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

Server is one of the basic configuration of enterprises, regardless of the size of the enterprise will almost configure a server for their own use. Well, while servers are becoming more common, server data recovery and server O & M services have not progressed and become popular at the same time. Today, I will tell you how to recover data if there is a bad hard disk in the server. The following is a real data recovery case of a server hard disk failure: There are multiple hard disks in the failed server that form a raid disk array. One day, the server suddenly crashed, and the administrator found that there are already two hard disks in the array. The status light is red (if the administrator often checks the server status and replaces the first hard disk failure in time, the current failure will not occur). At this point, the administrator did what most people do: restart the server. Because the server has self-test function, after restart, the offline hard disk is online and the automatic data synchronization starts. During the data synchronization process, the administrator forces the server to shut down (-_-!) In this case, we first consider whether the data in the hard disk is complete from the perspective of server data recovery, so mirror all the hard disks in the server (including those that have been dropped and those that have not been dropped). In the process of mirroring, hardware data recovery engineers are involved to detect hardware. If there is hardware failure, hardware data recovery engineers will repair and extract data. After the hard disk data backup is completed, the server data recovery engineer analyzes the mirrored data, reorganizes the raid structure and performs XOR verification. The verification result shows that part of the data is damaged. The reason for the destruction is that the administrator restarted the server and the offline hard disk was synchronized, which caused the data to be incomplete.

see Figure below

Server data recovery engineers tried many different methods for data extraction, but the results were all the same,

See below:

The way to extract data directly from the server did not work, and the data recovery engineer decided to change the way of thinking and scan and analyze the data files. However, due to the synchronous operation of the server, the directory of the data files was destroyed and could not be seen.

See figure below.

Due to directory loss, server data recovery engineers scan the free partition space data, scanning out a large number of file fragments, although the fragmentation is an extremely boring and cumbersome work, but other data recovery methods are not feasible, also had to be so. After a long process of analysis and aggregation, the data was finally extracted, and the newly extracted data was verified again and found to be fine.

The data recovery engineer informed the customer administrator to accept the data, and all the data passed the verification. The server data recovery was 100% successful.

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.