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

In this case, the failed storage type is HP P2000, the virtualization platform is vmware exsi, and there are 10 hard disks composed of raid5(hard disk capacity is 1t, and disk 6 is a hot spare disk). Due to some failures, two hard disks in the array are lit yellow and offline. The serial number of the hard disk cannot be read, and the hard disk cannot be read on the SAS expansion card.

After the failure occurs, the user realizes that the failure is serious, and must bring the RAID-5 disk array to our company for detection to determine whether the failure of each hard disk in the array is a logical failure or a physical failure. At the same time, link other normal hard disks to the mirror server and use the mirror tool to perform sector-level mirroring.

First, connect the bad disk to the external SAS expansion card. After power-on, it is judged that the hard disk motor can start rotating according to the working sound of the hard disk, but the magnetic head does not perform seek operation. Therefore, try to separate the hard disk PCB and clean the oxidized part of the HDA component. After restoring the PCB, the fault remains. Therefore, after communicating with the customer to replace the good PCB of No.6 hot spare disk with the faulty disk for trial repair, and then replace the ROM chip on the PCB of the faulty disk with the good PCB of No.6 disk, the hard disk starts to rotate and the magnetic head seek sound is normal, but after the seek is completed, there is obvious disk knocking sound, so it is judged that the magnetic head may be damaged. After communicating with the user, try to replace the failed disk with the good head in the hot spare disk No. 6 to read the data. After the operation of opening and replacing the magnetic head of the faulty disk in the clean room, frombyte.com connected the faulty disk to the professional hard disk maintenance work for detection, and found that the faulty disk could not be identified and the data could not be read.

Because the user has two failed disks, one of which was tried to repair before, communicate with the customer again to try to repair the other failed disk. As with the previous failed disk, the disk failure was still head damage, because the user's HP OEM disk was expensive, so he tried to buy the same model of ST original hard disk online for head replacement. After the head replacement of this hard disk is completed, the device can recognize the hard disk normally, so all the sectors of the failed disk are completely mirrored to a backup disk of the same capacity.

3. Reorganizing RAID-5 Steps

[Determine the starting sector] All hard disks can be reorganized after mirroring, open 9 disks with WinHex (hot spare disk does not need to be added to the reorganization), and first parse the image file into a disk. You can see that sector 0 of these nine disks has the "55AA" mark, as shown in Figure 1.

Fig. 1

The search result is shown in Figure 2. 0x01C2H indicates the type of the partition. Here,"05" is displayed, indicating that this is an extended partition. So from sector 0 this is an abnormal MBR partition structure.

Fig. 2

Continue to look down according to Figure 1, and find the "55AA" mark on disk 9 and disk 8 respectively. The query results of No. 9 disk are shown in Figure 3. This is a normal MBR partition with a value at 0x01 C6 indicating that the next sector is the head of GPT.

Fig. 3

The query results of No. 8 disk are shown in Figure 4. The value at 0x01 C6 points to the next sector. But the next sector is clearly not the head of GPT.

Fig. 4

From this, it can be determined that disk 9 is the first disk, and disk 8 may be the last disk. The sector where the GPT partition is located starts at sector 172032, so it is tentatively determined that the LUN starts at sector 172032.

Stripe size: Stripe is also called block, which is the basic unit of RAID processing data. Stripe sizes vary from RAID to RAID. RAID-5 has one parity area in one stripe group, and the size of one parity area is equal to the size of one stripe. According to this point, this RAID-5 case is analyzed. If you are unfamiliar with VMFS file systems, stripe sizes can be determined by comparison. For example, if the verification area in a band group may be significantly different from the non-verification area in this band group, you can find the band size by viewing and comparing WinHex. In this case, it is determined that a stripe is 1024 sectors.

[Judge RAID-5 Member Disk Order] Divide according to 1024 sectors, so that one record is the size of one stripe, as shown in Figure 5. And 9 discs jump to the same record 283123.

Fig. 5

When all 9 disks are located in the same position, the direction of the verification area can be judged by comparison, and then the direction of the entire RAID-5 can be judged. Before it has been judged that drive 9 is the first disk, put drive 9 in the first position, and then you can judge the direction, as shown in Figure 6 (drive9 is the fourth disk). RAID-5 is determined to be left-sided, and the disk order is 9, 2, 3, 4, 10, 1, 7, 8, 5.

Fig. 6

The previous content initially determined that the starting sector of the LUN is sector 172032. Use WinHex to jump to sector 172032 and observe the actual situation of each hard disk. If sector 172032 is the starting sector of the LUN, disk 5 in the stripe to which this sector belongs should be the parity area, but disk 8 in this stripe shows that disk 8 is the verification area. According to the left direction of RAID-5 in this case, disk 5's parity area should be in sector 172032-1024=171008, i.e., the previous stripe. Jump to sector 171008 and find that the check area is disk 5. Therefore, it can be determined that the LUN starts with sector 171008.

[Reorganize RAID-5] Use a professional recovery tool to group and add the disk according to the determined disk order, as shown in Figure 7. Select RAID-5, Stripe size 512KB, Left Asynchronous.

Fig. 7

Click Build to regroup. After the group is completed, since the data starts from 1024*8=8192 sectors, if the professional recovery tool does not have the function of jumping to this sector, then the newly organized RAID must be reorganized with a file. Select 8192 as the RAID start sector (Startsectors_frompyte.com). You can select the start sector and Count sectors of this file arbitrarily, as shown in Figure 8 and Figure 9. Figure 10 shows RAID-5 assembled.

Fig. 8

Fig. 9

Fig. 10

4. handover data

After the entire RAID-5 is rebuilt, our company will contact the user for acceptance of the data. The user will confirm that the data is OK. After signing the acceptance contract, the complete RAID-5 data can be handed over. When handing over, the data is handed over to the new disk brought by the user according to the user's request. The recovered data is kept on our company's server for 3 days, after which the data is automatically destroyed by the system.

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.