CN103150128A - Implementation method of solid state drive (SSD) and disk-based reliable mixed storage system - Google Patents
Implementation method of solid state drive (SSD) and disk-based reliable mixed storage system Download PDFInfo
- Publication number
- CN103150128A CN103150128A CN2013100969311A CN201310096931A CN103150128A CN 103150128 A CN103150128 A CN 103150128A CN 2013100969311 A CN2013100969311 A CN 2013100969311A CN 201310096931 A CN201310096931 A CN 201310096931A CN 103150128 A CN103150128 A CN 103150128A
- Authority
- CN
- China
- Prior art keywords
- disk
- data
- ssd
- redirect
- write
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Techniques For Improving Reliability Of Storages (AREA)
Abstract
The invention discloses an implementation method of a solid state drive (SSD) and disk-based reliable mixed storage system. According to the method, a mixed storage system is constructed by using an SSD and a disk; the SSD responds to a read-write request of a user, the disk serves as a backup, and when the user sends a read request, the SSD completely responds to the read request; when the user sends a write request and writes data into the SSD, the data are backed up by the disk; if the SSD fails, the data of the SSD can be restored according to the data on the disk; and if the disk fails, the data in the disk can be reconstructed according to the data on the SSD. In order to overcome the defects of lower SSD write function, short life and higher cost in the prior art, according to the implementation method disclosed by the embodiment of the invention, the mixed storage system is implemented by using a cheap disk as the backup, the high performance of the SSD is ensured, the reliability of the entire mixed storage system is improved, and the cost of the mixed storage system is controlled within an acceptable range. The method has the advantages of high read-write performance, high system reliability, difficulty in occurrence of relevant failure, long service life and low cost.
Description
Technical field
The present invention relates to the computer memory system field, be specifically related to a kind of with disk as the backup of SSD to realize a kind of reliable and cheap reliable mixing storage system implementation method based on SSD and disk.
Background technology
SSD (Solid State Drive, solid-state disk) based on flash memory is a kind of novel mass storage devices.Than disk, it has lower read-write and postpones; Because the concurrency that it is intrinsic, SSD can provide very high bandwidth; In addition, it also have low in energy consumption, quality light, to advantages such as the working environment requirement are lower.In a word, SSD is expected to become follow-on mainstream storage device.But the price of flash memory is also very high at present, and the SSD of same capability and disk price differ nearly ten times, mainly is used in high-end system so SSD is current, perhaps is used as the buffer memory of disk.Another defective of SSD is that the life-span is limited.Each storage unit of flash memory can only be born the write operation of limited number of times, and after write operation outnumbers this restriction, it is unreliable that this storage unit will become, and finally may cause loss of data.
Typical storage system comprises a large amount of memory devices, and it is a kind of frequent event that memory device lost efficacy, and how utilizing insecure memory device to build reliable storage system is a major challenge.Traditional method that is used for improving reliability comprises RAID (Redundant Array of Independent Disks, Redundant Array of Independent Disks (RAID)) technology, redundancy technique etc.
RAID technology commonly used comprises RAID1, RAID5, RAID6.RAID1 is with the user data piecemeal, and every blocks of data is write portion on two different memory devices, when a memory device lost efficacy, can pass through another memory device restoring user data.RAID5 produces check information when writing user data, and user data and check information are write on different memory devices.When user data loss, can pass through the check information restoring user data.RAID5 can tolerate that a memory device lost efficacy.RAID6 preserves more check information, can tolerate that two memory devices lost efficacy simultaneously.
Redundancy technique is preserved a plurality of copies of user data on distinct device, keep consistance between these copies with specific agreement, as long as there is a copy effective, user data just can not lost.Redundancy technique can be regarded loose RAID1 array as.The Google file system is exactly to guarantee reliability by redundancy technique.
But above-mentioned RAID technology and redundancy technique all can not directly be used in storage system based on SSD.At first, the number of times that writes of each storage unit of SSD is limited.The RAID technology all can regenerate check information when each write operation.In principle, the data volume of the check information of generation and renewal is the same large.So responding the data that a write request finally writes is twices of actual request, this reduced by half with regard to the life-span that makes SSD.Secondly, the write delay of SSD is larger, and the write operation of part RAID technology (as RAID5) can cause twice read operation and twice write operation, and this has just increased the delay of write operation more.In a word, both reduce the performance of SSD based on the RAID technology of SSD, also shortened the life-span of SSD.Redundancy technique is preserved a plurality of copies, needs more storage space, and the price of SSD is relatively high at present, is only suitable for as the high-end devices that guarantees memory property, does backup with it obviously too expensive.
In sum, the storage system of prior art all can't realize both having given full play to the performance advantage of SSD, can make up again the life-span defective of SSD simultaneously.If design a kind of storage system based on SSD of cheapness, can give full play to the performance advantage of SSD, can make up again the life-span defective of SSD, will greatly promote SSD in the application in Mass storage field.
Summary of the invention
The technical problem to be solved in the present invention is to provide that a kind of readwrite performance is high and system reliability good, be difficult for occuring relevant fault, long service life, the reliable mixing storage system implementation method based on SSD and disk with low cost.
In order to solve the problems of the technologies described above, the technical solution used in the present invention is:
A kind of reliable mixing storage system implementation method based on SSD and disk, implementation step is as follows:
1) SSD and disk is ready, wherein the storage space of disk is greater than the storage space of SSD, and redirect is carried out next step;
2) format SSD and disk address space are divided into data block and free block with the linear address space of disk take piece as unit, all data blocks on described disk form a linear address space, and are corresponding one by one with the linear address space of SSD; Redirect is carried out next step;
3) head position of initialization disk, redirect is carried out next step;
4) receive the operating system transmission to the read-write requests of SSD, redirect is carried out next step;
5) judge the request type of described read-write requests, if request type is read request, redirect execution in step 6); If request type is write request, redirect execution in step 7);
6) according to the SSD address of read request from the SSD reading out data, return data is waited for and is come then redirect execution in step 4 in next read-write requests to operating system);
7) according to the SSD address of write request, the data of write request are write SSD, redirect is carried out next step;
8) the SSD address translation with described write request becomes disk address, and redirect is carried out next step;
9) distance between the magnetic head current location of more described disk address and disk, if the distance between the magnetic head current location of described disk address and disk is less than assign thresholds, judge that the magnetic head current location of described disk address and disk closes on, redirect execution in step 10); Otherwise redirect execution in step 11);
10) according to the disk address of write request, the data of write request are sequentially write disk, wait for and come then redirect execution in step 4 in next read-write requests);
11) select the free block nearest with the magnetic head current location of disk in disk, redirect is carried out next step;
12) data that write write request to described free block record disk address corresponding to ephemeral data as ephemeral data, and redirect is carried out next step;
13) wait for the fixed time, redirect is carried out next step;
14) whether judgement has write request to arrive within the described fixed time, if having redirect execution in step 4); Otherwise redirect is carried out next step;
15) select the magnetic head current location recently and comprise the free block of ephemeral data, redirect is carried out next step;
16) read the ephemeral data of storing in described free block, redirect is carried out next step;
17) wait for the fixed time, redirect is carried out next step;
18) whether judgement has write request to arrive within the described fixed time, if having redirect execution in step 4); Otherwise, redirect next step;
19) ephemeral data of storing in described free block is write corresponding disk address, redirect is carried out next step;
20) judge in the free block of disk whether still have the free block that comprises ephemeral data, if all ephemeral datas have all been transferred to data block, wait in next read-write requests and come then redirect execution in step 4); Otherwise redirect execution in step 13) continue ephemeral data is transferred to data block.
As further improvement in the technical proposal of the present invention:
Described step 1) also comprise the non-volatile main memory of setting up for temporary write request and data thereof, described step 5) request type of judging described read-write requests as write request after, write request and data thereof are temporarily stored in described non-volatile main memory, if the data of write request finally are written into corresponding disk address, it is deleted from described non-volatile main memory, each starting up checks in described non-volatile main memory whether have residual write request and data thereof, if any redirect execution in step 8).
Described step 2) concrete steps that in, the linear address space of disk are divided into data block and free block take piece as unit as: the linear address space of disk is divided into the data segment of a plurality of continuous distribution, each data segment comprises a plurality of data blocks and a free block.
Described step 8) specifically refer to according to formula (1), the SSD address translation of read-write requests be become disk address;
Addr
Disk＝(Addr
Backup/(B×S))×(S+1)×B+(Addr
Backup％(B×S))(1)
In formula (1), Addr
DiskBe disk address; Addr
BackupRepresent the disk address of SSD address corresponding data piece in the disk linear address space; Addr
Backup/ (B * S) represents Addr
DiskThe data segment at place; Addr
Backup(B * S) represents Addr to %
DiskData segment in skew, % represents modulo operation, the size of (S+1) * B representative data section; S represents the data block quantity in each data segment, and B represents the byte number of each data block.
Described step 11) specifically refer in select the free block nearest with the magnetic head current location of disk according to formula (2) in disk;
B
f＝(H
Disk/((S+1)×B))×(S+1) (2)
In formula (2), B
fBe the free block address nearest with the magnetic head current location of disk, H
Disk/ (data segment at the magnetic head current location place of (S+1) * B) expression disk, the size of (S+1) * B representative data section; S represents the data block quantity in each data segment, and B represents the byte number of each data block.
The present invention has following advantage:
1. the mixing storage system of the present invention's design is isomery, is not easy to occur relevant fault.Prior art is easy to occur relevant fault based on the RAID array of SSD, when namely a SSD breaks down, second SSD follows closely and breaks down, this is mainly because in homogeneous arrays, the out of order trigger condition of each equipment is the same, residing working environment too, when working environment triggers a SSD inefficacy, probably cause another piece disk failure.By contrast, the array of the present invention's design is isomery, and the probability that disk and SSD lost efficacy simultaneously is very little, is not easy to occur relevant fault.
2. the commingled system readwrite performance of the present invention's proposition is higher.In mixing storage system implementation method based on SSD and disk of the present invention, read request fully by the SSD response, can take full advantage of SSD and read the high advantage of performance, so read better performances; For write request, when data are write SSD, also data must be write on disk, although disk performance is lower, can not reduce the performance of whole mixing storage system, reason such as following some: at first, disk of the present invention only responds write request and does not respond read request, SSD not only responds write request but also respond read request, and the duty factor SSD of disk is low, and the performance bottleneck of whole mixing storage system is SSD and not at disk; Secondly, the present invention strengthens the succession of disk operating from many aspects, improves disk performance.For example, step 11) Backup Data is write in the nearest free block of magnetic head, strengthen the succession of write operation; Step 16) read the data of a plurality of write requests by an operation, reduce the number of disk read-write operation.By means of these measures, disk of the present invention can back up the data of SSD in time.By contrast, prior art based on the RAID of SSD read that performance and the present invention propose to mix storage system suitable, but write performance is more relatively poor than write performance of the present invention, because when writing user data to the RAID array, also must write check information.
3. in the mixing storage system implementation method based on SSD and disk of the present invention, read request is responded by SSD fully, each reads or writes the quantity that the memory access quantity of SSD is actual request, and disk only responds write request and does not respond read request, SSD not only responds write request but also respond read request, the load of SSD and disk respectively has equilibrium, so the mixing storage system life-span that the present invention proposes is longer, and the life-span of mixing storage system is identical with SSD.By contrast, shorter based on the RAID array life-span of SSD, because when writing user data, also must write check information.
4, the present invention also further sets up that to be used for temporary request type be the read-write requests of write request and the non-volatile main memory of writing data thereof, the request type of judging read-write requests as write request after, write request and data thereof are temporarily stored in non-volatile main memory, if the data of writing of read-write requests finally are written into disk address it are deleted from non-volatile main memory, each starting up checks whether there is residual read-write requests in non-volatile main memory and writes data, writes synchronous processing if any automatically carrying out disk; Can overcome the unexpected asynchronous problem of SSD and disk write operation, the outage even meet accident, also can automatically complete automatically synchronous with SSD and disk write operation next time when starting shooting, therefore can improve the reliability based on the mixing storage system of SSD and disk.
Description of drawings
Fig. 1 is the method flow schematic diagram of the embodiment of the present invention.
Embodiment
As shown in Figure 1, the present embodiment is as follows based on the implementation step of the reliable mixing storage system implementation method of SSD and disk:
1) SSD and disk is ready, wherein the storage space of disk is greater than the storage space of SSD, and redirect is carried out next step.
2) format SSD and disk address space are divided into data block and free block with the linear address space of disk take piece as unit, all data blocks on disk form a linear address space, and are corresponding one by one with the linear address space of SSD; Redirect is carried out next step.
In the present embodiment, step 2) concrete steps that in, the linear address space of disk are divided into data block and free block take piece as unit as: the linear address space of disk is divided into a large amount of data segments, each data segment comprises S data block and a free block, and each data block or idle block size are the B byte.All data blocks on disk form a linear address space, and are corresponding one by one with the linear address space of SSD.If there is a logical address Addr in the linear address space of SSD
SSD, have corresponding logical address Addr on the linear address space that data block forms
BackupData block is used to back up the data of SSD, as the address Addr to SSD
SSDWhen sending a write request, also should be to Addr
BackupSend same write request.Because disk is read poor-performing, data write Addr
SSDShi Keneng can't synchronously write Addr with data
BackupAt this moment, data temporarily can be write in the free block of disk, these data are called as ephemeral data.When disk is idle, the ephemeral data that is temporarily stored in free block is transferred to Addr
BackupOn.
3) the head position H of initialization disk
Disk, redirect is carried out next step.
The technological difficulties that the present embodiment will solve are exactly the succession that strengthens disk write operation, and the present embodiment needs to follow the tracks of magnetic head in the position of disk linear address space, variable H in order to dispatch the write operation in disk
DiskBe used for the position of write head, it is initialized to 0, shows that magnetic head is in the reference position of disk.
4) receiving operating system sends read-write requests to SSD (the SSD address is Addr
SSD), redirect is carried out next step.
5) judge the request type of read-write requests, if request type is read request, redirect execution in step 6); If request type is write request, redirect execution in step 7).
6) according to the SSD address Addr of read request
SSDFrom the SSD reading out data, return data is waited for and is come then redirect execution in step 4 in next read-write requests to operating system).
7) according to the SSD address Addr of write request
SSDThe data of write request are write SSD, and redirect is carried out next step.
8) with the SSD address Addr of write request
SSDConvert disk address Addr to
Disk, redirect is carried out next step.
In the present embodiment, step 8) specifically refer to according to formula (1), the SSD address translation of write request be become disk address;
Addr
Disk＝(Addr
Backup/(B×S))×(S+1)×B+(Addr
Backup％(B×S))(1)
In formula (1), Addr
DiskBe disk address; Addr
BackupRepresent the disk address of SSD address corresponding data piece in the disk linear address space; Addr
Backup/ (B * S) represents Addr
DiskThe data segment at place; Addr
Backup(B * S) represents Addr to %
DiskData segment in skew, % represents modulo operation, the size of (S+1) * B representative data section; S represents the data block quantity in each data segment, and B represents the byte number of each data block.Step 2 in the present embodiment) during the partitioning discs address space, each SSD address Addr
SSDAddress Addr on the linear space that forms of corresponding data piece all
Backup, write Addr
SSDData should backup to Addr
BackupOn.But, Addr
BackupIt not the address of disk linear address space.During to the disk data writing, need to be with Addr
BackupConvert disk address Addr to
DiskShown in (1).
9) compare disk address Addr
DiskHead position and the magnetic head current location H of disk
DiskBetween distance, if disk address Addr
DiskMagnetic head current location H with disk
DiskBetween distance less than assign thresholds, judge disk address Addr
DiskMagnetic head current location H with disk
DiskClose on redirect execution in step 10); Otherwise redirect execution in step 11).By above-mentioned determining step, can reduce the operating lag of disk when busy, improve the performance gap of disk data writing and SSD, thereby improve the reliability of mixing storage system.
In the present embodiment, if Addr
DiskWith H
DiskClose on, explanation need not be dispatched magnetic head and just Backup Data can be write on disk, redirect execution in step 10), Backup Data is sequentially write disk; If Addr
DiskDistance H
DiskFar away, can not sequentially data be write Addr
Disk, redirect execution in step 11) and data are write the free block nearest apart from magnetic head, still guarantee the sequential write of disk.So-called judgement disk address Addr
DiskHead position and the magnetic head current location H of disk
DiskClose on, refer to Addr
DiskWith H
DiskDistance less than assign thresholds T
Threshold, namely | Addr
Disk-H
Disk|≤T
Threshold
10) according to the disk address Addr of write request
DiskThe data of write request are sequentially write disk, wait for and come then redirect execution in step 4 in next read-write requests).
11) the magnetic head current location H of selection and disk in disk
DiskNearest free block B
f, redirect is carried out next step.This step is by calculating distance H
DiskNearest free block B
f, Backup Data is write B temporarily
fIn, due to B
fVery near apart from magnetic head, avoided the magnetic head scheduling, still guarantee the sequential write of disk, can reduce the operating lag of disk when busy, improve the performance gap of disk data writing and SSD, thereby improve the reliability of mixing storage system.
In the present embodiment, step 11) specifically refer in select the free block nearest with the magnetic head current location of disk according to formula (2) in disk;
B
f＝(H
Disk/((S+1)×B))×(S+1) (2)
In formula (2), B
fBe the free block address nearest with the magnetic head current location of disk, H
Disk/ (data segment at the magnetic head current location place of (S+1) * B) expression disk, the size of (S+1) * B representative data section; S represents the data block quantity in each data segment, and B represents the byte number of each data block.
12) to free block B
fWhat write read-write requests writes data as ephemeral data, records disk address corresponding to ephemeral data, and redirect is carried out next step.In order to distinguish over the Backup Data of writing data blocks, write free block B
fBackup Data be called as ephemeral data.Free block B
fMiddlely may comprise the ephemeral data that some had before write, the ephemeral data that only need current preparation is write this moment is attached to thereafter.
13) wait for fixed time T, redirect is carried out next step.
The present embodiment is by waiting for fixed time T, namely by in when disk is idle, data in free block being transferred to data block.If at the appointed time T does not have write request to arrive disk in the period, illustrate that disk is idle.Pass through above-mentioned steps, can carry out data transfer operation when disk is idle, can improve the access performance of disk, can reduce the operating lag of disk when busy, improve the performance gap of disk data writing and SSD, thereby improve the reliability of mixing storage system.
14) judge that at the appointed time in T, whether request type being arranged is the read-write requests of write request, if having redirect execution in step 4) the follow-up write request of response; Otherwise redirect is carried out next step and is attempted the ephemeral data in free block is transferred in data block.
15) select magnetic head current location H
DiskRecently and comprise the free block B of ephemeral data
d, redirect is carried out next step.
16) read free block B
dThe ephemeral data of middle storage, redirect is carried out next step.
In the present embodiment, step 12) data that can not the order writing data blocks are temporarily write in free block, free block B
dMay comprise data corresponding to a plurality of write requests.At this moment, with B
dIn all ephemeral datas read, this way has not only reduced the number of read operation, has also strengthened the succession of disk operating, helps to improve disk performance.
17) wait for fixed time T, redirect is carried out next step.
18) judge at the appointed time whether have write request to arrive in T, if having redirect execution in step 4); Otherwise redirect is carried out next step.
If at the appointed time do not have write request to arrive in T, illustrating that disk is idle, with step 16) ephemeral data read writes when making the data writing data blocks in data block, scheduling magnetic head that can be the least possible, thus improve the readwrite performance of disk.
19) ephemeral data of storing in free block is write corresponding disk address Addr
Disk, redirect is carried out next step.
20) judge in the free block of disk whether still have the free block that comprises ephemeral data, if all ephemeral datas have all been transferred to data block, wait in next read-write requests and come then redirect execution in step 4); Otherwise redirect execution in step 13) continue ephemeral data is transferred to data block.
The present embodiment is by above-mentioned steps 1)～step 20), can realize the mixing storage system based on SSD and disk, and SSD is responsible for the read and write response, disk is responsible for writing backup, can take full advantage of the readwrite performance of SSD, is keeping SSD to read on the basis of performance, back up reading and writing data by disk, in case data are made mistakes, can utilize Backup Data in disk to carry out data and recover, readwrite performance is high and system reliability good, with low cost; The present embodiment is the mixing storage system of isomery based on the mixing storage system of SSD and disk, is difficult for occuring relevant fault; The present embodiment is based on the mixing storage system of SSD and disk, and SSD is responsible for the read and write response, and disk is responsible for writing backup, can realize that load shares, and can not increase the read-write burden of SSD, has advantages of long service life, with low cost.The present embodiment builds the storage system of mixing with SSD and disk, and wherein SSD responds user's read-write requests, and disk improves reliability as backup.When the user sends read request, responded by SSD fully; When the user sends write request, when data are write SSD, back up on disk.If SSD lost efficacy, can recover according to the data on disk the data of SSD; If disk failure can be according to the data in the data reconstruction disk on SSD.Than SSD, the disk read-write poor-performing.So, how the key problem in technology problem that the present embodiment solves is in time the data of SSD are backuped on disk, due to disk sequential write better performances, the present invention is devoted to strengthen the succession of write operation in disk, improve the write performance of disk, thereby on the data that guarantee SSD can back up to disk in time.The characteristics that the present embodiment is relatively poor for the SSD write performance, the life-span is limited, price is more expensive, do the backup of SSD with the disk of cheapness, realize a kind of storage system of mixing, both guaranteed the high-performance of SSD, improve again the reliability of whole mixing storage system, to mix simultaneously the cost Control of storage system within acceptable scope, and have advantages of that readwrite performance is high and system reliability good, be difficult for occuring relevant fault, long service life, with low cost.
In the present embodiment, step 1) also comprise the non-volatile main memory that foundation is used for keeping in write request and writes data, step 5) request type of judging read-write requests as write request after, write request and data thereof are temporarily stored in non-volatile main memory, if the data of write request finally are written into disk, it is deleted from non-volatile main memory, each starting up checks whether have residual write request and data thereof in non-volatile main memory, if any redirect execution in step 8).Therefore, the present embodiment is by the temporary write request of non-volatile main memory and data thereof, can overcome the unexpected asynchronous problem of SSD and disk write operation, the outage even meet accident, also can automatically complete automatically synchronous with SSD and disk write operation next time when starting shooting, therefore can improve the reliability based on the mixing storage system of SSD and disk.
The above is only the preferred embodiment of the present invention, and protection scope of the present invention also not only is confined to above-described embodiment, and all technical schemes that belongs under thinking of the present invention all belong to protection scope of the present invention.Should be pointed out that for those skilled in the art, in the some improvements and modifications that do not break away under principle of the invention prerequisite, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (5)
1. reliable mixing storage system implementation method based on SSD and disk is characterized in that implementation step is as follows:
1) SSD and disk is ready, wherein the storage space of disk is greater than the storage space of SSD, and redirect is carried out next step;
2) format SSD and disk address space are divided into data block and free block with the linear address space of disk take piece as unit, all data blocks on described disk form a linear address space, and are corresponding one by one with the linear address space of SSD; Redirect is carried out next step;
3) head position of initialization disk, redirect is carried out next step;
4) receive the operating system transmission to the read-write requests of SSD, redirect is carried out next step;
5) judge the request type of described read-write requests, if request type is read request, redirect execution in step 6); If request type is write request, redirect execution in step 7);
6) according to the SSD address of read request from the SSD reading out data, return data is waited for and is come then redirect execution in step 4 in next read-write requests to operating system);
7) according to the SSD address of write request, the data of write request are write SSD, redirect is carried out next step;
8) the SSD address translation with described write request becomes disk address, and redirect is carried out next step;
9) distance between the magnetic head current location of more described disk address and disk, if the distance between the magnetic head current location of described disk address and disk is less than assign thresholds, judge that the magnetic head current location of described disk address and disk closes on, redirect execution in step 10); Otherwise redirect execution in step 11);
10) according to the disk address of write request, the data of write request are sequentially write disk, wait for and come then redirect execution in step 4 in next read-write requests);
11) select the free block nearest with the magnetic head current location of disk in disk, redirect is carried out next step;
12) data that write write request to described free block record disk address corresponding to ephemeral data as ephemeral data, and redirect is carried out next step;
13) wait for the fixed time, redirect is carried out next step;
14) whether judgement has write request to arrive within the described fixed time, if having redirect execution in step 4); Otherwise redirect is carried out next step;
15) select the magnetic head current location recently and comprise the free block of ephemeral data, redirect is carried out next step;
16) read the ephemeral data of storing in described free block, redirect is carried out next step;
17) wait for the fixed time, redirect is carried out next step;
18) whether judgement has write request to arrive within the described fixed time, if having redirect execution in step 4); Otherwise, redirect next step;
19) ephemeral data of storing in described free block is write corresponding disk address, redirect is carried out next step;
20) judge in the free block of disk whether still have the free block that comprises ephemeral data, if all ephemeral datas have all been transferred to data block, wait in next read-write requests and come then redirect execution in step 4); Otherwise redirect execution in step 13) continue ephemeral data is transferred to data block.
2. the reliable mixing storage system implementation method based on SSD and disk according to claim 1, it is characterized in that: described step 1) also comprise and set up the non-volatile main memory that is used for temporary write request and data thereof, described step 5) request type of judging described read-write requests as write request after, write request and data thereof are temporarily stored in described non-volatile main memory, if the data of write request finally are written into corresponding disk address, it is deleted from described non-volatile main memory, each starting up checks in described non-volatile main memory whether have residual write request and data thereof, if any redirect execution in step 8).
3. the reliable mixing storage system implementation method based on SSD and disk according to claim 2, it is characterized in that: the concrete steps that described step 2), the linear address space of disk are divided into data block and free block take piece as unit as: the linear address space of disk is divided into the data segment of a plurality of continuous distribution, each data segment comprises a plurality of data blocks and a free block.
4. the reliable mixing storage system implementation method based on SSD and disk according to claim 3 is characterized in that: described step 8) specifically refer to according to formula (1), the SSD address translation of read-write requests be become disk address;
Addr
Disk＝(Addr
Backup/(B×S))×(S+1)×B+(Addr
Backup％(B×S)) (1)
In formula (1), Addr
DiskBe disk address; Addr
BackupRepresent the disk address of SSD address corresponding data piece in the disk linear address space; Addr
Backup/ (B * S) represents Addr
DiskThe data segment at place; Addr
Backup(B * S) represents Addr to %
DiskData segment in skew, % represents modulo operation, the size of (S+1) * B representative data section; S represents the data block quantity in each data segment, and B represents the byte number of each data block.
5. the reliable mixing storage system implementation method based on SSD and disk according to claim 4, is characterized in that: specifically refer to described step 11) select the free block nearest with the magnetic head current location of disk according to formula (2) in disk;
B
f＝(H
Disk/((S+1)×B))×(S+1) (2)
In formula (2), B
fBe the free block address nearest with the magnetic head current location of disk, H
Disk/ (data segment at the magnetic head current location place of (S+1) * B) expression disk, the size of (S+1) * B representative data section; S represents the data block quantity in each data segment, and B represents the byte number of each data block.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100969311A CN103150128A (en) | 2013-03-25 | 2013-03-25 | Implementation method of solid state drive (SSD) and disk-based reliable mixed storage system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100969311A CN103150128A (en) | 2013-03-25 | 2013-03-25 | Implementation method of solid state drive (SSD) and disk-based reliable mixed storage system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103150128A true CN103150128A (en) | 2013-06-12 |
Family
ID=48548236
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013100969311A Pending CN103150128A (en) | 2013-03-25 | 2013-03-25 | Implementation method of solid state drive (SSD) and disk-based reliable mixed storage system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103150128A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103713973A (en) * | 2014-01-08 | 2014-04-09 | 浪潮(北京)电子信息产业有限公司 | Mixed storage backup method and system based on HDD and SSD |
| CN104111894A (en) * | 2014-07-17 | 2014-10-22 | 记忆科技(深圳)有限公司 | Method for improving multi-partition sequential read-write performance and system thereof |
| CN105630415A (en) * | 2015-12-24 | 2016-06-01 | 创新科软件技术（深圳）有限公司 | RAID5 system and method for continuing to write data after RAID5 system failure |
| CN106469119A (en) * | 2015-08-10 | 2017-03-01 | 北京忆恒创源科技有限公司 | A kind of data write buffer method based on NVDIMM and its device |
| CN106648458A (en) * | 2016-09-30 | 2017-05-10 | 郑州云海信息技术有限公司 | Memorysystem with a plurality of disks and a plurality of SSDs |
| CN106933494A (en) * | 2015-12-31 | 2017-07-07 | 伊姆西公司 | Mix the operating method and device of storage device |
| CN107870744A (en) * | 2017-10-27 | 2018-04-03 | 上海新储集成电路有限公司 | The hybrid hard disk array storage system and method for a kind of asynchronous mirror image |
| CN109814811A (en) * | 2019-01-30 | 2019-05-28 | 哈尔滨工业大学 | A method of reducing NVMe SSD operating lag influences high-speed data storage apparatus writing speed |
| CN110109620A (en) * | 2019-04-25 | 2019-08-09 | 上海淇毓信息科技有限公司 | Mix storage method, device and electronic equipment |
| CN111007991A (en) * | 2015-08-13 | 2020-04-14 | 北京忆恒创源科技有限公司 | Method for separating read-write request based on NVDIMM and computer thereof |
| CN114640688A (en) * | 2022-03-23 | 2022-06-17 | 中国兵器装备集团自动化研究所有限公司 | Heterogeneous dual-redundancy equipment running state data storage device and access method |
| CN115016740A (en) * | 2022-07-14 | 2022-09-06 | 杭州优云科技有限公司 | Data recovery method and device, electronic equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110258363A1 (en) * | 2010-04-19 | 2011-10-20 | International Business Machines Corporation | Sub-lun input/output profiling for ssd devices |
| CN102541466A (en) * | 2011-10-27 | 2012-07-04 | 忆正存储技术（武汉）有限公司 | Hybrid storage control system and method |
| CN102662459A (en) * | 2012-04-22 | 2012-09-12 | 复旦大学 | Method for reducing energy consumption of server by using mixed storage of solid-state drive and mechanical hard disk |
| CN102667704A (en) * | 2009-12-22 | 2012-09-12 | 国际商业机器公司 | Hybrid storage subsystem |
| US20130042049A1 (en) * | 2011-08-08 | 2013-02-14 | International Business Machines Corporation | Enhanced copy-on-write operation for solid state drives |
-
2013
- 2013-03-25 CN CN2013100969311A patent/CN103150128A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102667704A (en) * | 2009-12-22 | 2012-09-12 | 国际商业机器公司 | Hybrid storage subsystem |
| US20110258363A1 (en) * | 2010-04-19 | 2011-10-20 | International Business Machines Corporation | Sub-lun input/output profiling for ssd devices |
| US20130042049A1 (en) * | 2011-08-08 | 2013-02-14 | International Business Machines Corporation | Enhanced copy-on-write operation for solid state drives |
| CN102541466A (en) * | 2011-10-27 | 2012-07-04 | 忆正存储技术（武汉）有限公司 | Hybrid storage control system and method |
| CN102662459A (en) * | 2012-04-22 | 2012-09-12 | 复旦大学 | Method for reducing energy consumption of server by using mixed storage of solid-state drive and mechanical hard disk |
Non-Patent Citations (1)
| Title |
|---|
| 陈志广等: "一种用磁盘备份SSD的高性能可靠存储系统", 《计算机研究与发展》 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103713973B (en) * | 2014-01-08 | 2017-11-17 | 浪潮(北京)电子信息产业有限公司 | A kind of mixing storage backup method and system based on HDD and SSD |
| CN103713973A (en) * | 2014-01-08 | 2014-04-09 | 浪潮(北京)电子信息产业有限公司 | Mixed storage backup method and system based on HDD and SSD |
| CN104111894A (en) * | 2014-07-17 | 2014-10-22 | 记忆科技(深圳)有限公司 | Method for improving multi-partition sequential read-write performance and system thereof |
| CN106469119A (en) * | 2015-08-10 | 2017-03-01 | 北京忆恒创源科技有限公司 | A kind of data write buffer method based on NVDIMM and its device |
| CN111007991B (en) * | 2015-08-13 | 2024-01-26 | 北京忆恒创源科技股份有限公司 | Method for separating read-write requests based on NVDIMM and computer thereof |
| CN111007991A (en) * | 2015-08-13 | 2020-04-14 | 北京忆恒创源科技有限公司 | Method for separating read-write request based on NVDIMM and computer thereof |
| CN105630415B (en) * | 2015-12-24 | 2019-04-30 | 创新科软件技术（深圳）有限公司 | A kind of RAID5 system and in the subsequent method for continuing data of RAID5 thrashing |
| CN105630415A (en) * | 2015-12-24 | 2016-06-01 | 创新科软件技术（深圳）有限公司 | RAID5 system and method for continuing to write data after RAID5 system failure |
| CN106933494B (en) * | 2015-12-31 | 2019-10-18 | 伊姆西公司 | The operating method and device of mixing storage equipment |
| US10339021B2 (en) | 2015-12-31 | 2019-07-02 | EMC IP Holding Company LLC | Method and apparatus for operating hybrid storage devices |
| CN106933494A (en) * | 2015-12-31 | 2017-07-07 | 伊姆西公司 | Mix the operating method and device of storage device |
| CN106648458A (en) * | 2016-09-30 | 2017-05-10 | 郑州云海信息技术有限公司 | Memorysystem with a plurality of disks and a plurality of SSDs |
| CN107870744A (en) * | 2017-10-27 | 2018-04-03 | 上海新储集成电路有限公司 | The hybrid hard disk array storage system and method for a kind of asynchronous mirror image |
| CN109814811A (en) * | 2019-01-30 | 2019-05-28 | 哈尔滨工业大学 | A method of reducing NVMe SSD operating lag influences high-speed data storage apparatus writing speed |
| CN109814811B (en) * | 2019-01-30 | 2022-04-29 | 哈尔滨工业大学 | Method for reducing influence of NVMe SSD response delay on write speed of high-speed data storage equipment |
| CN110109620A (en) * | 2019-04-25 | 2019-08-09 | 上海淇毓信息科技有限公司 | Mix storage method, device and electronic equipment |
| CN110109620B (en) * | 2019-04-25 | 2023-08-04 | 上海淇毓信息科技有限公司 | Hybrid storage method and device and electronic equipment |
| CN114640688A (en) * | 2022-03-23 | 2022-06-17 | 中国兵器装备集团自动化研究所有限公司 | Heterogeneous dual-redundancy equipment running state data storage device and access method |
| CN115016740A (en) * | 2022-07-14 | 2022-09-06 | 杭州优云科技有限公司 | Data recovery method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103150128A (en) | Implementation method of solid state drive (SSD) and disk-based reliable mixed storage system | |
| US8307159B2 (en) | System and method for providing performance-enhanced rebuild of a solid-state drive (SSD) in a solid-state drive hard disk drive (SSD HDD) redundant array of inexpensive disks 1 (RAID 1) pair | |
| CN102024044B (en) | Distributed file system | |
| US7984325B2 (en) | Storage control device, data recovery device, and storage system | |
| US8549381B2 (en) | Storage subsystem | |
| US20090077302A1 (en) | Storage apparatus and control method thereof | |
| CN101354633B (en) | Method for improving writing efficiency of virtual storage system and virtual storage system thereof | |
| CN102122235B (en) | RAID4 (redundant array of independent disks) system and data reading and writing method thereof | |
| CN103049220B (en) | Storage controlling method, memory control device and solid-state memory system | |
| US9098404B2 (en) | Storage array, storage system, and data access method | |
| US9195530B1 (en) | Systems and methods for improved data management in data storage systems | |
| TW201324130A (en) | Data storage method for a plurality of RAID systems and data storage system thereof | |
| CN102945207A (en) | Cache management method and system for block-level data | |
| CN103019893A (en) | Multi-disk fault-tolerant two-dimensional hybrid disk RAID4 system architecture and read-write method thereof | |
| CN103713973A (en) | Mixed storage backup method and system based on HDD and SSD | |
| US11853162B2 (en) | Controller and storage device | |
| US11093339B2 (en) | Storage utilizing a distributed cache chain and a checkpoint drive in response to a data drive corruption | |
| CN109739777B (en) | Reliable flash memory storage method and system based on local and distributed collaborative design | |
| CN102609224A (en) | Redundant array of independent disk system and initializing method thereof | |
| CN101169705A (en) | Method for implementing file class mirror-image under multiple hard disk based on nude file system | |
| US20210334022A1 (en) | Zone-based device with control level selected by the host | |
| CN114077517A (en) | Data processing method, equipment and system | |
| US11989457B2 (en) | System and method for performing data read-write operations in zoned storage devices | |
| CN114415968B (en) | Storage system and data writing method thereof | |
| CN101149946B (en) | Method for treating stream media data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130612 |