CN101515255A - Method and device for storing data - Google Patents
Method and device for storing data Download PDFInfo
- Publication number
- CN101515255A CN101515255A CNA2009101294209A CN200910129420A CN101515255A CN 101515255 A CN101515255 A CN 101515255A CN A2009101294209 A CNA2009101294209 A CN A2009101294209A CN 200910129420 A CN200910129420 A CN 200910129420A CN 101515255 A CN101515255 A CN 101515255A
- Authority
- CN
- China
- Prior art keywords
- memory
- data
- blank
- storage
- memory block
- 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
- Memory System Of A Hierarchy Structure (AREA)
Abstract
The embodiment of the invention discloses a method for storing data, which includes steps of obtaining at least two blank memory blocks on different buffer memory plates according to received writing request, wherein each buffer memory plate comprises more than one blank memory block; storing the data in the at least two obtained blank memory blocks, wherein the data stored in the memory blocks is same; and sending the data in one memory block of the at least two memory blocks to a storing medium. The embodiment of the invention also discloses a storing device. When Cache plate is damaged and can not be used, mirror relation of each Cache plate is not needed to construct, after receiving the writing request sent by the host, at least two blank memory blocks on different Cache plates are obtained, the data sent by the host is stored in the obtained memory blocks, thereby increasing operating efficiency of the storing device.
Description
Technical field
The present invention relates to field of computer technology, be specifically related to a kind of storage means and memory storage of data.
Background technology
Storage system in data is in service, and the reliability of storage system is very important, because the loss that is caused of losing of data is difficult to retrieve.Present stage, buffer memory in the storage system (Cache) module has been brought into play vital role because of processing speed on the different caused contradictions solving between processor and the hard disk, mainly show: the one, when user's write operation, when user's write data writes in the Cache module, system will respond write operation and finish, the user carries out the next one and wants the operation carried out, at this moment, in storage system inside, user's write data is collectively written into storage medium (storage medium is disk, hard disk normally) at set intervals in the Cache module.The 2nd, when user's read operation, because deposited the reading frequency higher data in the Cache module, therefore, the request of user's reading of data can not need access storage media, and directly from the Cache module, read, solved the slow problem of physical disk reading speed.In order to prevent that the needs of storing in the Cache module from writing the loss of data in the storage medium, need write the data backup in the storage medium to the needs of storing in the Cache module, the data (being dirty data Dirty Data) that are stored in the Cache module and also do not store in the storage medium are backed up, prevent when storing dirty data into disk, loss of data takes place, and causes forever losing of data.
For the method that in the easy to understand prior art dirty data in the Cache module is backed up, at first the logical architecture in the storage system is done explanation, referring to shown in Figure 1, be a kind of storage system logic framework that can realize the backup of the dirty data in the Cache module, include controller A and controller B in this storage system, the structure of each controller is identical, all comprise control module, Subscriber Interface Module SIM, Cache module and disk interface module, the data that the user will write write the Cache module by Subscriber Interface Module SIM, become dirty data, dirty data is written in the disk by the disk interface module, more than said process control by control module.Dirty data among the controller A in the Cache module backups in the Cache module among the controller B by mirror-image channels, and in like manner, the dirty data among the controller B in the Cache module also can pass through mirror-image channels, backups in the Cache module among the controller A.For the ease of the following description, the Cache module that needs to understand is the Cache plate in physics realization, and promptly the Cache module is identical with the Cache plate, and such understanding is all arranged in full.
In the existing high-end memory device, a plurality of controllers are arranged usually, the structure of each storer is similar to controller A, and the dirty data of Cache module is realized backup by the circulation mirror method in the controller.Referring to shown in Figure 2, the arrow among the figure is represented the memory address of data backup or mirror image, and the dirty data in the previous Cache plate is mirrored in the next Cache plate, and the dirty data in last Cache plate is mirrored in first Cache plate.According to this mirror that circulates, carry out backup more than each controller in the memory device of two controllers to dirty data.A damage is arranged in a plurality of Cache plates, be assumed to be Cache plate 2 among Fig. 2, data in this Cache plate 2 are stored in the storage medium timely, when not having new Cache plate to replace the Cache plate 2 of this damage, need to make up the mirror between Cache plate 1 and the Cache plate 3, the write data that is about in the Cache plate 1 stores in the Cache plate 3; When new Cache plate is replaced the Cache plate 2 that damages, need to make up Cache plate 1 and new Cache plate, new Cache plate and Cache plate 3 these two mirrors, the write data that is about in the Cache plate 1 is mirrored in the new Cache plate, and the write data that configuration receives new Cache plate is mirrored to the information in the Cache plate 3.
In research and practice process to prior art, the present inventor finds, in this memory device with a plurality of controllers, if wherein the Cache plate in certain controller has damaged, replace before the Cache plate that damages at the Cache plate made good use of and afterwards, all need to rebuild the mirror between each Cache plate, be unfavorable for improving the work efficiency of memory device.
Summary of the invention
The embodiment of the invention provides a kind of storage means and memory storage of data, breaks mirror fixing between each Cache plate, realizes improving the purpose of memory device work efficiency.
The embodiment of the invention provides a kind of storage means of data, comprising:
According to the write request that receives, obtain not at least two blank memory blocks on same buffer memory plate, each buffer memory plate comprises the blank memory block more than 1;
In the blank memory block of described at least two of obtaining, the data of storing in the described memory block are identical data with data storage;
Data in one of them memory block in described at least two blank memory blocks are sent to storage medium.
The embodiment of the invention also provides a kind of memory storage of data, comprising:
Storage unit comprises the buffer memory plate more than at least two, and each buffer memory plate comprises the memory block more than 1;
Receiving element is used to receive write request;
Acquiring unit, be used for according to the write request that receives, obtain not at least two blank memory blocks on same buffer memory plate, in the blank memory block of described at least two of obtaining, the data of storing in described at least two memory blocks are identical data with data storage;
First transmitting element is used for the data in one of them memory block of described at least two memory blocks are sent to storage medium.
The embodiment of the invention adopts memory storage data to be write at least two blank memory blocks that do not belong to same Cache plate respectively, owing to write the blank memory block that obtains before the data and be the memory block from the operable storage space that each Cache plate is formed at every turn, not fixedly to belong to certain Cache plate, when the damage of Cache plate can not be used, do not need to rebuild the mirror between each Cache plate, but after receiving write request, just need obtain not at least two blank memory blocks on same buffer memory Cache plate, with data storage in described at least two memory blocks that obtain, thereby improved the work efficiency of memory device.
Description of drawings
In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art, to do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art below, apparently, accompanying drawing in describing below only is some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the simple and easy logical architecture that has the memory storage realization mirror image of two controllers in the prior art;
Fig. 2 is the simple and easy logical diagram that has in the prior art more than data backup in the memory storage of two controllers;
Fig. 3 is the simple and easy process flow diagram of the storage means of a kind of data of providing of the embodiment of the invention one;
Fig. 4 is a simplified schematic diagram of representing mirror between the memory block in the embodiment of the invention one;
Fig. 5 is the simple and easy process flow diagram of the storage means of a kind of data of providing of the embodiment of the invention two;
Fig. 6 is the logical block figure of a kind of memory storage of providing of the embodiment of the invention three;
Fig. 7 is the module map of a kind of memory storage of providing of the embodiment of the invention four.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the invention, the technical scheme in the embodiment of the invention is clearly and completely described, obviously, described embodiment only is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that is obtained under the creative work prerequisite.
Embodiment one
The embodiment of the invention provides a kind of storage means of data, and referring to shown in Figure 3, this method comprises:
Step 1: receive write request;
Wherein, the write request that memory storage receives in the step 1 can directly be sent by main frame, also can be that main frame arrives this memory storage by forwarded.
Step 2: according to the write request that receives, obtain the not memory block of at least two blank on same buffer memory Cache plate, each Cache plate comprises the memory block more than 1;
Wherein, it is identical that the Cache plate in the step 2 and Cache module can be thought, a plurality of Cache modules can be a plurality of Cache plates in physics realization, and the storage space on each Cache plate can be divided into a plurality of memory blocks again.As shown in Figure 4, can think that each Cache plate is divided into three parts, comprising: read the space, write space and mirror image space, wherein, the data that the mirror image space is used for storing are backups of the write data of other Cache plates.As shown in Figure 4, suppose to comprise in this memory device 4 Cache plates, each Cache plate in this memory device be not fix as mirror of the prior art, but the control of the storage in the Cache plate is divided into littler memory block, according to the mirror image between the memory block, reach the purpose of Backup Data.Arrow among Fig. 4 is represented the mirror of two memory blocks.
The Cache plate is divided into the memory block more than 1 in embodiments of the present invention, the memory source that is different from the prior art storage data is to be unit with the Cache plate, the memory source of storing data in the present embodiment is to be unit with littler memory block, therefore, do not need to be provided with specially the mirror between each Cache plate, but, when needs storage data, obtain at least two blank memory blocks, set up the mirror between the memory block.The memory block that the Cache plate is divided is many more, and the available internal memory resource is abundant more.
Also need to prove, obtain at least two blank on same Cache plate not memory block can at random, Hash (HASH) algorithm or according to wherein any method of the load condition of Cache plate, selecting is not memory block on the same Cache plate.
Step 3: data storage at least two that obtain blank memory blocks, is made that the data of every storage are identical in the memory block;
Wherein, main frame sends in the step 3 or network in the data that send of transmission equipment will be by the user interface in the memory storage, the data that make main frame send are stored in the memory block that obtains.The data that are written in two or more the memory block are identical, and the data in two or more the memory block are mirror image (perhaps backuping each other) each other.
Step 4: the data in one of them memory block at least two memory blocks after the complete step 3 are sent to storage medium.
By step 4, the data storage in the buffer memory plate in storage medium (as disk), has been realized the reliable memory to data.
The explanation of the storage means by above a kind of data that the embodiment of the invention is provided, memory storage writes the data that main frame sends respectively in two or more memory blocks that do not belong to same Cache plate, owing to write the blank memory block that obtains before the data and be the memory block from the operable storage space that each Cache plate is formed at every turn, not fixedly to belong to certain Cache plate, when the damage of Cache plate can not be used, do not need to rebuild the mirror between each Cache plate, but after the write request that receives the main frame transmission, just need obtain not two or more blank memory blocks on same buffer memory Cache plate, the data storage that main frame is sent in described two or more memory blocks that obtain, thereby improved the work efficiency of memory device.
Need to prove that this method can also comprise:
Step 5: preserve the mirror between at least two blank memory blocks that obtain;
Need to prove, at least two blank memory blocks that obtain in the step 2 are to be used for to storing identical data, promptly data are backed up, therefore, the mirror of arranging in the storer in difference and the prior art, in step 4, need to set up the mirror of two or more memory blocks, guarantee that a memory block (for example: when disk) sending mistake is writing data storage medium, can get access to Backup Data according to this mirror, again Backup Data is write storage medium, avoid loss of data.
The relation of establishing mirror image can be the corresponding relation of setting up between two or more memory blocks, when address that learns one of them memory block or sign, just can know the address with the memory block of its mirror image.
By above this method is increased step 5, make memory device when the write operation mistake, can be easy to obtain backed up data, avoid loss of data.
Embodiment two
The embodiment of the invention provides a kind of storage means of data, this method is similar to embodiment one, can realize when having the Cache plate to damage in the memory storage, replace back Cache plate and be used as the memory block resource, and do not need to make up immediately the mirror of new Cache plate.Present embodiment is specifying of embodiment one, below this embodiment is done explanation.
In conjunction with Fig. 5 present embodiment is done explanation, this method comprises:
Steps A 1: receive the write request that main frame sends;
The write request that main frame sends in the steps A 1 can be that main frame directly sends to memory storage, also can be that main frame passes through certain network (as certain LAN (Local Area Network)) and sends to memory storage.
Steps A 2: according to the write request that receives, judge whether the load of each Cache plate is identical, if, enter steps A 3, if not, enter steps A 4;
Wherein, need to prove that the Cache plate can be divided into three parts according to the purposes of storage data, comprising: read the space, write space and mirror image space.Therefore, the load on the Cache plate comprises three kinds: the one, be stored on the Cache plate, and send to the data that main frame is read; The 2nd, be stored in the data that will write disk on the Cache plate; The 3rd, the mirror image data of loss of data, i.e. Backup Data when preventing write operation.Judging in the memory storage whether load is the same on each Cache plate, can specifically be to judge operating factor of memory space on each Cache plate, and the load of the Cache plate that utilization factor is high is heavy, and the load of the Cache plate that utilization factor is low is light.
Steps A 3: select two Cache plates arbitrarily, from these two Cache plates, obtain the memory block of each blank, enter steps A 5;
Steps A 4: select load according to by light to heavy order, be arranged in two Cache plates of first and second, from these two Cache plates, obtain the respectively memory block an of blank, enter steps A 5;
Wherein, also need to prove, the normally identical size of obtaining in steps A 3 or the steps A 4 of the storage space that belongs to two blank memory blocks on the Different Ca che plate, be that a kind of simple way is that the storage size of the memory block divided on each Cache plate in this memory storage is all identical, the concrete numerical value of the storage size of each memory block can be set according to designing requirement, usually, the memory block that the Cache plate is divided is many more, make that the memory source of storage data is abundanter, do not need to set up specially the relationship memory between the Cache plate, but the mirror between the stored memory piece has as required improved the work efficiency of memory storage.
Steps A 5: data are stored into respectively in two blank memory blocks that obtain, and wherein, the data of storing in two memory blocks are identical;
Steps A 6: the data in one of them memory block that stores in the steps A 5 in two memory blocks of data are sent to disk.
By steps A 6 with the data storage in the buffer memory in memory device.
The method that present embodiment provides can realize the one constructed effect with embodiment, and the data that are about to the main frame transmission write respectively in two blank memory blocks that do not belong to same Cache plate; When replacing when certain Cache plate damage and by good Cache plate, new Cache plate after the replacement is used as the memory block resource, does not need to rebuild mirror, but will write fashionable as data, just need obtain two blank memory blocks, improve the work efficiency of memory storage.By judging the load condition of each Cache plate in the memory storage, select to obtain two blank memory blocks on the Cache plate that do not coexist in this method, can realize the load balancing on each Cache plate, the performance of the memory storage of raising.
Need to prove that also in order to guarantee can to get access to backed up data rapidly when the loss of data, this method also further comprises:
Steps A 7: preserve the mirror between two memory blocks;
Wherein, the mirror of preserving in the steps A 7 specifically can be: set up two corresponding relations between the memory block, when the address that obtains one of them memory block obtains sign, just can obtain the address of the memory block of mirror image with it, thereby make things convenient for memory device to get access to Backup Data (the perhaps data of mirror image).
Need to prove also that memory storage is being carried out the data storage that once main frame sent in the process of disk, the storage size of memory block is normally fixing, therefore, the size that the data of main frame transmission may occur has exceeded the memory capacity of a memory block, and therefore, this method also comprises:
Steps A 8: judge whether the data of main frame transmission have surpassed the capacity of a memory block that obtains, if enter steps A 2; If not, then enter steps A 8 or end operation.
Therefore, the storage size of memory block also should be considered the size of each dirty data, and for the bigger situation of the dirty data of write-once, blank memory block is obtained in the data recirculation execution that will exceed the memory block capacity, makes that this method is more perfect.
Steps A 9: send to write and finish response to main frame.
By increasing steps A 9, make things convenient for main frame to write and finish other operations of response execution according to this.
More than to the explanation of steps A 1 to A9, make memory device when having the Cache plate to damage, the new Cache plate of replacement does not need to rebuild mirror, but writes fashionable as data, storage space on the new Cache plate as the memory block resource, is used in order to subsequent operation; By judging and selecting the light Cache plate of load, make load balancing on the Cache plate, improved the performance of memory storage;
Embodiment three
The embodiment of the invention provides a kind of memory storage, participates in shown in Figure 6ly, and this memory storage comprises: receiving element 100, acquiring unit 200, storage unit 300 and the first transmitting element 600a.
Wherein, receiving element 100 is used to receive the write request that main frame sends;
Acquiring unit 200, be used for according to write request, obtain the not memory block of at least two blank on same buffer memory Cache plate, in the blank memory block of described at least two of obtaining, the data of storing in described at least two memory blocks are identical data with data storage;
The first transmitting element 600a is used at least two data that store one of them memory block of data are sent to storage medium.
Need to prove, comprise the Cache plate more than 2 in the storage unit 300, comprise more than 1 memory block on each Cache plate.
After receiving element 100 receives the write request of main frame transmission, notice acquiring unit 200, acquiring unit 200 is according to the write request that receives, obtain at least two blank memory blocks that are not on the same buffer memory Cache plate, wherein, comprise memory block on each Cache plate more than 1, the data storage that main frame is sent is in two or more the memory block that obtains, and the data of storing in two or more the memory block are identical, therefore, data in two or more the memory block become mirror image data each other, when a Cache plate damages, memory storage can get access to data from the memory block of mirror image, store data into storage medium (as disk) again, thereby avoided loss of data.
By above explanation to present embodiment, the not memory block of at least two blank on same Cache plate that this memory storage obtains according to acquiring unit 200, the data that main frame is sent write respectively in each memory block that obtains that obtains, and the data in one of them memory block are sent to storage medium by the first transmitting element 600a.Owing to write the blank memory block that obtains before the data and be the memory block from the operable storage space that each Cache plate is formed at every turn, not fixedly to belong to certain Cache plate, when the damage of Cache plate can not be used, do not need to rebuild the mirror between each Cache plate, but after the write request that receives the main frame transmission, acquiring unit 200 just obtains not at least two blank memory blocks on same buffer memory Cache plate, the data storage that main frame is sent in the described memory block that obtains, thereby improved the work efficiency of memory device.
Optionally, the acquiring unit 200 in this memory storage can specifically comprise: judge load cell 201 and selected cell 202.
Wherein, judge load cell 201, be used for, judge whether the load of each Cache plate is the same according to the write request that receives; Judge that in judging load cell 201 load on each Cache plate is identical, then notify selected cell 202, selected cell 202 any selections are not each blank memory blocks on the same Cache plate; In judging load cell 201, judge load that there emerged a on the Cache plate not simultaneously, selected cell 202 select load according to by light to heavy order, be arranged in the respectively blank memory block on two Cache plates of first and second.
Owing to judge whether the load on each Cache plate is identical according to judging in the load cell 201 in the acquiring unit 200, select blank memory block on the light Cache plate of load by selected cell 202, make load balancing on each Cache plate, improved the system performance of memory storage.
Optionally, this memory storage can also comprise:
Wherein, storage concerns the mirror that unit 400 is stored between two memory blocks that obtain, so that when a Cache plate damages, can be easy to get access to backed up data.
Optionally, this memory storage can also comprise:
Judge and to overflow unit 500 whether the size of the data that are used to judge that main frame sends surpasses the capacity of a memory block that obtains;
The second transmitting element 600b is used for sending to write and finishing response to main frame when judging that overflow unit 500 judges the size of data and do not surpass the capacity of a memory block that obtains.
Wherein, if the size of data has surpassed the capacity of a memory block that obtains, notice acquiring unit 200, obtain two blank memory blocks on the Cache plate that do not coexist again, specifically can be if size of data has surpassed the capacity of a memory block, judge and overflow unit 500 the judgement load cell 201 in this result notification acquiring unit 200; If the size of data does not have to surpass the capacity of a memory block, then end operation is perhaps notified the second transmitting element 600b; The second transmitting element 600b sends to write and finishes response to main frame.Wherein, because the storage space of two pieces is normally identical, and the data of the storage in two memory blocks that obtain are identical, and when the size of data of main frame transmission surpassed the capacity of a memory block, size of data had also surpassed the capacity of memory block in the memory block of certain mirror image with it.Therefore, need circulation to carry out and obtain blank memory block, with data storage in memory block.
Embodiment four
Present embodiment provides a kind of memory storage, this embodiment is similar to a kind of memory storage that embodiment three provides, referring to shown in Figure 7, this memory storage 10 comprises: interface module 101, Switching Module 102, administration module 103, buffer memory Cache plate 104 and disk node 105.
Wherein, interface module 101 is used to receive the write request that the transmission equipment of main frame or network sends, with the receiver module functional similarity among the embodiment three.Outside equipment can be main frame, perhaps the transmission equipment in the network.Comprise a plurality of interface modules in the interface module 101, as interface module 101a among the figure, interface module 101b, interface module 101c, interface module 101d, different interface modules is represented the communication protocol of different main frames and memory storage usually, as: interface module 101a adopts optical-fibre channel (FC, Fiber Channel) communicate with main frame, interface module 101b adopts serial line interface (SAS, Serial Attached SCSI) and main-machine communication etc.Need the communication protocol of main frame and memory storage multiple in addition, do not do qualification herein.
Switching Module 102 can comprise a plurality of, as figure Switching Module 102a and Switching Module 102b, this sentences interface module 101a is example, Switching Module 102a is according to the strategy in the administration module 103, the data that interface module 101a is sent send to corresponding Cache plate, and perhaps, the data that the data on the Cache plate are sent send to corresponding interface module 101a, transfer out this memory storage from interface module 101a, all the other interface modules are identical with the function of interface module 101a; Switching Module 102b is according to the strategy in the administration module 103, with the data on the different Cache plates, and storage in the disk node 105 of the correspondence that sends to by this Switching Module 102b, Switching Module can be similar to explanation of the prior art.
Other modules in administration module 103 these memory storages of control, specify and comprise: after memory storage receives the write request of main frame transmission by its interface module 101, administration module 103 is according to this write request, obtain not two blank memory blocks on same Cache plate, each Cache plate comprises the memory block more than 1; The data storage that administration module 103 control sends main frame is in two blank memory blocks that obtain, the data that administration module 103 will store on one of them memory block in two memory blocks of data send to the disk node, the disk node receives the data that one of them memory block sends, and is stored in the disk node.
Need to understand the time, by having comprised the acquiring unit 200 of explanation among the embodiment three in the above explanation administration module 103, can also comprise in this administration module 103 that the judgement among the embodiment three overflows unit and transmitting element.But, it will be appreciated that, this administration module 103 cannot simply be thought the acquiring unit 200 among the embodiment three, this administration module can also comprise: judge and overflow unit 500, the first transmitting element 600a and the second transmitting element 600b, this administration module 103 is the control centers in this memory storage.
Understand for convenience, also to describing that the logical architecture beyond this memory storage carries out, promptly one or more main frame 70 can be connected to this memory storage by certain network 80 among Fig. 7, make main frame can with this storage communication.
Explanation by above a kind of memory storage that present embodiment is provided, administration module 103 in this memory storage writes the data that main frame sends respectively in two memory blocks that do not belong to same Cache plate, owing to write before the data at every turn, the blank memory block that administration module 103 obtains is the memory block from the operable storage space that each Cache plate is formed, not fixedly to belong to certain Cache plate, when the damage of Cache plate can not be used, do not need to rebuild the mirror between each Cache plate, but after the write request that receives the main frame transmission, just need obtain not two blank memory blocks on same buffer memory Cache plate, the data storage that main frame is sent in described two blank memory blocks that obtain, thereby improved the work efficiency of memory device.
One of ordinary skill in the art will appreciate that all or part of step in the whole bag of tricks of the foregoing description is to instruct relevant hardware to finish by program, this program can be stored in the computer-readable recording medium, and storage medium can comprise: ROM, RAM, disk or CD etc.
More than the storage means of a kind of data that the embodiment of the invention provided and the device of response are described in detail, used specific case herein principle of the present invention and embodiment are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (13)
1, a kind of storage means of data is characterized in that, comprising:
According to the write request that receives, obtain not at least two blank memory blocks on same buffer memory plate, each buffer memory plate comprises the blank memory block more than 1;
In the blank memory block of described at least two of obtaining, the data of storing in the described memory block are identical data with data storage;
Data in one of them memory block in described at least two memory blocks are sent to storage medium.
2, method according to claim 1 is characterized in that, the write request that described basis receives is obtained not at least two blank memory blocks on same buffer memory plate, and each buffer memory plate comprises the memory block more than 1, further comprises:
According to write request that receives and the hash algorithm that presets, obtain the not memory block of at least two blank on same buffer memory plate.
3, method according to claim 1 is characterized in that, the write request that described basis receives is obtained not at least two blank memory blocks on same buffer memory plate, and each buffer memory plate comprises the memory block more than 1, further comprises:
According to the write request that receives, judge whether the load on each buffer memory plate is identical, if judged result is that the load on each buffer memory plate is inequality, then from load according to by gently to heavy order, be arranged in two buffer memory plates of first and second and respectively obtain a blank memory block.
4, method according to claim 3 is characterized in that, when the load on each buffer memory plate of described judgement when being identical, described method also comprises: from respectively obtaining a blank memory block at least two buffer memory plates arbitrarily.
5, according to claim 1,3 or 4 each described methods, it is characterized in that described method also comprises:
Whether the size of judgment data surpasses the wherein capacity of a memory block in the blank memory block of described at least two of obtaining, if repeat the described not step of at least two blank memory blocks on same buffer memory plate of obtaining.
6, method according to claim 5, it is characterized in that, when described judgment data size did not surpass the capacity of the wherein memory block in the blank memory block of described at least two of obtaining, described method also comprised: send to write and finish response to described main frame.
7, method according to claim 1 is characterized in that, described method also comprises:
Preserve the mirror between the blank memory block of described at least two of obtaining.
8, a kind of memory storage is characterized in that, comprising:
Storage unit comprises the buffer memory plate more than at least two, and each buffer memory plate comprises the memory block more than 1;
Receiving element is used to receive write request;
Acquiring unit, be used for according to the write request that receives, obtain not at least two blank memory blocks on same buffer memory plate, in the blank memory block of described at least two of obtaining, the data of storing in described at least two memory blocks are identical data with data storage;
First transmitting element is used for the data in one of them memory block of described at least two memory blocks are sent to storage medium.
9, memory storage according to claim 8 is characterized in that, described acquiring unit further comprises:
Judge load cell, be used for judging according to the write request that receives whether the load on each buffer memory plate is identical;
Selected cell is used for being the load on each buffer memory plate when inequality when the judged result of described judgement load cell, from load according to by gently to heavy order, be arranged in two buffer memory plates of first and second and respectively obtain a blank memory block.
10, memory storage according to claim 9 is characterized in that, selected cell also is used for being the load on each buffer memory plate when identical, from respectively obtaining a blank memory block at least two buffer memory plates arbitrarily when the judged result of described judgement load cell.
11, according to Claim 8 to 10 each described memory storages, it is characterized in that described memory storage also comprises:
Judge and overflow the unit, be used for the wherein capacity of a memory block whether the judgment data size surpasses described at least two blank memory blocks that obtain, if notify described acquiring unit.
12, memory storage according to claim 11 is characterized in that, described memory storage also comprises:
Second transmitting element is used for whether surpassing when the size of described judgment data the wherein capacity of a memory block of described at least two blank memory blocks that obtain, and sends to write and finishes response to described main frame.
13, memory storage according to claim 8 is characterized in that, described memory storage also comprises:
Storage concerns the unit, is used to preserve the mirror between the blank memory block of described at least two of obtaining.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2009101294209A CN101515255A (en) | 2009-03-18 | 2009-03-18 | Method and device for storing data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2009101294209A CN101515255A (en) | 2009-03-18 | 2009-03-18 | Method and device for storing data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101515255A true CN101515255A (en) | 2009-08-26 |
Family
ID=41039717
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2009101294209A Pending CN101515255A (en) | 2009-03-18 | 2009-03-18 | Method and device for storing data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101515255A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102640118A (en) * | 2009-10-16 | 2012-08-15 | 赛门铁克公司 | De-duplication Storage System With Multiple Indices For Efficient File Storage |
| CN102880277A (en) * | 2012-09-05 | 2013-01-16 | 浪潮(北京)电子信息产业有限公司 | Protection method for uninterrupted power supply redundancy of double-control disk array |
| WO2014000300A1 (en) * | 2012-06-30 | 2014-01-03 | 华为技术有限公司 | Data buffer device, data storage system and method |
| CN105045685A (en) * | 2015-08-04 | 2015-11-11 | 浪潮电子信息产业股份有限公司 | Backup method and apparatus for cached data and recovery method and apparatus for cached data |
| CN107678689A (en) * | 2017-09-27 | 2018-02-09 | 郑州云海信息技术有限公司 | A kind of method and device of data write-in |
| CN108984104A (en) * | 2017-06-02 | 2018-12-11 | 伊姆西Ip控股有限责任公司 | Method and apparatus for cache management |
-
2009
- 2009-03-18 CN CNA2009101294209A patent/CN101515255A/en active Pending
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102640118A (en) * | 2009-10-16 | 2012-08-15 | 赛门铁克公司 | De-duplication Storage System With Multiple Indices For Efficient File Storage |
| CN102640118B (en) * | 2009-10-16 | 2015-09-09 | 赛门铁克公司 | For the deduplication storage system with multiple index that efficient File stores |
| WO2014000300A1 (en) * | 2012-06-30 | 2014-01-03 | 华为技术有限公司 | Data buffer device, data storage system and method |
| CN102880277A (en) * | 2012-09-05 | 2013-01-16 | 浪潮(北京)电子信息产业有限公司 | Protection method for uninterrupted power supply redundancy of double-control disk array |
| CN105045685A (en) * | 2015-08-04 | 2015-11-11 | 浪潮电子信息产业股份有限公司 | Backup method and apparatus for cached data and recovery method and apparatus for cached data |
| CN108984104A (en) * | 2017-06-02 | 2018-12-11 | 伊姆西Ip控股有限责任公司 | Method and apparatus for cache management |
| CN108984104B (en) * | 2017-06-02 | 2021-09-17 | 伊姆西Ip控股有限责任公司 | Method and apparatus for cache management |
| CN107678689A (en) * | 2017-09-27 | 2018-02-09 | 郑州云海信息技术有限公司 | A kind of method and device of data write-in |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101055511B (en) | Memory array system and its data operation method | |
| CN101446924B (en) | Method and system for storing and obtaining data | |
| US20090049240A1 (en) | Apparatus and method for storage management system | |
| CN101515255A (en) | Method and device for storing data | |
| CN106066890B (en) | Distributed high-performance database all-in-one machine system | |
| CN105242881A (en) | Distributed storage system and data read-write method for same | |
| CN101808127B (en) | Data backup method, system and server | |
| CN107329704B (en) | Cache mirroring method and controller | |
| CN104423889A (en) | Multi-path management method and system | |
| CN105934793A (en) | Method for distributing data in storage system, distribution apparatus and storage system | |
| US9336093B2 (en) | Information processing system and access control method | |
| CN102215268A (en) | Method and device for transferring file data | |
| KR20170027921A (en) | Server device includig interface circuits, memory modules and switch circuit connecting interface circuits and memory modules | |
| CN104424052A (en) | Automatic redundant distributed storage system and method | |
| CN101944066A (en) | Solid state disk, interface processing method thereof and storage system | |
| CN101710270A (en) | High-speed mass memory based on flash memory and chip data management method | |
| CN103577125A (en) | Cross controller group mirror image writing method and device applied to high-end disk array | |
| CN105930357A (en) | Distributed file system, and data node data storage processing method and device | |
| CN101510837B (en) | Ethernet bridge equipment, method and equipment for migrating polymerization mouth address | |
| US20090259814A1 (en) | Memory control apparatus and method for controlling the same | |
| CN103139098A (en) | Method, device and system of data transmission control, and fiber channel switch | |
| CN102982033A (en) | Small documents storage method and system thereof | |
| CN105068896B (en) | Data processing method and device based on RAID backup | |
| CN104572374A (en) | Storage processing method, storage processing device and storage equipment | |
| CN101714064A (en) | Data access method and server |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20090826 |