[server data recovery] HP StorageWorks series storage RAID5 two disk failure offline data recovery case

Server storage data recovery environment ：

HP StorageWorks Family storage ;

VMWARE ESX Virtual host platform ,VMFS file system ;

8 Block hard disk composition RAID5,1 A hot spare .

Server storage failure ：

When the server is in normal operation, the two hard disk indicators display yellow , The server administrator detected this 2 The hard disk is shown as ： Unable to read serial number , stay SAS Unrecognized on expansion card , Need to be right raid Data recovery . The administrator contacts our data recovery center for server raid Data recovery .

Server storage data recovery process ：

1、 Hard disk physical fault repair .

First, you need to determine the Group raid The reason why the disk dropped . The hardware engineer will 2 A fault disk is connected to an external SAS Power up detection on the expansion card , It is found that the magnetic head does not seek . The hardware engineer will PCB Check the hard disk after separation HDA Components , Partial oxidation was found , After cleaning the oxidized part, it is found that the magnetic head still does not seek the track after being powered on again . So the hardware engineer can only open and repair the faulty hard disk .

The hardware engineer unpacks the normal hot spare PCB Replace the faulty disk PCB, And then put the fault disk PCB Upper ROM Chip removal and installation to new PCB Repair on . After the hardware repair of the faulty hard disk is completed, it can be tested again and start normally , The sound of magnetic head seek is normal , But the sound of tapping is obvious . The hardware engineer will find another magnetic head to replace , The hard disk can be recognized normally , The data of the failed disk is mirrored successfully .

2、 restructuring RAID5.

Use tools to open 7 Disk found all hard disks 0 Every sector has “55 AA” sign ,0x01C2H Represents the type of the partition , This shows “05”, Represents that this is an extended partition . So from 0 Look, this is an abnormal MBR Partition structure . Follow this method to continue searching , Respectively in 7 Plate and 8 The number plate has been found “55 AA” The logo of .8 The query result of the number plate is shown in the figure below . This is a normal MBR Partition , Its 0x01C6 The value at represents the point

The next sector is GPT The head of .

7 The query result of the number plate is shown in the figure below . Its 0x01C6 The number at points to the next sector . But the next sector is obviously not GPT The head of .

From this, we can preliminarily judge 8 Plate number is the first plate ,7 The number plate may be the last one .GPT The sector of the partition starts from 172032 A sector , Therefore, it is preliminarily determined that LUN The starting sector of is 172032 A sector .

raid Little knowledge ： A strip is a strip raid Array block , It is RAID The basic unit of processing data . Different RAID The stripe size of the array is also different .RAID5 Of 1 Among the band groups are 1 A check area , According to this law, we can raid5 Analyze . We can determine the band size by comparison ： For example, the difference between the check area in a band group and the non check area in this band group may be very obvious , You can find the strip size by checking and comparing with the tool .

The server data recovery engineer is responsible for the server raid Analyze and make sure that the strip size is 1024 Sectors . according to 1024 Sector segmentation , Make a record the size of a strip , As shown in the figure below ,7 Jump to the same record 283123.

When 7 When all disks are in the same position , The direction of the check area can be judged by comparison , Then judge the whole RAID5 The direction of . It has been judged before that 8 Plate number is the first plate , hold 8 The number plate is in the first position , determine RAID5 To the left , The order is 8,2,3,4,1,7,5.

The server data recovery engineer preliminarily determines LUN The starting sector of is 172032 A sector , Use the data recovery tool to jump to 172032 The sector checks the use of the hard disk , Under normal circumstances, this sector belongs to the 5 Disk number should be the check area , But in fact, it shows that the verification area is 8 Number plate . According to the raid Law of left strike ,5 The check area of disk number should be in 172032-1024=171008 A sector , That's the last strip . Jump to 171008 A sector , The data recovery engineer of North Asia server found that the verification area is 5 Number plate . So it can be determined LUN The starting sector of is 171008 A sector .

Use raid The recombination tool recombines according to the determined disk sequence raid Add in , As shown in the figure below . choice RAID5,Stripe size 512KB, It's a left step .

Click on Build Restructuring . Group up RAID after , Because the data from 1024*8=8192 Sectors start . If the tool does not have the function of jumping to this sector , So, what we just assembled RAID It has to be done again with a file Build Recombination operation .RAID The starting sector of (Start sectors) choice 8192, This file can be arbitrarily selected from the starting sector and size (Count sectors), The figure below shows the restructured raid5.

Server storage data recovery result verification ：

RAID5 After the disk array reconstruction is completed, it is verified by the server administrator , Verification passed , Server data recovery is complete .