so data is often sacrificed to ensure the integrity of hardware equipment as soon as possible. According to the engineer in charge of the data recovery center, cases of complete destruction of the final data due to the repair of the after-sales service engineer of the data recovery center occur almost every week. After all, after-sales service engineers are basically biased towards hardware, while RAID data recovery is based on software operations. In addition, even simple verification operations may cause inconvenience to later data recovery. When the raid hard disk damage exceeds the fault tolerance, the image of the damaged hard disk must be completely extracted before the final recovery can be carried out. If the hard disk is energized frequently, it is easy to cause further fatigue damage to the magnetic head, or even scratch the disk. The correct way to deal with it is to consult the professional data recovery service provider immediately.

Select a professional data recovery center

When choosing a data recovery service provider, the company size, technical level, engineer team and word-of-mouth in the industry are all important indicators to be considered.

Data recovery method after erroneous deletion of VG from server raid disk array:

Raid disk Array data recovery case background:

The failed raid array synchronized the data after replacing the disk, and the VG information was deleted and the data synchronization was aborted. The RAID is a new RAID array after the backup VG information is imported by the operation and maintenance staff. generally speaking, once the data is synchronized, the new raid array will be synchronized into a blank disk and the data cannot be recovered. Fortunately, the technical staff of the manufacturer turned off the data synchronization function for the sake of safety, and it was this operation that saved the data, otherwise the data of the server would suffer irreversible damage. The customer chose the North Asia data recovery Center after comparing several data recovery companies in Beijing.

Raid disk array data recovery steps

1. All the disks in the server disk array are mirrored separately to the secure storage of the North Asia data recovery Center, and the original disk will no longer be operated to ensure the originality of the user's disk.

two。 Analyze the underlying structure, reconstruct the original RAID, virtual reorganize the RAID, and mirror it to the temporary storage of the North Asia data center again.

3. Recovery using VMWare ESX data recovery software: Frombyte Recovery For ESX developed by North Asia data recovery Center

4. After seeing all the data normally, hang the recovered storage to the ESXi server of the North Asia data recovery Center and verify all the data.

Server disk Array mdisk disk offline data recovery method:

Description of Raid disk offline process

There are eight groups of Mdisk in the server, one of which has a hard disk failure offline in one set of Mdisk. In the process of enabling and replacing the hot spare, another hard disk in the same group Mdisk is also offline, resulting in the synchronization failure of the hot spare and the invalidation of this group of Mdisk, resulting in the failure of the whole general-purpose volume.

Server disk array data recovery method:

1. I mirror each disk through the mirroring software, and all the data recovery work will be carried out on the mirrored disk in the later stage. The original disk always maintains its original state and does not affect the data. The main purpose is to prevent misoperation in the process of data recovery, once the operation error will lead to the second destruction of the original disk data, and the later data recovery will be more difficult.

two。 To analyze and reorganize Mdisk, first of all, according to part of the configuration information provided by customers, the hard drives are classified according to the Mdisk group. Analyze all the hard drives in each set of Mdisk to get the relevant raid information. Use professional data recovery software to carry out virtual reorganization of Mdisk.

3. Analyze all the Mdisk and get the relevant information of pool. Use professional data recovery software to virtual reassemble pool.

4. Drop analysis. Because of the inherent characteristics of raid5, we know that raid5 can allow at most one member disk to go offline, that is, if a member disk fails, raid5 can still be used normally. Therefore, combined with the previous raid failure, it can be judged that two disks failed at that time.

After analyzing the bottom data of each member disk, it is found that there are two hard disks, and the data in the front of the hard disk is exactly the same, so it is judged that one of the two hard disks is a hot spare disk and the other is a failure disk.

According to this, it can be preliminarily inferred that one member disk in raid fails first, and in the process of synchronization of hot spare, another member disk also fails, resulting in raid failure. When the later raid returns to normal, because the data on the hard disk that first failed is no longer up-to-date, the data in the normal raid appears some errors.

5. After the analysis of the raid structure of the hard disk and the drop of the member disk, the professional data recovery software is used to complete the construction of Mdisk and pool, and generate the data in it.

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.