Changzhi electricity supplier morning received a telephone company for help, a DELL-R430,3 disk 1TB, install CentOS, the system can not be started after an unexpected power failure, the aircraft is equipped with the company's OA, as well as marketing data. Environment LNMP - Nginx + MySQL + PHP server architecture Linux system and requires on-site field service
Telephone recovery program and given the price, the other party to consult with the leadership decided to bring the machine to our recovery. In fact, to our first recovery can save a lot of off-site costs, and secondly in the process of data recovery if there are other unforeseen circumstances are all good solutions, such as equipment, resources, etc., unless the data confidential!
Our clients reach already 17:00, and according to standard procedures:
To mirror each one physical disk. Its purpose is to ensure data security, and ensure that the most original state after a failure, and the subsequent working line system is to use the original image.
Mirroring the original disc and then divided into two steps simultaneously, this is our advantage:
Software engineer to analyze the image file, the method of determining the arrangement of the RAID group, the size of the block, to the other.
Analyzing hardware engineers with the original hard disk damage state machine, in conjunction with software RAID technology information given by the dirty plate removed, when properly configured to make the last normal system startup.
The system starts normally, the data recovery is complete, if they can not start the first group of software engineers to verify whether the data is correct.
After the image is complete, we will be on the hard disk starts up, first enter the RAID configuration program, it displays a raid5 volume, and has been downgraded newspaper yellow, which has a MISS state, RAID5 which has normally a disc miss is normal start, then we reboot into the system to see what problems
See a normal system boot interface, then that still remain in force in the RAID, but his system was being given.
Analysis of the error message, find a lv home alone mount, then we use a command that attempted to fix it
Fsck –y
After a bunch of code hopping after we reboot reboot after entering the command.
This picture is a bit familiar with it? Yes, we have the system repaired. After configuring the network environment, perfect viewed online
I wrote these words more than 10 minutes with, but in fact our entire work has been completed after 5:00 the next morning, the whole 12 hours, so-called "stand for ten minutes, the audience ten years of practice."