CN105808150A - Solid state disk caching system for hybrid storage device - Google Patents
Solid state disk caching system for hybrid storage device Download PDFInfo
- Publication number
- CN105808150A CN105808150A CN201410834245.4A CN201410834245A CN105808150A CN 105808150 A CN105808150 A CN 105808150A CN 201410834245 A CN201410834245 A CN 201410834245A CN 105808150 A CN105808150 A CN 105808150A
- Authority
- CN
- China
- Prior art keywords
- data
- solid state
- hard disc
- cache
- 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.)
- Granted
Links
Abstract
The present invention discloses a solid state disk caching system for a hybrid storage device. The system comprises a solid state disk which is used for storing caching data and divided into a repeat mode caching zone and a dynamic replacement caching zone; and a caching management module, including an input/output analysis unit which is used for detecting input/output requests for accessing blocks of a hard disk at a plurality of continuous detection time intervals and sequentially storing first data corresponding to the first block to the repeat mode caching zone, and a hot data search unit for detecting input/output requests for accessing the blocks of the hard disk at independent detection time intervals and sequentially storing second data corresponding to the second block to the dynamic replacement caching zone.
Description
Technical field
The present invention relates to a kind of solid state hard disc fast taking system, particularly relate to a kind of solid state hard disc fast taking system for hybrid storage facilities, by storing configuration mechanism, can have higher performance.
Background technology
Traditionally, in order to solve the too high input/output delay issue caused in data center or enterprise due to random access memory live load, it is possible to large-scale and excessive number magnetic hard-disk can be disposed.Consequently, it is possible to more hard disk read heads can reduce the probability on two same disks of continuous reading, and then improve the performance of overall access on hard disk.But, there is multiple shortcoming because of excessive deployment, for instance, more and more protect the use of equipment, exceptional space, system to run and more power and cooling system, and high maintenance cost etc..Additionally, the utilization rate that too much capacity increase causes system may reduce.
Recently, a kind of solution widely adopted is that quickly read (the random or order) taking solid state hard disc coordinates the operation of hard disk.It is obvious that the stocking system utilizing such solution is a hybrid storage facilities.Solid state hard disc is used as the function of a kind of cache.It is to say, only dsc data (data being often used most) is temporarily stored in solid state hard disc, it is provided that access.Once the data of these storages no longer so " heat ", they will be removed and original space is left other dsc data for and used.A lot of non-dsc datas can be stored in hard disk.Such stocking system has a lot of benefits, and first, for spike demand, solid state hard disc cache is provided that spiking behaviour numerical value more preferably.Secondly, solid state hard disc cache under virtual stocking system environment, can shift between the live load being used in conjunction with.Furthermore, allocate dsc data in advance and performance requirement can be done to solid state hard disc cache and respond flexibly.Under the background of solid state hard disc cache transfer, solid state hard disc cache can discharge the data being soon not used in advance, gets other live load soon transferring solid state hard disc, and other live load should be able to predicted or periodically occur.Really, if the peak period time of individual work load can be predicted, solid state hard disc cache also can only operate on single live load.
Solid state hard disc and random access memory all can become fast taking system.Although having excellent storge quality when cache operates together with hard disk, solid state hard disc cache does not defer to the principle of tradition random access memory cache.Its reading is far faster than write, and order read/write is faster than random read/write.The most important thing is, solid state hard disc can lose efficacy after the write of certain number of times, thus when running such hybrid storage facilities, the write controlling solid state hard disc cache is most important for opportunity.
Many existing patents provide new technology to meet above demand.Such as, U.S. Patent Publication No. the 20140244959th discloses a kind of stocking system having for different kenel storage facilities control devices.This stocking system comprises: a hard disc storage equipment；One solid state hard disc storage facilities；And a storage facilities controller, wherein this storage facilities controller is collected about the load information of each load of multiple districts in this hard disc storage equipment.It based on the load information of those collections, can select one and migrate candidate regions in hard disc storage equipment, and by the Data Migration in the candidate regions of this selection to solid state hard disc storage facilities.In storage facilities controller collection time per unit, the counting of input/output request is as this load information, and the average life that it is also based on from this load time delay selects candidate regions.This average life can calculate in the load elapsed time in each district of hard disc storage equipment by deducting further.It is clear that the selection moving to solid-state hard disc equipment data is based on the time delay that load information is lasting.This system does not include periodically repeating load, and those load possibilities will not continue a very long time, but containing regular and request mode frequently.This repeated boad from hard disk accesses the performance that definitely can injure this stocking system.
U.S. Patent Publication No. the 20140258668th provides in hybrid data storage system for managing the System and method for of storage area.The System and method for of this application is based on data source, data types, data function or other close parameter, and distribution data are to solid state hard disc (or other relatively high performance disk) and other Data Holding Equipment intelligently, in such as hard disk.Smart allocation between at least one solid state hard disc and other disk form at least one, it is allowed to effectively making for promoting systematic function by storage area.Storing in hybrid data and concentrate, this application can choose the fastest storage facilities adaptively from the option of link, maintains maximum performance and low volume data write most effectively simultaneously.Automatically control or specifically by user setup additionally, the storage facilities that data are assigned in hybrid data storage system is probably.Although many factors all consider distribution data, but the hit rate of solid state hard disc cannot be improved, and this further restricts the use of solid state hard disc.
Therefore, in order to operate hybrid storage facilities efficiently, a kind of solid state hard disc fast taking system being used for controlling to use together with other hard disk is in demand.Particularly, this system may operate to improve the hit rate of solid state hard disc use and reach higher performance.
Summary of the invention
In order to solve the problems referred to above, according to an aspect of the present invention, disclose a kind of solid state hard disc fast taking system for hybrid storage facilities.This system comprises a solid state hard disc and cache management module, and wherein, this solid state hard disc is in order to store caching data, and this solid state hard disc is divided into a repeat pattern cache district and a dynamic replacement cache district；And this cache management module comprises an input/output analytic unit and a dsc data searches unit, wherein, this input/output analytic unit is for detecting in multiple continuous detecting intervals, the input/output request that the block of one hard disk is accessed, and sequentially store the first data of corresponding first block to this repeat pattern cache district, in each continuous detecting interval, this first block is at least by repeated access secondary；And this dsc data search unit is for detecting in an independent detection interval, the input/output request that one hard disk is accessed, and sequentially store the second data of corresponding second block to this dynamic replacement cache district, in this independent detection interval, this second block is at least accessed secondary.
According to the present invention, this solid state hard disc fast taking system comprises a write hard disc module further, if to be write instruction requested and when this input/output analytic unit finds that the space, repeat pattern cache district of one first percentage ratio is occupied by data for these first data, in order to directly to store these first data in a hard disk.This first percentage ratio be 90% or any be higher than 90% percentage ratio.If to be write instruction requested and this input/output analytic unit finds that when the space in dynamic replacement cache district of one second percentage ratio is occupied by data, this write hard disc module is further used for directly storing these second data in a hard disk for these second data.This second percentage ratio be 90% or any be higher than 90% percentage ratio.First and second percentage ratios are all the Configuration Values set by system manager and consider system performance, SLA (ServiceLevelAgreement) or QoS (QualityofService), 90% or higher than 90% numerical value preferably.
Solid state hard disc fast taking system comprises a RAM module further, and this RAM module includes district's Virtual table, and wherein this district's Virtual table is a mapping table, for keeping track the Block status of this first block and the second block；If the correspondence reflection block that the logical blocks in district's Virtual table is in this solid state hard disc belongs to dynamic replacement cache district, the logical blocks in Ze district Virtual table is denoted as dynamic replacement cache block；If the correspondence reflection block that the logical blocks in district's Virtual table is in this solid state hard disc belongs to repeat pattern cache district, the logical blocks in Ze district Virtual table is denoted as repeat pattern cache block；When this input/output analytic unit finds that specific second data become first data in this dynamic replacement cache district, the dynamic replacement cache block of specific second data in this dynamic replacement cache district of then videoing is denoted as repeat pattern cache block, and this dynamic replacement cache district Central Plains data content does not do any change.
When preference, when this input/output analytic unit finds a group the first block by repeated access, before in next time, this first block is accessed, the first corresponding data will be stored in this repeat pattern cache district.When this input/output analytic unit finds that a group the first block is not repeated access, after in next time, this first block is accessed, the first corresponding data will remove from this repeat pattern cache district.When new the first data of a group and/or the second data predicting can be accessed at following particular point in time, this input/output analytic unit removes a part of space in this repeat pattern cache district and/or a part of space in this dynamic replacement cache district.One cache replacement algorithm is searched unit by this dsc data and is adopted, to remove the second data of a part in dynamic replacement cache district.This cache replacement algorithm is minimum nearest use (LeastRecentlyUsed, LRU) algorithm, minimum more often use (LeastFrequentlyUsed, LFU) algorithm, or self adaptation replacement cache (AdaptiveReplacementCaching, ARC) algorithm.Being searched the algorithm that uses of unit by the dsc data in this dynamic replacement cache district is from those cache replacement algorithm, according to selected by the occupation mode that storage facilities is different, or stops being applied to read or write instruction.When hybrid storage facilities run into routinely, input and output operation number per second high low with cache hit rate time, for avoiding low usefulness and insignificant cache, this dsc data is searched unit and is stopped storing the second data by reading instruction request.The form of this cache management module and write hard disc module is the adapter in server or a connector, or is installed on the software in server.
The present invention utilizes the advantage in repeat pattern cache district and the dynamic replacement cache district separated in solid state hard disc, each stores (or cache) and repeats and commonly used data.Other solid state hard disc cache mechanism may be lost reusable data and not get soon, so can reduce the usefulness of integer composite storage facilities.Thus, the present invention smooth and easy can operate a set of hybrid storage facilities efficiently.The most important thing is, the hit rate of solid state hard disc cache can be improved.
Accompanying drawing explanation
Fig. 1 is according to the present invention schematic diagram for the fast taking system of hybrid storage facilities；
Fig. 2 is the record accessing block in continuous detecting interval；
Fig. 3 is the record accessing block in independent detection interval；
Fig. 4 is distributed at a kind of live load of the hybrid storage facilities of ordinary day；
Fig. 5 is distributed at the another kind of live load of the hybrid storage facilities of ordinary day.
Description of reference numerals: 10-fast taking system；The hybrid storage facilities of 20-；100-solid state hard disc；102-repeat pattern cache district；104-dynamic replacement cache district；200-server；202-cache management module；2022-input/output analytic unit；2024-dsc data searches unit；204-writes hard disc module；206-RAM module；2062-district Virtual table；302-hard disk；304-hard disk；401-applied host machine；402-applied host machine；A1, A2, A3, B1, B2, B3, B4, C1, C2, C3, C4-block.
Detailed description of the invention
The present invention is by more particularly described below with reference to following embodiment.
Referring to Fig. 1, one embodiment of the invention is illustrated by this figure.One fast taking system 10 is used for improving the performance of a hybrid storage facilities 20.In the present embodiment, hybrid storage facilities 20 is made up of a solid state hard disc 100 and two hard disks 302 and 304.Being not limited to one or two according to the quantity of the spirit of the present invention, solid state hard disc and hard disk, need the demand of running according to a system, they can be any quantity.Hybrid storage facilities 20 is likely to be present in data center with the form of disk array, and it can also be only the mixed dish (SolidStateHardDrive, SSHD) of a solid state hard disc in single main frame.In the present embodiment, this hybrid storage facilities 20 is a disk array.
Fast taking system 10 includes solid state hard disc 100 and a server 200, and fast taking system 10 is to be reached by changing the original operating type of solid state hard disc 100.Therefore, any solid state hard disc with data below structure in hybrid storage facilities, all it is considered the practice of the present invention.Solid state hard disc 100 is divided into Liang Ge district: repeat pattern cache district 102 and a dynamic replacement cache district 104.Solid state hard disc 100 is used for storing caching data, and not all is from the data of applied host machine 401 and 402 write instruction.Repeat pattern cache district 102 and dynamic replacement cache district 104 also can be used to store caching data, only different classes of caching data.
Server 200 includes a cache management module 202, and writes hard disc module 204 and a RAM module 206, each item in these elements all will be described under.Cache management module 202 includes an input/output analytic unit 2022 and a dsc data searches unit 2024.Input/output analytic unit 2022 can detect the input/output request that hard disk 302 or 304 intra block is accessed.This that is input/output analytic unit 2022 detect the write instruction of hard disk 302 or 304 and read instruction, and its relevant block.This detection process, before the time instantly, continues multiple continuous detecting intervals.If in each continuous detecting interval, having the first data of corresponding first block, wherein the first block is at least by repeated access secondary, then those first data are sequentially stored in repeat pattern cache district 102.Preferably understand in order to this point is had, refer to Fig. 2.
Fig. 2 is the record being accessed block.Show the block array in hard disk 302 on the left of Fig. 2, figure only illustrates 2 blocks in order to illustrate.It is true that a hard disk is likely to comprise thousands of block, Fig. 2 is an illustration for purpose example, is not intended to limit the present invention.Some block is denoted as A1, A2, A3, B1, B2, B3, B4, C1, C2, C3 and C4, and they are detected and were once accessed (is not herein, it will be apparent that " access " refers to the reading instruction received, write instruction.In general, read instruction and be suitable for the present invention with write instruction).A1 to A3 represents complete data.B1 and B2 represents that data and B3 and B4 represent another data, and the two data are accessed simultaneously.C1 to C4 represents the data that another is complete.The sequential of right side description block access.Time t0 represents present.Ta1, ta2, ta3, be time point in the past with ta4.Interval between any two adjacent time points is exactly continuous detecting interval, for instance between ta2 and ta3.Can see that A1, A2, and A3 is accessed 3 times between ta1 to t0, B1, B2, B3 and B4 occur 2 times, respectively between ta3 to ta2 and ta2 to ta1.C1, C2, C3 and C4 occur in each continuous detecting interval.If definition is the first block at all blocks being accessed of each continuous detecting interval, the first data of its correspondence can by sequentially cache (storage) to repeat pattern cache district 102, then C1, C2, C3 and C4 are the first blocks.
If the block that definition is all accessed in 3 continuous detecting intervals is the first block, first data of its correspondence can by sequentially cache (storage) to repeat pattern cache district 102, and that B1, B2, B3, B4, C1, C2, C3 and C4 are the first blocks.These first blocks are sequentially stored in the use in repeat pattern cache district 102 for order reading in future.Because those blocks of application claims are in each continuous detecting interval, should at least by repeated access twice, although A1, A2, and A3 is concentrated access between ta1 to t0, and they are not the first blocks.Simultaneously, minimum access times can change according to the performance requirement of any hybrid storage facilities 20, continuous detecting interval depends on live load and service level agreement (ServiceLevelAgreement, SLA) definition or in service quality (QualityofService, QoS) demand.Generally, it can be several seconds (hard work load and superior performance demand) to several tens minutes (light work load and low performance demand).
Dsc data is searched unit 2024 and is used for detecting in an independent detection interval, the input/output request that a hard disk is accessed.Similarly, dsc data is searched unit 2024 and is detected write instruction and the reading instruction of hard disk 302 or 304 and relevant block thereof.Being different from input/output analytic unit 2022, dsc data is searched unit 2024 and is sequentially stored the second data of corresponding second block in dynamic replacement cache district 104, and wherein the second block is at least accessed secondary in this independent detection interval.Refer to Fig. 3.Access block identical in hard disk 302 in Fig. 2, if the twice that independent detection interval the is continuous detecting interval (time equal to tb2 to the tb1 time of ta4 to ta2, the time equal to tb1 to the t0 time of ta2 to t0), according to definition, in independent detection interval tb2 to tb1, C1, C2, C3, it is the second block with C4, and in independent detection interval tb1 to t0, A1, A2, A3, C1, C2, C3, be the second block with C4.Certainly, independent detection interval needs not be the twice of continuous detecting interval, can have any multiple proportion between independent detection interval and continuous detecting interval.
If to be write instruction requested and input/output analytic unit 2022 finds that repeat pattern cache district 102 is soon full for the first data, write hard disc module 204 can directly store the first data in a hard disk 302 or 304.As discussed above, the first data being stored in repeat pattern cache district 102 may be from write instruction or read instruction.To write caching, because the first data write in hard disk 302 or 304 the most at last, some first data cannot be stored into repeat pattern cache district 102 and be likely to the performance without compromising on hybrid storage facilities 20.This is because the first data being directly stored in hard disk 302 or 304 are likely to will not often be accessed in the future.Thus, if repeat pattern cache district 102 is written into rapidly, or the space, repeat pattern cache district 102 of one first percentage ratio occupied by data time, the first data occurred after a while will directly be stored in hard disk 302 or 304.In implementation, the first percentage ratio can be 90% or any be higher than 90% percentage ratio.Similarly, if to be write instruction requested and input/output analytic unit 2022 finds that dynamic replacement cache district 104 is written into rapidly for the second data, or one second space, dynamic replacement cache district 104 of percentage ratio when being occupied by data, write hard disc module 204 also can directly store the second data in hard disk 302 or 304.Second percentage ratio can also be 90% or any be higher than 90% percentage ratio.First and second percentage ratios are all that the Configuration Values set by system manager considers system performance, service level agreement or QoS requirement in the lump.90% or higher than 90% numerical value preferably.
In order to improve the efficiency in repeat pattern cache district 102 and dynamic replacement cache district 104, RAM module 206 applies district's Virtual table 2062.District's Virtual table 2062 is a mapping table, is used for keeping track the Block status of solid state hard disc 100 (repeat pattern cache district 102 and dynamic replacement cache district 104).If belonging to dynamic replacement cache district 104 at correspondence reflection block in solid state hard disc 100 of the logical blocks of district's Virtual table 2062, the logical blocks in district's Virtual table 2062 is denoted as dynamic replacement cache block.On the other hand, if belonging to repeat pattern cache district 102 at correspondence reflection block in solid state hard disc 100 of the logical blocks of district's Virtual table 2062, the logical blocks in district's Virtual table 2062 is denoted as repeat pattern cache block.When input/output analytic unit 2022 finds that in this dynamic replacement cache district, specific one second data 104 become the first data, it is not necessary to remove these data from dynamic replacement cache district 104 and be rewritten to repeat pattern cache district 102.For reflection to the dynamic replacement cache block of these specific second data in dynamic replacement cache district 104, they only can be directly denoted as repeat pattern cache block.In dynamic replacement cache district, specific second data 104 do not do any change.Because random access memory running speed is faster than solid state hard disc 100, repeat pattern cache district 102 and dynamic replacement cache district 104 must be modified by hybrid storage facilities 20 with slower speed, and RAM module 206 performs little change is definitely of value to hybrid storage facilities 20.Therefore the performance of hybrid storage facilities 20 can promote.
It should be noted that a kind of cache replacement algorithm adopted by dsc data search unit 2024 is for removing the second data that a part is stored in dynamic replacement cache district 104.This cache replacement algorithm is used for removing the second data of some less use in dynamic replacement cache district 104, in order to the second data of new demand can be cached in dynamic replacement cache district 104.Several cache replacement algorithm is had to use.Such as, minimum nearest use (LeastRecentlyUsed, LRU) algorithm, minimum more often use (LeastFrequentlyUsed, LFU) algorithm, or self adaptation replacement cache (AdaptiveReplacement cache, ARC) algorithm, different operational versions need different algorithms.It is from those cache replacement algorithm that a kind of dsc data by dynamic replacement cache district 104 searches the algorithm that unit 2024 uses, and according to selected by the occupation mode that storage facilities is different, or stops being applied to read or write instruction.
According to different occupation modes, each algorithm can represent than other algorithm better performance in some behaviour in service.Have the problem that the situation that request occurs do not known by hybrid storage facilities 20, working environment there is also too many variable.But it is useless that particular phenomenon can demonstrate most recently used cache replacement algorithm.Such as, if hybrid storage facilities 20 (or dynamic replacement cache district 104) runs into routinely, input and output operation number per second high and cache hit rate is low, this would imply that the cache replacement algorithm of this application is an invalid selection.Thus, dsc data searches unit 2024 will stop storing the second data from reading instruction request, the substitute is another kind of cache replacement algorithm.That is read cache to temporarily cease and write caching is unaffected.
According to the spirit of the present invention, before some measurable (or repetition) situation occurs, input/output analytic unit 2022 can adjust the caching data in dynamic replacement cache district 104.Asking for an interview Fig. 4, this figure is shown in the live load distribution situation of the hybrid storage facilities 20 of ordinary day.A live load from applied host machine 401 and 402 shows that a kind of live load sharply increased is present approximately in 9:00, and this is probably is reached the standard grade work by employee, will ask for what data caused to server 200.A group the first block D is detected in this stage, and every day all can by repeated access on ordinary days.Certainly, if this group the first block D is stored in hard disk 302, when people require its access is not made any changes, input and output operation number per second will uprise, and postpone also can occur.Therefore, when input/output analytic unit 2022 finds that this group the first block D repeats to be accessed, before in next time, this first block is accessed, the first corresponding data will be stored in repeat pattern cache district 102.Such as, the first data will be stored into repeat pattern cache district 102 at 8:30 on ordinary days.That is based on input/output analytic unit 2022, data pre-fetching goes out in repeat pattern cache district 102, rather than waits that the cache replacement algorithm of dsc data search unit 2024 dynamically updates dynamic replacement cache district 104.
On the contrary, if input/output analytic unit 2022 finds that a group the first block " or not is by repeated access (they must previously some moment is repeatedly accessed), after in next time, this first block is accessed, the first corresponding data will be removed by repeat pattern cache district 102.Referring to Fig. 5, this figure shows another live load of hybrid storage facilities 20.It is fairly heavy for will be apparent from this live load at 9:00 to 21:00.During this period, a group the first block E is observed and is not accessed.Therefore, this group the first block E will be removed.Such as, he will in the next day 8:50 divide after remove.Remove storage data in advance and be provided that other live load more cache space in dynamic replacement cache district 104.
The above-mentioned example of the present embodiment is based on the data of history detection.If live load can be predicted or gives, the present invention is also suitable.In other words, when new the first data of a group and/or the second data are predicted accesses at following particular point in time, input/output analytic unit 2022 can remove the space in a part of repeat pattern cache district 102 and/or the space in a part of dynamic replacement cache district 104.Thus, the first new data and/or the second data can store at the time point of this prediction.Any so the suitable method of service, algorithm or module can be provided to use.A kind of storage facilities discharge model provided in U.S. Patent Application No. 14/290,533 by identical inventor is provided, constructed collective reference can be obtained by this application case.The size removing space in repeat pattern cache district 102 and/or dynamic replacement cache district 104 should depend on live load.
It should be noted that in the present embodiment, the form of cache management module 202 and write hard disc module 204 is the adapter in server 200 or a connector.It is true that they also can be the software being installed in server 200 and realize.
Although the present invention is disclosed above with embodiment; so it is not limited to the present invention; any art has usually intellectual; without departing from the spirit and scope of the present invention; when doing a little change and retouching, therefore protection scope of the present invention ought be defined depending on appended claims scope and is as the criterion.
Claims (14)
1. the solid state hard disc fast taking system for hybrid storage facilities, it is characterised in that it comprises a solid state hard disc and cache management module, wherein,
This solid state hard disc is in order to store caching data, and this solid state hard disc is divided into a repeat pattern cache district and a dynamic replacement cache district；And
This cache management module comprises an input/output analytic unit and a dsc data searches unit, wherein,
This input/output analytic unit is for detecting in multiple continuous detecting intervals, the input/output request that the block of one hard disk is accessed, and sequentially store the first data of corresponding first block to this repeat pattern cache district, in each continuous detecting interval, this first block is at least by repeated access secondary；And
This dsc data searches unit for detecting in an independent detection interval, the input/output request that one hard disk is accessed, and sequentially store the second data of corresponding second block to this dynamic replacement cache district, in this independent detection interval, this second block is at least accessed secondary.
2. solid state hard disc fast taking system as claimed in claim 1, comprise a write hard disc module further, if to be write instruction requested and when this input/output analytic unit finds that the space in the repeat pattern cache district of one first percentage ratio is occupied by data for these first data, in order to directly to store these first data in a hard disk.
3. solid state hard disc fast taking system as claimed in claim 2, wherein this first percentage ratio be 90% or any be higher than 90% percentage ratio.
4. solid state hard disc fast taking system as claimed in claim 2, if wherein these second data are that write instruction is requested and this input/output analytic unit finds that when the space in dynamic replacement cache district of one second percentage ratio is occupied by data, this write hard disc module directly stores these second data further in a hard disk.
5. solid state hard disc fast taking system as claimed in claim 4, wherein this second percentage ratio be 90% or any be higher than 90% percentage ratio.
6. solid state hard disc fast taking system as claimed in claim 2, wherein this cache management module is the adapter in server or a connector with write hard disc module, or is installed on the software in server.
7. solid state hard disc fast taking system as claimed in claim 1, comprise a RAM module further, this RAM module includes district's Virtual table, and wherein this district's Virtual table is a mapping table, for keeping track the Block status of this first block and the second block；If the correspondence reflection block that the logical blocks in district's Virtual table is in this solid state hard disc belongs to dynamic replacement cache district, the logical blocks in Ze district Virtual table is denoted as dynamic replacement cache block；If the correspondence reflection block that the logical blocks in district's Virtual table is in this solid state hard disc belongs to repeat pattern cache district, the logical blocks in Ze district Virtual table is denoted as repeat pattern cache block；When this input/output analytic unit finds that specific second data become first data in this dynamic replacement cache district, the dynamic replacement cache block of specific second data in this dynamic replacement cache district of then videoing is denoted as repeat pattern cache block, and this dynamic replacement cache district Central Plains data content does not do any change.
8. solid state hard disc fast taking system as claimed in claim 1, wherein when this input/output analytic unit finds a group the first block by repeated access, before in next time, this first block is accessed, the first corresponding data will be stored in this repeat pattern cache district.
9. solid state hard disc fast taking system as claimed in claim 1, wherein when this input/output analytic unit finds that a group the first block is not repeated access, after in next time, this first block is accessed, the first corresponding data will remove from this repeat pattern cache district.
10. solid state hard disc fast taking system as claimed in claim 1, wherein when new the first data of a group and/or the second data predicting are taken when following particular point in time will deposit, then this input/output analytic unit removes a part of space in this repeat pattern cache district and/or a part of space in this dynamic replacement cache district.
11. solid state hard disc fast taking system as claimed in claim 1, wherein a cache replacement algorithm is adopted by this dsc data search unit, to remove a part the second data in dynamic replacement cache district.
12. solid state hard disc fast taking system as claimed in claim 11, wherein this cache replacement algorithm is that minimum nearest use algorithm, minimum more often use algorithm or self adaptation replace cache algorithm.
13. solid state hard disc fast taking system as claimed in claim 12, the algorithm wherein used by the dsc data search unit in this dynamic replacement cache district is for from those cache replacement algorithm, according to selected by the occupation mode that storage facilities is different, or stop being applied to read or write instruction.
14. solid state hard disc fast taking system as claimed in claim 11, wherein when hybrid storage facilities run into routinely, input and output operation number per second high low with cache hit rate time, this dsc data is searched unit and is stopped storing the second data by reading instruction request.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410834245.4A CN105808150B (en) | 2014-12-29 | 2014-12-29 | Solid state disk fast taking system for hybrid storage facilities |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410834245.4A CN105808150B (en) | 2014-12-29 | 2014-12-29 | Solid state disk fast taking system for hybrid storage facilities |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105808150A true CN105808150A (en) | 2016-07-27 |
| CN105808150B CN105808150B (en) | 2018-09-07 |
Family
ID=56980498
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410834245.4A Active CN105808150B (en) | 2014-12-29 | 2014-12-29 | Solid state disk fast taking system for hybrid storage facilities |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105808150B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10496541B2 (en) | 2017-11-29 | 2019-12-03 | Samsung Electronics Co., Ltd. | Dynamic cache partition manager in heterogeneous virtualization cloud cache environment |
| US10795583B2 (en) | 2017-07-19 | 2020-10-06 | Samsung Electronics Co., Ltd. | Automatic data placement manager in multi-tier all-flash datacenter |
| US11334274B2 (en) | 2018-02-09 | 2022-05-17 | Seagate Technology Llc | Offloaded data migration between storage devices |
| CN117041256A (en) * | 2023-10-08 | 2023-11-10 | 深圳市连用科技有限公司 | Network data transmission and storage method and system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030140207A1 (en) * | 2002-01-21 | 2003-07-24 | Hitachi. Ltd. | Hierarchical storage apparatus and control apparatus thereof |
| CN101458613A (en) * | 2008-12-31 | 2009-06-17 | 成都市华为赛门铁克科技有限公司 | Method for implementing mixed hierarchical array, the hierarchical array and storage system |
| US20110173395A1 (en) * | 2010-01-12 | 2011-07-14 | International Business Machines Corporation | Temperature-aware buffered caching for solid state storage |
| CN102129472A (en) * | 2011-04-14 | 2011-07-20 | 上海红神信息技术有限公司 | Construction method for high-efficiency hybrid storage structure of semantic-orient search engine |
| CN103186350A (en) * | 2011-12-31 | 2013-07-03 | 北京快网科技有限公司 | Hybrid storage system and hot spot data block migration method |
| CN103399713A (en) * | 2013-08-02 | 2013-11-20 | 浙江大学 | Data buffering method for balancing multistage memory property and solid-state disk service life |
-
2014
- 2014-12-29 CN CN201410834245.4A patent/CN105808150B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030140207A1 (en) * | 2002-01-21 | 2003-07-24 | Hitachi. Ltd. | Hierarchical storage apparatus and control apparatus thereof |
| CN101458613A (en) * | 2008-12-31 | 2009-06-17 | 成都市华为赛门铁克科技有限公司 | Method for implementing mixed hierarchical array, the hierarchical array and storage system |
| US20110173395A1 (en) * | 2010-01-12 | 2011-07-14 | International Business Machines Corporation | Temperature-aware buffered caching for solid state storage |
| CN102129472A (en) * | 2011-04-14 | 2011-07-20 | 上海红神信息技术有限公司 | Construction method for high-efficiency hybrid storage structure of semantic-orient search engine |
| CN103186350A (en) * | 2011-12-31 | 2013-07-03 | 北京快网科技有限公司 | Hybrid storage system and hot spot data block migration method |
| CN103399713A (en) * | 2013-08-02 | 2013-11-20 | 浙江大学 | Data buffering method for balancing multistage memory property and solid-state disk service life |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10795583B2 (en) | 2017-07-19 | 2020-10-06 | Samsung Electronics Co., Ltd. | Automatic data placement manager in multi-tier all-flash datacenter |
| US10496541B2 (en) | 2017-11-29 | 2019-12-03 | Samsung Electronics Co., Ltd. | Dynamic cache partition manager in heterogeneous virtualization cloud cache environment |
| US11334274B2 (en) | 2018-02-09 | 2022-05-17 | Seagate Technology Llc | Offloaded data migration between storage devices |
| US11893258B2 (en) | 2018-02-09 | 2024-02-06 | Seagate Technology Llc | Offloaded data migration between storage devices |
| CN117041256A (en) * | 2023-10-08 | 2023-11-10 | 深圳市连用科技有限公司 | Network data transmission and storage method and system |
| CN117041256B (en) * | 2023-10-08 | 2024-02-02 | 深圳市连用科技有限公司 | Network data transmission and storage method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105808150B (en) | 2018-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107943867B (en) | High-performance hierarchical storage system supporting heterogeneous storage | |
| US8909887B1 (en) | Selective defragmentation based on IO hot spots | |
| CN109154917B (en) | Storage system and solid state disk | |
| CN104317742B (en) | Automatic thin-provisioning method for optimizing space management | |
| US8782369B2 (en) | Apparatus to manage efficient data migration between tiers | |
| CN107506314B (en) | Method and apparatus for managing storage system | |
| US20160179683A1 (en) | Ssd caching system for hybrid storage | |
| US20140304452A1 (en) | Method for increasing storage media performance | |
| EP3089039B1 (en) | Cache management method and device | |
| CN107077300A (en) | For balancing segmentation removing and the rate-matched technology of I/O workloads | |
| CN106687910A (en) | Optimized segment cleaning technique | |
| WO2013093994A1 (en) | Storage system, data rebalancing program and data rebalancing method | |
| CN107018172A (en) | System and method for the adaptive partition in distributed caching memory | |
| Lin et al. | Hot random off-loading: A hybrid storage system with dynamic data migration | |
| KR20120120186A (en) | Efficient Use of Hybrid Media in Cache Architectures | |
| US9823875B2 (en) | Transparent hybrid data storage | |
| CN104503703B (en) | The treating method and apparatus of caching | |
| JP2010198056A (en) | Allocation control program and allocation control device | |
| CN107615254A (en) | The cache memory architectures and algorithm of blending objects storage device | |
| US10198180B2 (en) | Method and apparatus for managing storage device | |
| CN105808150A (en) | Solid state disk caching system for hybrid storage device | |
| JP2003280950A (en) | File management system | |
| CN107422989A (en) | A kind of more copy read methods of Server SAN systems and storage architecture | |
| CN105095495A (en) | Distributed file system cache management method and system | |
| CN106547477A (en) | For the method and apparatus for reducing buffer memory device online |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 1226833 Country of ref document: HK |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |