CN100363903C

CN100363903C - hard disk array verification method

Info

Publication number: CN100363903C
Application number: CNB200410056480XA
Authority: CN
Inventors: 李勇
Original assignee: Via Technologies Inc
Current assignee: Via Technologies Inc
Priority date: 2004-08-11
Filing date: 2004-08-11
Publication date: 2008-01-23
Anticipated expiration: 2024-08-11
Also published as: CN1588320A

Abstract

The invention provides a method and a related device for hard disk array verification. When a verification instruction is received to verify a certain part in a hard disk array, the invention simultaneously verifies each hard disk in the hard disk array, and the actually verified part of each hard disk surpasses the part which is indicated to be verified in the verification instruction, and records the verification result for the surpassed part. When the next verification instruction is received to verify another part of the hard disk array, if the part is verified before, the verification result can be directly recorded to respond to the verification instruction, and the verification efficiency of the hard disk array is improved.

Description

The hard disk array verification method

Technical field

The invention provides a kind of method and relevant apparatus of verifying hard disk array, refer to especially a kind of according to each hard disk the checking ability but not the hard disk array data framework carries out the method and the relevant apparatus of hard disk array checking.

Background technology

Computer system can convenience, computing promptly, access, management lot of data, data and audio and video information, has become one of most important hardware foundation of advanced information society; And information manufacturer also endeavours to research and develop various technology, so that computer system can quicker, more safely be handled more data.

Hard disk is to use one of non-volatile storage the most widely in the modem computer systems; Along with the cost of hard disk reduces, capacity raises, modern computer system can several hard disks of integration management, these hard disks are formed a hard disk array carry out the access of data.For instance, at the cheap hard disk array (RAID of redundancy, redundant array of inexpensive disks) under the division formula data framework (similarly being the data framework that RAID 0, RAID 2 to 5 etc. are adopted), in the time will being stored in data in the hard disk array, these data can be divided (stripe) and be different parts, and each several part can be transferred to the different hard disks in the hard disk array simultaneously and store.When reading these data in will be by hard disk array, then can be simultaneously by the different piece that reads these data in each hard disk, recombinant goes out original data.In addition, also have the hard disk array data framework of expanding (SPAN/JBOD, JBOD are just a bunch of disks), under this kind data framework, the storage space of each hard disk can be one by head and the tail connection ground polyphone in the hard disk array; In other words, the data that deposit hard disk array in can preferentially be deposited in first hard disk, if first hard disk not have residue, available storage space, just understand being deposited in to second hard disk, by that analogy.In other words, just with the storage space of first hard disk first as hard disk array global storage space, the storage space of second hard disk is as the second portion in hard disk array global storage space, by that analogy.

As be familiar with known to the operator, guarantee that hard disk/hard disk array can correctly preserve data, must verify hard disk/hard disk array (verify).In general, when hard disk/hard disk array being formatd (format) or disk scanning (scan disk), will verify hard disk/hard disk array whether can correctly preserve data and in addition access with the each several part of checking in hard disk/hard disk array.In general, the checking of hard disk often will be taken the suitable time, owing to be provided with a plurality of hard disks in the hard disk array, verify all hard disks of complete hard disk array, may will expend more time.Yet the hard disk array checking is again one of steps necessary of guaranteeing data security, so will how to promote the checking usefulness of hard disk array, also just becomes one of emphasis of information manufacturer research and development.

In existing prior art, the checking of hard disk array all is that the data framework according to hard disk array carries out.For instance, in the redundant cheap hard disk array of the division formula of forming with N hard disk, data respectively have the part of 1/N information and are stored in respectively in the hard disk owing to can be divided into N part, so when this hard disk array is verified, if will verify that M unit in the hard disk array (similarly is with byte or sector, sector is unit) storage space the time, prior art can be verified the storage space of about M/N unit respectively in N hard disk in fact.Because, according to the data framework of redundant cheap hard disk array, gather M/N the unit that is verified respectively in N the hard disk, just be equivalent in hard disk array, verify the storage space of M unit.After having verified M unit's storage space of hard disk array, hard disk array can continue to verify the storage space of an other M unit in regular turn, by that analogy, all is verified up to all storage spaces of DISK to Image array.

Yet above-mentioned existing hard disk array verification technique also has shortcoming; For instance, this prior art must be controlled each hard disk continually and verifies.When control one hard disk is verified, it (similarly is the read verify instruction under the ATA specification that computer system needs hard disk is sent hard disk checking instruction, ATA is Advanced Technology Attachment), to verify which part, have the storage space of much capacity with the indication hard disk; And need to require the checking storage space to have a maximal value in the single hardware checking instruction.That is to say that if this maximal value is the storage space of K unit, then single hardware checking instruction can only be controlled a hard disk at most the storage space of its K unit is verified.Yet, in general, in above-mentioned existing hard disk array verification technique, when each control hard disk is verified, the maximum memory space that the storage space of its checking often can be verified much smaller than single hardware checking instruction.Brought forward is described; in the time will in hard disk array, verifying the storage space of M unit; prior art can be with the storage space of M/N unit of hard disk checking instruction control one a hard disk checking; but the storage space of this M/N unit usually can be less than the storage space of K unit, the maximum memory space that the instruction of single hardware checking just can be indicated.Because sending hard disk checking instruction each time, prior art require the storage space of checking all to verify the maximum verification space that instruction can be verified less than single hardware, so will verify a hard disk, prior art will be instructed to the hard disk checking that hard disk sends more repeatedly, consume more resource for computer system; And this makes that also the usefulness of existing hard disk array verification technique is lower.

On the other hand, under expanding hard disk array data framework, because first hard disk of hard disk array can provide the storage space of first in the hard disk array, second hard disk can provide the storage space of hard disk array second portion, so when the hard disk array to this type carries out the hard disk array checking, prior art also can be verified first hard disk in the hard disk array earlier, verifies second hard disk again after having verified first hard disk, by that analogy.Certainly, the shortcoming of this kind prior art also is that usefulness is low, because verify complete the time that hard disk array is required, is equivalent to verify the summation of indivedual hard disk required times.

Summary of the invention

Therefore, fundamental purpose of the present invention promptly is to propose a kind of method and the relevant apparatus that can promote hard disk array checking usefulness, can make the hard disk array of different pieces of information framework all finish the hard disk array checking, to overcome the lower shortcoming of prior art usefulness with higher usefulness.

The present invention can be implemented in the driver of hard disk array.When the present invention receives the checking instruction (for example being the instruction that the user of computer system sends via operating system) of hard disk array and when the storage space of M unit is verified in will be to hard disk array, no matter what hard disk array adopted is any data framework, the present invention all together with the time control the relatively large storage space of each hard disk checking (in preferred embodiment, all indicate the maximum storage space of its checking when each time hard disk being sent hard disk checking instruction), make the storage space of each hard disk actual verification can be equivalent to the storage space of Me unit in the hard disk array, and the storage space of this Me unit can be greater than the storage space of M unit.In other words, the M unit's storage space that not only will verify originally is verified, and the present invention has also additionally verified more storage space.For the storage space of extra checking, the present invention can get off its checking outcome record earlier; For the M unit's storage space that originally will verify, then can repay the checking result at once.When the present invention receives another hard disk array checking instruction and will verify other storage space of hard disk array the time, if the storage space that requirement will be verified in this instruction has been covered by the storage space of before additionally verifying, just the previous checking immediate yield as a result of record in advance needn't actually again can be sent hard disk checking instruction control hard disk and is verified.

In other words, the present invention can make full use of the chance of each time hard disk being sent hard disk checking instruction, finishes the checking of hard disk array efficiently.Concerning the cheap hard disk array of redundancy, the present invention can reduce the actual number of times that sends hard disk checking instruction of hard disk, reduces the running burden of computer system, the usefulness of promoting the hard disk array checking.Concerning the expanding hard disk array, because meeting of the present invention is verified each hard disk in the hard disk array simultaneously, just when first hard disk of hard disk array is verified, second (and in the hard disk array other hard disk) also all can verify simultaneously, so the present invention can significantly shorten the time of hard disk array checking.

The invention provides a kind of method that is used for verifying hard disk array, this hard disk array includes a plurality of hard disks, and each hard disk is used to provide a storage space to form the total storage space in this hard disk array; And this method includes: make the part of the storage space of each hard disk corresponding to total storage space in this hard disk array; The Receipt Validation instruction, this checking instruction system is used to refer to the instruction verification portion, and this instruction verification portion system is used for representing the part that need be verified in total storage space of this hard disk array in this checking instruction; After receiving this checking instruction, make each hard disk the verifying its storage space respectively in this hard disk array to certain portions, make part that each hard disk verifies an equivalent verification portion, and make these effects card parts greater than this instruction verification portion corresponding to this total storage space; After each hard disk is verified, to according with the part of this instruction verification portion in these effect card parts, repayment checking result instructs to respond this checking, and after each hard disk is verified, to surmounting the part of this instruction verification portion, the result of record checking in these effect card parts.

Description of drawings

Graphic simple declaration

The 1st figure is the function block schematic diagram of computer system of the present invention.

What the 2nd figure illustrated is the schematic flow sheet that computer system is carried out the hard disk array checking among the 1st figure.

3rd, the 4th figure signal is the situation that computer system is carried out the hard disk array checking among the 1st figure to the cheap hard disk array of redundancy.

What the 5th to 7 figure illustrated is the situation that computer system is carried out the hard disk array checking among the 1st figure to the expanding hard disk array.

Graphic symbol description

10 computer systems, 12 central processing units

14 chipsets, 16 hard disk array controllers

18 storeies, 20 drivers

22A interface routine 22B control program

22C repayment program 22D logging program

22E determining program 22F skips program

The 102-116 step

The part of M1, M2, Me, Mr, M1s, M2s storage space

HD (1)-HD (N) hard disk

Embodiment

Please refer to the 1st figure.The 1st figure is the synoptic diagram of an embodiment of computer system 10 of the present invention.Be provided with central processing unit 12, chipset 14, hard disk array controller 16, storer 18 in the computer system 10, and a plurality ofly be used for forming the hard disk HD (1) of hard disk array to HD (N).Central processing unit 12 is used for the running of host computer system 10, is responsible for operating system (OS, operation system) and various program implementation, and the processing of data, data or the like.Storer 18 as system storage can be a dynamic RAM, is used for storing required program and data during central processing unit 12 running.Chipset 14 is electrically connected between central processing unit 12 and the storer 18, is used for data transmission between diode-capacitor storage 18, central processing unit 12 and each peripheral device.16 of hard disk array controllers (similarly are IDE, SATA or SCSI bus with bus, IDE is Intelligent Drive Electronic, SATA is serial ATA, and SCSI is small computer system interface) passage (channel) that forms is connected to each hard disk HD (1) to HD (N), these hard disks are integrated into a hard disk array, and the storage space that each hard disk can provide also just can form total storage space of hard disk array via the integration of hard disk array controller 16.In addition, hard disk array controller 16 also can be integrated in the chipset.By the execution of corresponding driver 20, central processing unit 12 just can come total storage space of access hard disk array via hard disk array controller 16.

For instance, as the user of computer system 10 when requirement will be stored to data in the hard disk array by operating system, operating system can make central processing unit 12 carry out drivers 20 and handle this requirement.By the execution of driver 20, these data can suitably be handled according to the data framework of hard disk array, deposit in to suitable hard disk via the control of hard disk array controller 16 again.For example, if hard disk array is RAID 5 data frameworks under the redundant cheap hard disk array, execution via driver 20, these data will be divided into the N part, and calculating corresponding parity check sign indicating number, which hard disk the information and the parity check sign indicating number that determine this N part again will deposit in respectively to.Next, hard disk controller 16 will write instruction (similarly being the instruction that writes under the ATA/ATAPI specification) with suitable hard disk by each channel, control each hard disk HD (1) respectively to HD (N), deposit in dispersedly to different hard disks with information and parity check sign indicating number with this N part.In like manner, when the user requires to carry out the checking of hard disk array, operating system (or other verification management application program) can be sent different checking instructions to driver in regular turn, require driver 20 in regular turn the different piece of hard disk array to be verified, up to the checking of finishing the DISK to Image array.

In order to realize high-effect hard disk array verification technique of the present invention, be provided with interface routine 22A, control program 22B, repayment program 22C, logging program 22D, determining program 22E in the driver 20 of the present invention in addition and skip program 22F.Send checking instruction and requirement will verify certain part (similarly being some sector) in the hard disk array time in operating system, driver 20 can receive the hard disk array checking instruction that operating system is sent via the execution of interface routine 22A, makes hard disk controller 16 send hard disk checking instruction (similarly being the read verify instruction under the ATA/ATAPI specification) to each hard disk HD (1) to HD (N) practically by the execution of control program 22B again.In the present invention, the chance of sending hard disk checking instruction each time to each hard disk is used in the execution meeting of control program 22B efficiently, makes the actual part of verifying of each hard disk surmount the part of operating system checking that requires in the checking instruction.In other words, in hard disk array, not only the part of operating system call checking is verified, and the present invention has also additionally verified other part.For the part of operating system call checking, can utilize repayment program 22C to repay its checking result to operating system; For the checking result of the extra verification portion of the present invention, then can utilize logging program 22D record in addition, similarly be to be recorded in the storer 18.By the time operating system is sent the checking instruction of another hard disk array and requirement will verify another part of hard disk array the time, utilize determining program 22E, driver 20 just can judge in this checking instruction to require the part of checking whether to be included in the part of before additionally verifying; If just can need not send hard disk checking instruction to hard disk is actual, and repay program 22C again and directly the checking result of precedence record is repaid to operating system, as response to its checking instruction via the execution of skipping program 22F.So, just can effectively improve the usefulness of hard disk array checking.

For further specifying the process that the present invention carries out the hard disk array checking, please continue with reference to the 2nd figure (and in the lump with reference to the 1st figure).The 2nd figure is the schematic flow sheet that computer system 10 is carried out the hard disk array checking; Be provided with the following step in this flow process:

Step 102:

Driver 20 receives the checking instruction that operating system is sent.As previously mentioned, in the time will carrying out the hard disk array checking, operating system can be sent different checking instructions to driver 20 in regular turn, requires the different piece of driver 20 checking hard disk arrays in each checking instruction respectively, all finishes checking up to total storage space of DISK to Image array.Therefore, the instruction checking progress that all includes an appointment in each checking instruction of operating system, being used to refer to driver 20 should verify which part (for example being which sector) in the hard disk array, and the part of checking command request checking just can be described as the instruction verification portion.And in this step, driver 20 promptly begins to receive a checking instruction, and learns the instruction checking progress that requires in this checking instruction with this.

Step 104:

Hard disk array controller 16 sends hard disk checking instruction to each hard disk in the hard disk array simultaneously, and hard disk is actual verifies in indication.In the present invention, no matter the data framework of hard disk array why, no matter the instruction checking progress also in the checking instruction why, hard disk on each channel is verified, make that when the part of each hard disk actual verification is reflected in total storage space of hard disk array its part of occupying can be greater than the instruction verification portion in total storage space.As previously mentioned, in hard disk checking instruction, the storage space that can require hard disk to verify has a maximal value.And in preferred embodiment of the present invention, when each hard disk sends hard disk checking instruction, just can require hard disk to verify peaked storage space in reality.In other words, in preferred embodiment of the present invention, can make full use of the chance of sending hard disk checking instruction each time to hard disk, make each hard disk when carrying out single hardware checking instruction, all verify maximum storage spaces.

Step 106:

Driver 20 can calculate these storage spaces of having verified according to each hard disk storage space of actual verification in step 104 and be equivalent to which part in the total storage space of hard disk array, and calculate first-class effect card progress according to this, just in total storage space of hard disk array, there is which part in step 104, to be verified.Since total storage space of hard disk array is to be integrated and got by the storage space of each hard disk, Zong the storage space that all hard disks had been verified in step 104 also must be equivalently corresponding to some part in the storage space, be exactly the equivalent verification portion of the equivalent checking progress of these parts representative.

In the present invention, owing to step 104 can make each hard disk verify the storage space of more (or maximum) when each hard disk being sent hard disk checking instruction, so equivalent checking progress should be greater than the checking of the instruction in the step 102 progress.In other words, except operating system was verified instruction checking progress specified in the instruction, the present invention had also made full use of the chance of hard disk being sent hard disk checking instruction, had verified other part extraly in total storage space.

Step 108:

Because actual equivalence checking progress of carrying out has been higher than instruction checking progress indicated in the checking instruction among the present invention, the instruction verification portion in total storage space should be covered by in the equivalent verification portion.For the part that meets instruction checking progress in the equivalence checking progress, just can respond its checking of in step 102, sending instruction to operating system repayment checking result.

Step 110:

Surmount in the part of instruction checking progress for equivalence checking progress, can write down the result of its checking earlier.That is to say that the part to the present invention's extra checking in total storage space writes down the result of its checking.

Step 112:

After the completing steps 108,110, just can receive another checking instruction of operating system.Certainly, also indicated another instruction checking progress in this checking instruction.

Step 114:

Whether the instruction checking progress that is received in the determining step 112 has been included in the equivalence checking progress of before having verified.Because the present invention can additionally verify the other parts in total storage space in step 104, the progress of these extra checkings has contained the instruction checking progress in the step 112 probably.If the way it goes, then can proceed to step 116; If not, then recurrence proceeds to step 104.

Step 116:

If the checking of the instruction in the step 112 progress is in before having reached, the present invention does not just need reality that each hard disk is sent hard disk checking instruction again, as long as directly the checking result with precedence record is reported to operating system, just can satisfy the checking instruction that operating system is sent in step 112, and get back to the continued access of step 112 relaying and receive another checking instruction.

For further specifying the situation of flow implementation among the 2nd figure, below will illustrate that respectively the present invention applies to the situation of redundant cheap hard disk array and expanding hard disk array with different embodiment.At first please refer to the 3rd figure and the 4th figure (and in the lump with reference to the 1st figure and the 2nd figure); What the 3rd figure and the 4th figure illustrated is computer system 10 of the present invention is carried out the hard disk array checking in the cheap hard disk array of redundancy synoptic diagram.

In the embodiment of the 3rd figure, the 4th figure, suppose that computer system 10 is to integrate out a cheap hard disk array of redundancy (similarly being the hard disk array of RAID 5) with four hard disk HD (1) to HD (4), forms total storage space of hard disk array integral body with the storage space of each hard disk.In the 3rd figure, suppose that operating system sent a checking instruction, require the M1 in 20 pairs of total storage spaces of driver partly to verify (in other words, M1 partly is the instruction verification portion); And driver 20 will begin to carry out the flow process among the 2nd figure, is begun to carry out by step 102.Under the data framework of the cheap hard disk array of redundancy, total storage space of M1 part can be shared by the storage space of each hard disk, so concerning each hard disk HD (1) to the HD (4), if only will satisfy checking to the M1 part, the storage space of each its required checking of hard disk can be less than the M1 part, just the M1s part that is indicated among the 2nd figure.In the prior art, also only can be respectively respectively verify the M1s part to HF (4) with each hard disk HD (1) of hard disk checking instruction control.But in step 104 of the present invention, can make full use of each time each hard disk HD (1) is sent the chance of hard disk checking instruction to HD (4), allow the more storage space of each hard disk checking, just the Mes part that indicates among the 2nd figure.Gather each hard disk HD (1) to HD (4) respectively the checking the Mes part, just corresponding to the Me part in the total storage space of hard disk array, just equivalent verification portion, such as among the 2nd figure sign.In other words, the present invention is sending total storage space of being verified after the hard disk checking instruction respectively to HD (4) to each hard disk HD (1), not only contained the M1 part of instruction checking progress, is also additionally verifying the Mr part in total storage space more.

Along with the carrying out of flow process among the 2nd figure, to the M1 checking result partly of total storage space, but real-time repaying is to operating system (step 108 of the 2nd figure); To the Mr part that additionally is verified in total storage space, then the checking result of this part (whether similarly be the address that takes place of defectiveness, defective etc.) can be noted (step 108) earlier.

Please refer to the 4th figure.Continue the situation among the 3rd figure, in the 4th figure, hypothesis driven program 20 begins to receive next checking instruction (step 112); In the checking of this hard disk array instruction, operating system continues requirement will verify M2 part in total storage space, such as among the 4th figure sign.But, can find out by the 4th figure, this M2 part has been included in the Mr part of the extra checking of previous the present invention in fact, so the flow process of the present invention in the 2nd figure just can proceed directly to step 116, to meet M2 checking immediate yield as a result partly to operating system (step 116) in the Mr part, and not need to verify instruction to the actual hard disk that sends of each hard disk again.

If the indicated instruction checking progress of the checking of driver 20 follow-up receptions instruction all also is covered by the Me part in the equivalent checking progress, the present invention just can continue to skip the access control to each hard disk, just, needn't carry out step 104 in step 112,114,116 circulations; Verify that up to the instruction of new checking instruction progress has surmounted the Me part in the equivalent checking progress, just can proceed to step 104 once again.

By above discussion as can be known, in the time of in being used in redundant cheap hard disk array, the present invention can reduce the number of times that each hard disk is sent hard disk checking instruction as far as possible, make hard disk array checking of the present invention can be fast, carry out efficiently, also promote the overall efficiency of computer system.

Apply to the situation of expanding hard disk array as for the present invention, please refer to the 5th figure to the 7 figure (and in the lump with reference to the 1st figure and the 2nd figure); The 5th figure to the 7 figure are the present invention carries out the hard disk array checking in the expanding hard disk array synoptic diagram.Suppose in the embodiment of the 5th figure to the 7 figure, computer system 10 is the hard disk arrays that form expanding with identical hard disk HD (1) of two capacity and HD (2), so the storage space of hard disk HD (1) just becomes the first half of the total storage space of hard disk array, the storage space of hard disk HD (2) becomes the latter half of of storage space.At first, shown in the 5th figure, hypothesis driven program 20 begins to receive hard disk array checking instruction, and requirement will be verified the M1 part in total storage space of hard disk array.So driver 20 begins to carry out the checking flow process among the 2nd figure, and in step 104, simultaneously hard disk HD (1) and HD (2) are sent hard disk checking instruction, hard disks all in the hard disk array is all verified.According to the data framework of expanding hard disk array, because the M1 in total storage space is partly provided by hard disk HD (1), so, only need hard disk HD (1) is sent hard disk checking instruction if will satisfy the checking instruction of M1 part.And prior art is really also only in hard disk HD (1) checking M1 part.But in step 104 of the present invention, because hard disk array controller 16 can send hard disk checking instruction to the hard disk on each channel simultaneously, so meeting of the present invention makes hard disk HD (1) and HD (2) all verify simultaneously.Shown in the 5th figure, in the present invention, not only hard disk HD (1) can verify the M1 part, and hard disk HD (2) also can verify the M1s part, a later half part in just total storage space.This M1s partly is exactly the part of the present invention's extra checking in step 104, and its checking result can be prior to going on record in the step 110.

Continue the situation of the 5th figure, in the 6th figure, driver 20 proceeds to step 112, begins to receive next checking instruction; This checking instruction can require the M2 in 20 pairs of total storage spaces of driver partly to verify.And driver 20 will send hard disk checking instruction to hard disk HD (1), HD (2) once again in step 104.So, not only the storage space of meeting checking M2 part is to satisfy the requirement (step 108) that checking is instructed among the hard disk HD (1), and hard disk HD (2) also can verify the M2s part extraly, latter half of another part in just total storage space.Checking result to the M2 part also can go on record in step 110.In comparison, prior art still only can only be controlled hard disk HD (1) checking M2 part according to the data framework of expanding hard disk array.

Along with driver 20 by operating system Receipt Validation instruction constantly, hard disk HD (1) and HD (2) can verify.By the time after the operating system checking instruction of sending the first half of total storage space (just hard disk HD (1) provide storage space) checking has been provided, because the present invention verifies all simultaneously to two hard disk HD (1) and HD (2), so the latter half of storage space that is provided by hard disk HD (2) of total storage space also is verified, just shown in the 7th figure.In other words, in the 7th figure, being verified the storage space that finishes among the hard disk HD (2) is exactly the storage space of the extra checking of the present invention.When sending the checking command request and will verify total storage space latter half of etc. operating system, because the latter half of of total storage space has been included in the extra storage space of verifying of the present invention, so just can circulating apace to follow-up checking instruction, driver 20 carries out step 112 to 116, needn't control that each hard disk is actual to be verified, as long as suitably the checking result with precedence record is reported to operating system, just can satisfy all follow-up checking instructions of operating system.

In other words, if form the hard disk array of an expanding with N hard disk HD (1) that waits capacity to HD (N), just only when the preceding 1/N to total storage space partly verifies, the present invention just can send hard disk to each hard disk and verify and instruct.When the instruction progress in the checking instruction begins requirement follow-up (the N-1)/N of total storage space is partly verified, because all can belonging to the present invention, these parts additionally verify the part of finishing, so driver of the present invention just needn't actual send hard disk checking instruction to each hard disk, instruct and under unactual situation of carrying out the hard disk checking, satisfy all follow-up checkings apace.

In summary,,, additionally verify some storage spaces, needn't be subject to the data framework of hard disk array so the present invention can make full use of the chance of each time hard disk being sent hard disk checking instruction because the checking of hard disk array does not relate to the access of data.Compared to prior art,, guarantee the data security in the hard disk array no matter the present invention, all can effectively promote the usefulness of hard disk array checking to cheap hard disk array of redundancy or expanding hard disk array.As be familiar with the operator and can understand, except in the driver of hard disk array controller correspondence, realizing technology of the present invention, also can suitably but some program in the driver of the present invention be implemented in the hard disk array controller with hardware, firmware or program circuit, to reach purpose of the present invention.

The above only is preferred embodiment of the present invention, and all equalizations of being done according to the present patent application claim change and modify, and all should belong to the covering scope of patent of the present invention.

Claims

1. A method for verifying a hard disk array, the hard disk array includes a plurality of hard disks, each hard disk is used to provide a storage space to form the total storage space in the hard disk array; and the method includes:

Make the storage space of each hard disk correspond to a part of the total storage space in the hard disk array;

receiving a verification instruction, the verification instruction is used to indicate the instruction verification part, and the instruction verification part is used to represent the part of the total storage space of the hard disk array that needs to be verified in the verification instruction;

After receiving the verification instruction, each hard disk in the hard disk array is made to verify a given part of its storage space, so that the part of each hard disk to be verified corresponds to an equivalent verification part of the total storage space, and the The equivalent verification part is larger than the instruction verification part;

After each hard disk is verified, the part of the equivalent verification part that complies with the instruction verification part is reported to respond to the verification command, and after each hard disk is verified, the equivalent verification part exceeds The part of the directive that validates the section that records the results of the validation.

2. The method according to claim 1, further comprising:

receiving a second verification instruction, the verification instruction is used to indicate a second instruction verification part, and the second instruction verification part is used to indicate a part of the total storage space that needs to be verified in the second verification instruction;

determining whether the second instruction verification part is covered by the equivalent verification part; and

If the second instruction verification part is included in the equivalent verification part, report a second verification result according to the recorded verification result in response to the second verification instruction.

3. The method of claim 2, further comprising:

If the second instruction verification part is included in the equivalent verification part, then do not make each hard disk to perform verification, that is, report the second verification result.