Shulou(Shulou.com)06/03 Report--

Description of server data recovery failure

The physical condition of the failed server is introduced as usual before introducing the data recovery case. The server of this data recovery is linux operating system, some brand 730 series server, MD3200 series storage. The reason for the data loss is that the system can not start normally due to the unexpected power outage in the computer room. after repairing the inaccessible server, the customer administrator enters the system to view the data, and some of the server files have been lost. So the customer administrator contacted the data recovery center for server data recovery.

Second, failure analysis of server data recovery.

1. Back up customer server data

After receiving the customer's server, the data recovery engineer first checked the server and first mapped the stored lun in read-only mode to the data recovery dedicated storage device in the data recovery center. Then mirror the client's server at the sector level for two purposes, one is to allow the customer to retrieve the original server to prevent the later data recovery process from occupying the client's equipment, and the other is to protect the customer's original data, because data recovery requires a lot of data analysis and attempts, so it is possible to try many times. In this way, the data recovery operation in the image file can avoid the operation on the customer's original server and protect the integrity of the customer's original data. In case our recovery fails, customers can also bring the original server to other companies for data recovery operations, which is a way to be responsible for customer data.

2. Analyze the cause of server failure

After the backup is completed, the server data recovery engineer looks at the underlying data and finds that the directory entries of the server have been corrupted. Fortunately, the destruction of these directory entries does not affect the important data of the server, but only destroys some of the directory entries, which can be repaired manually. Because the server administrator of the customer has carried out a repair operation, this leads to the fact that the damaged directory items are not repaired successfully, but are named after the node number and stored in the lost+found folder, and the corresponding datazone index is automatically cleared by the server. Engineers have handled many similar cases of data loss in the past, in which case it is only necessary to match and piece together fragments in free space according to the file system and file type, and finally recover the data of the entire server.

III. Implementation process of server data recovery

The node information in this data recovery case has been cleared and the data cannot be restored according to the node information. The server data recovery engineer extracts the file names under the lost+found folder and matches them one by one according to the file directory item node numbers of the missing files, so as to analyze the missing directory structure. Then continue to analyze the underlying data, scan and extract the information that meets the lost directory structure conditions in the corresponding position of the underlying space according to the structure information of the file system, then integrate with the directory item node number, record the scanned directory item node number into the database, and then match the file record number in the lost+found with the record number in the database.

4. Server data recovery results

In this server data recovery case, the client's server was damaged by an abnormal power outage, and then the file system was manually repaired by the administrator, resulting in the loss of the directory structure of a large number of files. in the process of repair and inspection, the server also wrote some new data, which directly led to the server data recovery process is more complex than the normal data loss. As the engineer has many years of experience in server data recovery, after a period of analysis and reorganization, finally successfully extracted the data lost by the client server. The customer validates the successfully recovered data. After verification, all the data in the original server of the customer are successfully restored and can be used normally. The customer agrees with the results of the server data recovery.

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.