CN105573675B - User based on training mechanism in cache management device is accustomed to acquisition methods and device - Google Patents
User based on training mechanism in cache management device is accustomed to acquisition methods and device Download PDFInfo
- Publication number
- CN105573675B CN105573675B CN201510941425.7A CN201510941425A CN105573675B CN 105573675 B CN105573675 B CN 105573675B CN 201510941425 A CN201510941425 A CN 201510941425A CN 105573675 B CN105573675 B CN 105573675B
- Authority
- CN
- China
- Prior art keywords
- page address
- user
- hit
- data page
- count
- 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.)
- Active
Links
- 230000007246 mechanism Effects 0.000 title claims abstract description 20
- 238000000034 method Methods 0.000 title claims abstract description 19
- 239000007787 solid Substances 0.000 claims abstract description 28
- 230000008859 change Effects 0.000 claims description 8
- 238000007726 management method Methods 0.000 description 25
- 230000006399 behavior Effects 0.000 description 11
- 238000003860 storage Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 4
- 230000003321 amplification Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000003199 nucleic acid amplification method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
- G06F3/0604—Improving or facilitating administration, e.g. storage management
- G06F3/0605—Improving or facilitating administration, e.g. storage management by facilitating the interaction with a user or administrator
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0655—Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
- G06F3/0659—Command handling arrangements, e.g. command buffers, queues, command scheduling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
- G06F3/0671—In-line storage system
- G06F3/0673—Single storage device
- G06F3/0679—Non-volatile semiconductor memory device, e.g. flash memory, one time programmable memory [OTP]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Memory System Of A Hierarchy Structure (AREA)
Abstract
The invention discloses the users based on training mechanism in cache management device to be accustomed to acquisition methods, includes the following steps：S1, the operation note for obtaining user's operation solid state disk, and the operating habit of user is obtained according to operation note；S2, judge whether the operating habit is new operating habit, if so, old operating habit is replaced with new operating habit, if it is not, being then continuing with old operating habit；S3, the data page address that user will access is predicted according to operating habit；S4, will be in data page address caching to cache management device.Correspondingly, the invention also discloses the users based on training mechanism in cache management device to be accustomed to acquisition device.It is trained in the behavior that the present invention operates solid state disk repeatedly by user under new custom, it makes a prediction to be newly accustomed to the operation that lower user will carry out, to guide the cache management device of solid state disk to be buffered to the data page address that will be accessed in caching in advance, cache access hit rate is greatly improved.
Description
Technical field
The present invention relates in solid-state hard disk management method and device more particularly to cache management device based on training mechanism
User is accustomed to acquisition methods and device.
Background technology
Under cloud computing environment, the challenge brought is asked with user in order to cope with mass data, traditional database is solved and faces
Large-scale data access bottleneck, distributed caching technology is introduced into, and provides high-performance, High Availabitity, telescopic number to the user
According to buffer service.Enterprise uses storage medium of the high-speed internal memory as data object, data to be stored in the form of key/value.
Solid state disk (Solid State Drive, SSD) is manufactured hard disk with solid-state electronic storage chip array,
It is made of control unit and storage unit (including FLASH chip and dram chip).Solid state disk the specification of interface and definition,
It is identical with common hard disc in function and application method, it is also completely consistent with common hard disc in product design and size.Gu
State hard disk has the characteristics that the fast reading and writing that traditional mechanical hard disk do not have, light weight, low energy consumption and small.But its price
Still costly, capacity is relatively low, once the durability (service life) of hardware damage, the more difficult recovery of data, and solid state disk is opposite
It is shorter.
Since the erasable number of solid state disk flash memory is limited, the MLC flash chip service life of 34nm is approximately 5000 P/E, and
The service life of 25nm is approximately 3000 P/E.One of the optimizing index of SSD firmware algorithms is to provide less unnecessary writing.
The performance of caching, which is also embodied in, replaces on efficiency of algorithm, and the purpose that algorithm is replaced in optimization is to improve the hit rate of caching
With bit hit rate.Influence efficiency of algorithm because be known as cache size, data cached size, it is data cached be not hit by expense, when
Between locality and long tail effect etc..Business system comes usually using FIFO replacement strategy in SSD cache servers at present
Content is updated, however, carrying out analysis by the access situation to cache contents finds that FIFO policy can reduce access hit
Rate causes cache server that more request background data center storages, increased bandwidth demand is needed to increase in back-end data
The I/O pressure of the heart.Combine the replacement policy of more optimizing factors that can effectively improve cache hit rate and ratio using LRU etc.
Special hit rate.
But defect born SSD --- amplification is write, determines that FIFO replacement strategy can will write amplification and be preferably minimized,
Other replacement policies such as LRU, LFU can all cause it is serious write amplification, shorten the service life of SSD.Enterprise is from cost
Upper consideration, FIFO can extend the service life of SSD, reduce the purchase of SSD, existing slow although FIFO can be such that hit rate reduces
Deposit system still uses FIFO replacement strategy.
Invention content
It captures and analyzes by user being used for a long time the custom of solid state disk in solid state disk cache management device
Imminent operation after the completion of current user operation can be predicted, to guide the cache management device of solid state disk will i.e.
Data cached by hit is buffered in caching in advance, helps to improve cache access hit rate.In this case, can
User is perceived using the custom of solid state disk, is obtained, especially when user is changed using the custom of solid state disk
When, the change of user's custom can be perceived, the new custom until adapting to user, the operation that will be carried out for the lower user of new custom
Making a prediction just becomes extremely important.
In order to solve the above technical problem, the present invention provides user's customs based on training mechanism in cache management device
Acquisition methods include the following steps：
S1, the operation note for obtaining user's operation solid state disk, and the operation habit of user is obtained according to the operation note
It is used；
S2, judge whether the operating habit is new operating habit, practised if so, replacing old operation with new operating habit
It is used, if it is not, being then continuing with old operating habit；
S3, the data page address that user will access is predicted according to the operating habit；
S4, will be in the data page address caching to cache management device.
Specifically, the step S3 is specially：
The sequence of the data page address accessed according to active user provides optimal, suboptimum and worst priority is recommended, in advance
Survey the data page address that user will access.
Further, the step S4 includes the following steps：
S41, will be in the data page address caching to cache management device recommended according to priority；
S42, the data page address for really accessing the data page address that will be accessed recommended according to priority and user
It is compared, obtains hit-count, and according to the hit-count,
The corresponding page address of hit-count highest is labeled as to next data page address that will be accessed of optimal recommendation,
The high corresponding page address of hit-count time is labeled as next data page address that will be accessed that suboptimum is recommended,
The worst corresponding page address of hit-count is labeled as to next data page address that will be accessed of worst recommendation；
S43, according to hit-count, readjust optimal, suboptimum and worst priority recommended, and execute step again
S3。
Specifically, the step S42 is specifically included：
S421, the data page address that the data page address that will be accessed of optimal recommendation is really accessed with user is compared
It is right,
If comparison result is consistent, the hit-count of optimal recommendation adds one,
If comparison result is inconsistent, enter next step；
S422, the data page address that the data page address that will be accessed that suboptimum is recommended really is accessed with user is compared
It is right,
If comparison result is consistent, the hit-count that suboptimum is recommended adds one,
If comparison result is inconsistent, enter next step；
S423, the data page address that the data page address that will be accessed of worst recommendation is really accessed with user is compared
It is right,
If comparison result is consistent, the hit-count of worst recommendation adds one.
Further, judge whether the operating habit is new operating habit in the step S2, include the following steps：
It will be compared according to the hit-count of optimal, suboptimum and the recommendation of worst priority, wherein optimal hit time
Number label is that the hit-count of suboptimum is labeled as b, and worst hit-count is labeled as c,
If a≤b or a≤c, judge that the operating habit of user changes, current operation custom is new operating habit；
If a>b>C then judges that user operation habits are constant.
Correspondingly, the present invention also provides the users based on training mechanism in a kind of cache management device to be accustomed to obtaining dress
It sets, including：
Acquisition module, the operation note for obtaining user's operation solid state disk, and used according to the operation note
The operating habit at family；
Judgment module, for judging whether the operating habit is new operating habit, if so, being replaced with new operating habit
Old operating habit, if it is not, being then continuing with old operating habit；
Prediction module, the data page address for that will be accessed according to operating habit prediction user；
Cache module, being used for will be in the data page address caching to cache management device.
Further, the prediction module, the sequence for being specifically used for the data page address accessed according to active user provide
Optimal, suboptimum and worst priority are recommended, the data page address that prediction user will access.
Further, the cache module is specifically used for：
It will be in the data page address caching to cache management device recommended according to priority；
The data page address that the data page address that will be accessed recommended according to priority and user are really accessed carries out
It compares, obtains hit-count, and according to the hit-count,
The corresponding page address of hit-count highest is labeled as to next data page address that will be accessed of optimal recommendation,
The high corresponding page address of hit-count time is labeled as next data page address that will be accessed that suboptimum is recommended,
The worst corresponding page address of hit-count is labeled as to next data page address that will be accessed of worst recommendation；
According to hit-count, readjusts optimal, suboptimum and worst priority is recommended.
Further, the cache module further includes comparing module, and the comparing module is specifically used for：
The data page address that will be accessed of optimal recommendation is compared with the data page address that user really accesses,
If comparison result is consistent, the hit-count of optimal recommendation adds one,
If comparison result is inconsistent, by the data page address that will be accessed that suboptimum is recommended and the number that user really accesses
It is compared according to page address,
If comparison result is consistent, the hit-count that suboptimum is recommended adds one,
If comparison result is inconsistent, number that the data page address that will be accessed of worst recommendation and user are really accessed
It is compared according to page address,
If comparison result is consistent, the hit-count of worst recommendation adds one.
Further, the judgment module further includes comparison module：
The comparison module is used to be compared according to the hit-count of optimal, suboptimum and the recommendation of worst priority,
Wherein, optimal hit-count is labeled as a, and the hit-count of suboptimum is labeled as b, and worst hit-count is labeled as c,
If a≤b or a≤c, judgment module is used to judge that the operating habit of user to change, and current operation custom is new behaviour
Work is accustomed to；
If a>b>C, then judgment module is for judging that user operation habits are constant.
User based on training mechanism in the cache management device of the present invention is accustomed to acquisition methods and device, and have has as follows
Beneficial effect：
1, the method for the present invention changes priority according to hit-count and recommends next data page address that will be accessed at any time,
This records the process changed at any time and can be considered as in the behavior for operating solid state disk repeatedly for user to recommending to enter at any time
The training of cache page address behavior, user's custom is more stable, then it is more accurate into cache page address behavior to recommend.
2, apparatus of the present invention can perceive the change of user's custom, and can repeatedly be operated under new custom by user
It is trained in the behavior of solid state disk, the new custom until adapting to user, the operation that will be carried out for the lower user of new custom is made
Go out prediction, works as to guide the cache management device of solid state disk that the data page address that will be accessed is buffered to caching in advance
In, greatly improve cache access hit rate.
3, the method for the present invention and device realize cache replacement algorithm and automatically update evolution according to the change that user is accustomed to
Function, since the present invention can ensure that user automatically obtains most when forming new custom according to the new custom adjustable strategies of user
Good access speed experience with artificial without being adjusted.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is the flow chart of user's custom acquisition methods based on training mechanism in cache management device of the present invention；
Fig. 2 is that the user based on training mechanism in the cache management device in the embodiment of the present invention is accustomed to acquisition system frame
Figure；
Fig. 3 is the structure diagram of user's custom acquisition device based on training mechanism in cache management device of the present invention.
In figure：The data page address that 1- user really accesses, 2- hit address comparators, 3- hit-count counters, 4-
Page address stream logger, next segment address register for entering caching that 5- highest priorities are recommended, 6- suboptimum priority push away
The next segment address register for entering caching recommended, next sector address deposit for entering caching that the worst priority of 7- is recommended
Device, the corresponding hit-count register of 8- highest priorities, the corresponding hit-count register of 9- suboptimum priority, 10- are worst
The corresponding hit-count register of priority, 11- hit-counts sequence generator.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art obtained under the premise of not making creative work it is all its
His embodiment, shall fall within the protection scope of the present invention.
The present invention provides the users based on training mechanism in cache management device to be accustomed to acquisition methods, including following step
Suddenly：
S1, the operation note for obtaining user's operation solid state disk, and the operation habit of user is obtained according to the operation note
It is used；
S2, judge whether the operating habit is new operating habit, practised if so, replacing old operation with new operating habit
It is used, if it is not, being then continuing with old operating habit.
S3, the data page address that user will access is predicted according to the operating habit；
S4, will be in the data page address caching to cache management device.
Specifically, the step S3 is specially：
The sequence of the data page address accessed according to active user provides optimal, suboptimum and worst priority is recommended, in advance
Survey the data page address that user will access.
Wherein, the step S4 includes the following steps：
S41, will be in the data page address caching to cache management device recommended according to priority；
S42, the data page address for really accessing the data page address that will be accessed recommended according to priority and user
It is compared, obtains hit-count, and according to the hit-count,
The corresponding page address of hit-count highest is labeled as to next data page address that will be accessed of optimal recommendation,
The high corresponding page address of hit-count time is labeled as next data page address that will be accessed that suboptimum is recommended,
The worst corresponding page address of hit-count is labeled as to next data page address that will be accessed of worst recommendation；
S43, according to hit-count, readjust optimal, suboptimum and worst priority recommended, and execute step again
S3。
Wherein, the step S42 is specifically included：
S421, the data page address that the data page address that will be accessed of optimal recommendation is really accessed with user is compared
It is right,
If comparison result is consistent, the hit-count of optimal recommendation adds one,
If comparison result is inconsistent, enter next step；
S422, the data page address that the data page address that will be accessed that suboptimum is recommended really is accessed with user is compared
It is right,
If comparison result is consistent, the hit-count that suboptimum is recommended adds one,
If comparison result is inconsistent, enter next step；
S423, the data page address that the data page address that will be accessed of worst recommendation is really accessed with user is compared
It is right,
If comparison result is consistent, the hit-count of worst recommendation adds one.
Wherein, judge whether the operating habit is new operating habit in the step S2, include the following steps：
It will be compared according to the hit-count of optimal, suboptimum and the recommendation of worst priority, wherein optimal hit time
Number label is that the hit-count of suboptimum is labeled as b, and worst hit-count is labeled as c,
If a≤b or a≤c, judge that the operating habit of user changes, current operation custom is new operating habit；
If a>b>C then judges that user operation habits are constant.
Correspondingly, the present invention also provides the users based on training mechanism in a kind of cache management device to be accustomed to obtaining dress
It sets, including：
Acquisition module, the operation note for obtaining user's operation solid state disk, and used according to the operation note
The operating habit at family；
Judgment module, for judging whether the operating habit is new operating habit, if so, being replaced with new operating habit
Old operating habit, if it is not, being then continuing with old operating habit；
Prediction module, the data page address for that will be accessed according to operating habit prediction user；
Cache module, being used for will be in the data page address caching to cache management device.
Wherein, the prediction module, the sequence offer that is specifically used for the data page address accessed according to active user is optimal,
Suboptimum and worst priority are recommended, the data page address that prediction user will access.
Wherein, the cache module is specifically used for：
It will be in the data page address caching to cache management device recommended according to priority；
The data page address that the data page address that will be accessed recommended according to priority and user are really accessed carries out
It compares, obtains hit-count, and according to the hit-count,
The corresponding page address of hit-count highest is labeled as to next data page address that will be accessed of optimal recommendation,
The high corresponding page address of hit-count time is labeled as next data page address that will be accessed that suboptimum is recommended,
The worst corresponding page address of hit-count is labeled as to next data page address that will be accessed of worst recommendation；
According to hit-count, readjusts optimal, suboptimum and worst priority is recommended.
Wherein, the cache module further includes comparing module, and the comparing module is specifically used for：
The data page address that will be accessed of optimal recommendation is compared with the data page address that user really accesses,
If comparison result is consistent, the hit-count of optimal recommendation adds one,
If comparison result is inconsistent, by the data page address that will be accessed that suboptimum is recommended and the number that user really accesses
It is compared according to page address,
If comparison result is consistent, the hit-count that suboptimum is recommended adds one,
If comparison result is inconsistent, number that the data page address that will be accessed of worst recommendation and user are really accessed
It is compared according to page address,
If comparison result is consistent, the hit-count of worst recommendation adds one.
Wherein, the judgment module further includes comparison module：
The comparison module is used to be compared according to the hit-count of optimal, suboptimum and the recommendation of worst priority,
Wherein, optimal hit-count is labeled as a, and the hit-count of suboptimum is labeled as b, and worst hit-count is labeled as c,
If a≤b or a≤c, judgment module is used to judge that the operating habit of user to change, and current operation custom is new behaviour
Work is accustomed to；
If a>b>C, then judgment module is for judging that user operation habits are constant.
More specifically, when cache page address to be executed is replaced, user is accustomed to acquisition device and recommends highest priority
Next data page address that will be accessed be substituted into caching, wherein it is described it is next will access data page address storage
In segment address register, after caching, data page address that user is really accessed and highest priority recommend it is next into
Enter the segment address register 5 of caching compared to, i.e., is compared in hit address comparator 2, is generated if comparing unanimously
One pulse generates that primary plus 1 counts in the corresponding hit-count register of highest priority 8 to being hit-count counter 3；
If inconsistent, with suboptimum priority recommend it is next enter caching segment address register 6 compared with pair, i.e., in hit address
It is compared in comparator 2, a pulse is generated if comparing unanimously to being hit-count counter 3, in suboptimum priority
Corresponding hit-count register 9 generates primary plus 1 counting；If inconsistent, next entrance with the recommendation of worst priority
The segment address register 7 of caching is compared compared to Dui in hit address comparator 2, one is generated if comparing unanimously
A pulse generates that primary plus 1 counts in the corresponding hit-count register of worst priority 10 to being hit-count counter 3,
After this is compared three times, hit-count sorts generator 11 to the corresponding hit-count register 8 of highest priority, and suboptimum is excellent
How much the corresponding hit-count register 9 of first grade, the corresponding hit-count register 10 of worst priority, weigh according to hit-count
Next segment address register 5 for entering caching that new adjustment highest priority is recommended, next entrance that sub-priority is recommended
The segment address register 6 of caching, the sequence for next segment address register 7 for entering caching that worst priority is recommended, operation
When will enter buffer storage the case where be stored in successively according to sequencing in page address stream logger 4.So, total energy
Ensure the most preferential recommendation of hit-count, reaches and obtained in the behavior for operating solid state disk repeatedly under new custom by user
Training, the new custom until adapting to user, the operation that will be carried out for the lower user of new custom is made a prediction, to guide solid-state hard
The cache management device of disk by will hit it is data cached be buffered in caching in advance, improve cache access hit rate.
User based on training mechanism in the cache management device of the present invention is accustomed to acquisition methods and device, and have has as follows
Beneficial effect：
1, the method for the present invention changes priority according to hit-count and recommends next data page address that will be accessed at any time,
This records the process changed at any time and can be considered as in the behavior for operating solid state disk repeatedly for user to recommending to enter at any time
The training of cache page address behavior, user's custom is more stable, then it is more accurate into cache page address behavior to recommend.
2, apparatus of the present invention can perceive the change of user's custom, and can repeatedly be operated under new custom by user
It is trained in the behavior of solid state disk, the new custom until adapting to user, the operation that will be carried out for the lower user of new custom is made
Go out prediction, works as to guide the cache management device of solid state disk that the data page address that will be accessed is buffered to caching in advance
In, greatly improve cache access hit rate.
3, the method for the present invention and device realize cache replacement algorithm and automatically update evolution according to the change that user is accustomed to
Function, since the present invention can ensure that user automatically obtains most when forming new custom according to the new custom adjustable strategies of user
Good access speed experience with artificial without being adjusted.
The above is the preferred embodiment of the present invention, it is noted that for those skilled in the art
For, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also considered as
Protection scope of the present invention.
Claims (6)
1. the user based on training mechanism in cache management device is accustomed to acquisition methods, which is characterized in that include the following steps：
S1, the operation note for obtaining user's operation solid state disk, and the operating habit of user, the use are obtained according to the operation note
The operating habit at family refers to the custom that solid state disk is used for a long time in user；
S2, judge whether the operating habit is new operating habit, if so, old operating habit is replaced with new operating habit, if
It is not to be then continuing with old operating habit；
S3, the data page address that user will access is predicted according to the operating habit；The step S3 is specially：According to current
The sequence for the data page address that user accesses provides optimal, suboptimum and worst priority is recommended, what prediction user will access
Data page address；
S4, will be in the data page address caching to cache management device；
The step S4 includes the following steps：S41, by the data page address caching recommended according to priority to cache management device
In；
S42, the data page address for really accessing the data page address that will be accessed recommended according to priority and user carry out
It compares, obtains hit-count, and according to the hit-count, the corresponding page address of hit-count highest is labeled as optimal recommendation
Next data page address that will be accessed, the high corresponding page address of hit-count time is recommended labeled as suboptimum next
The data page address that will be accessed will access the worst corresponding page address of hit-count labeled as the next of worst recommendation
Data page address；
S43, according to hit-count, readjust optimal, suboptimum and worst priority recommended, and execute step S3 again.
2. the user based on training mechanism in cache management device according to claim 1 is accustomed to acquisition methods, feature
It is, step S42 is specifically included：
S421, the data page address that will be accessed of optimal recommendation is compared with the data page address that user really accesses,
If comparison result is consistent, the hit-count of optimal recommendation adds one, if comparison result is inconsistent, enters next step；
S422, the data page address that will be accessed that suboptimum is recommended is compared with the data page address that user really accesses,
If comparison result is consistent, the hit-count that suboptimum is recommended adds one, if comparison result is inconsistent, enters next step；
S423, the data page address that will be accessed of worst recommendation is compared with the data page address that user really accesses,
If comparison result is consistent, the hit-count of worst recommendation adds one.
3. the user based on training mechanism in cache management device according to claim 2 is accustomed to acquisition methods, feature
It is, judges whether the operating habit is new operating habit, is included the following steps in the step S2：It will be according to optimal, secondary
The hit-count that excellent and worst priority is recommended is compared, wherein optimal hit-count is labeled as a, the hit of suboptimum
Number is labeled as b, and worst hit-count is labeled as c, if a≤b or a≤c, judges that the operating habit of user changes, currently
Operating habit is new operating habit；
If a>b>C then judges that user operation habits are constant.
4. the user based on training mechanism in cache management device is accustomed to acquisition device, which is characterized in that including：Acquisition module,
Operation note for obtaining user's operation solid state disk, and the operating habit of user is obtained according to the operation note, it is described
The operating habit of user refers to the custom that solid state disk is used for a long time in user；
Judgment module, for judging whether the operating habit is new operating habit, if so, replacing old behaviour with new operating habit
Work is accustomed to, if it is not, being then continuing with old operating habit；
Prediction module, the data page address for that will be accessed according to operating habit prediction user；The prediction module, tool
The sequence for the data page address that body is used to be accessed according to active user provides optimal, suboptimum and worst priority is recommended, prediction
The data page address that user will access；
Cache module, being used for will be in the data page address caching to cache management device；The cache module is specifically used for：
It will be in the data page address caching to cache management device recommended according to priority；
The data page address that will be accessed recommended according to priority is compared with the data page address that user really accesses,
Hit-count is obtained, and according to the hit-count, by the corresponding page address of hit-count highest labeled as under optimal recommendation
One data page address that will be accessed, will labeled as the next of suboptimum recommendation by the high corresponding page address of hit-count time
The worst corresponding page address of hit-count is labeled as next number that will be accessed of worst recommendation by the data page address of access
According to page address；
According to hit-count, readjusts optimal, suboptimum and worst priority is recommended.
5. the user based on training mechanism in cache management device according to claim 4 is accustomed to acquisition device, feature
It is, the cache module further includes comparing module, and the comparing module is specifically used for：By the number that will be accessed of optimal recommendation
It is compared with the data page address that user really accesses according to page address, if comparison result is consistent, the hit time of optimal recommendation
Number plus one, if comparison result is inconsistent, by the data page address that will be accessed that suboptimum is recommended and the number that user really accesses
It is compared according to page address,
If comparison result is consistent, the hit-count that suboptimum is recommended adds one, if comparison result is inconsistent, what it is by worst recommendation is
The data page address of access is compared with the data page address that user really accesses,
If comparison result is consistent, the hit-count of worst recommendation adds one.
6. the user based on training mechanism in cache management device according to claim 5 is accustomed to acquisition device, feature
It is, the judgment module further includes comparison module：The comparison module is used for will be according to optimal, suboptimum and worst priority
The hit-count of recommendation is compared, wherein optimal hit-count is labeled as a, and the hit-count of suboptimum is labeled as b, worst
Hit-count be labeled as c, if a≤b or a≤c, judgment module be used for judge user operating habit change, current operation
Custom is new operating habit；
If a>b>C, then judgment module is for judging that user operation habits are constant.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510941425.7A CN105573675B (en) | 2015-12-16 | 2015-12-16 | User based on training mechanism in cache management device is accustomed to acquisition methods and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510941425.7A CN105573675B (en) | 2015-12-16 | 2015-12-16 | User based on training mechanism in cache management device is accustomed to acquisition methods and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105573675A CN105573675A (en) | 2016-05-11 |
| CN105573675B true CN105573675B (en) | 2018-10-23 |
Family
ID=55883869
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510941425.7A Active CN105573675B (en) | 2015-12-16 | 2015-12-16 | User based on training mechanism in cache management device is accustomed to acquisition methods and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105573675B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108572799B (en) * | 2018-04-12 | 2021-05-04 | 上海理工大学 | Data page migration method of heterogeneous memory system of bidirectional hash chain table |
| US11176556B2 (en) * | 2018-11-13 | 2021-11-16 | Visa International Service Association | Techniques for utilizing a predictive model to cache processing data |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103927203A (en) * | 2014-03-26 | 2014-07-16 | 上海新储集成电路有限公司 | Computer system and control method |
| CN104111900A (en) * | 2013-04-22 | 2014-10-22 | 中国移动通信集团公司 | Method and device for replacing data in cache |
| CN104915345A (en) * | 2014-03-11 | 2015-09-16 | 华为技术有限公司 | Method and terminal for recommending service information |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8990506B2 (en) * | 2009-12-16 | 2015-03-24 | Intel Corporation | Replacing cache lines in a cache memory based at least in part on cache coherency state information |
-
2015
- 2015-12-16 CN CN201510941425.7A patent/CN105573675B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104111900A (en) * | 2013-04-22 | 2014-10-22 | 中国移动通信集团公司 | Method and device for replacing data in cache |
| CN104915345A (en) * | 2014-03-11 | 2015-09-16 | 华为技术有限公司 | Method and terminal for recommending service information |
| CN103927203A (en) * | 2014-03-26 | 2014-07-16 | 上海新储集成电路有限公司 | Computer system and control method |
Non-Patent Citations (1)
| Title |
|---|
| 《用户访问特征驱动的中间件语义缓存替换策略》;陈宁江，林盘;《广西大学学报：自然科学版》;20101031;第35卷(第5期);第787-792页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105573675A (en) | 2016-05-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Eisenman et al. | Flashield: a hybrid key-value cache that controls flash write amplification | |
| US10430084B2 (en) | Multi-tiered memory with different metadata levels | |
| US11263149B2 (en) | Cache management of logical-physical translation metadata | |
| TWI684099B (en) | Profiling cache replacement | |
| CN104809076B (en) | Cache management method and device | |
| CN106708425A (en) | Distributed multimode storage management | |
| WO2017117734A1 (en) | Cache management method, cache controller and computer system | |
| US10997080B1 (en) | Method and system for address table cache management based on correlation metric of first logical address and second logical address, wherein the correlation metric is incremented and decremented based on receive order of the first logical address and the second logical address | |
| US10296466B2 (en) | Information processing device, method of controlling a cache memory, and storage medium | |
| CN104834607A (en) | Method for improving distributed cache hit rate and reducing solid state disk wear | |
| US9558123B2 (en) | Retrieval hash index | |
| Chai et al. | WEC: Improving durability of SSD cache drives by caching write-efficient data | |
| US10990531B2 (en) | Cloud-based frequency-based cache management | |
| US20200285582A1 (en) | Adaptive readahead cache manager based on detected active streams of read commands | |
| US20090132769A1 (en) | Statistical counting for memory hierarchy optimization | |
| CN110888600A (en) | A buffer management method for NAND flash memory | |
| Wang et al. | A novel buffer management scheme based on particle swarm optimization for SSD | |
| CN105512051A (en) | Self-learning type intelligent solid-state hard disk cache management method and device | |
| CN105573675B (en) | User based on training mechanism in cache management device is accustomed to acquisition methods and device | |
| US9851925B2 (en) | Data allocation control apparatus and data allocation control method | |
| CN109783006B (en) | Computing system and method of operating a computing system | |
| Yang et al. | CacheSack: Theory and experience of Google’s admission optimization for datacenter flash caches | |
| KR101940382B1 (en) | Prefetching method and apparatus for pages | |
| US20250156326A1 (en) | Managing a system level cache | |
| CN108664217A (en) | A kind of caching method and system reducing the shake of solid-state disc storaging system write performance |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |