CN105988720A - Data storage device and method - Google Patents
Data storage device and method Download PDFInfo
- Publication number
- CN105988720A CN105988720A CN201510067474.2A CN201510067474A CN105988720A CN 105988720 A CN105988720 A CN 105988720A CN 201510067474 A CN201510067474 A CN 201510067474A CN 105988720 A CN105988720 A CN 105988720A
- Authority
- CN
- China
- Prior art keywords
- data
- write
- layer
- data layer
- memory
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a data storage method. The data storage method comprises the following steps: after receiving a data writing request, writing data into a first data layer or a second data layer according to a data writing operation degree; and after receiving a data reading request, inquiring the second data layer or the first data layer according to the data reading operation degree and returning back an inquiring result. The invention further discloses a data storage device, comprising a processor, the first data layer and the second data layer.
Description
Technical field
The present invention relates to data storage technology, particularly relate to a kind of data storage device and method.
Background technology
Along with development and the application of database technology, the data volume of database purchase is from the eighties in 20th century
Million (M) bytes and gigabit (G) byte develop into present billion (T) byte and peta-(P) word
Joint.Meanwhile, the query demand of user also becomes increasingly complex, and relate to is not only inquiry or handles a pass
It is one or several record in table, and the data of ten million bar record in multiple tables is carried out data analysis
And informix, relational database system can not all meet this requirement.Existing big data process side
Case has the number of employing massive parallel processing (MPP, Massively Parallel Processing) framework
According to storehouse system Greenplum, the Hadoop etc. of employing distributed system architecture.Further, by a large number
X86PC adds the framework of local hard drive (HDD, Hard Disk Drive) storage composition large-scale calculations cluster
Having become as the main operation platform of the big data solution of great majority, substantial amounts of server local hard disk also becomes
The storage mode that big data age is main.
But, traditional data is all based on hard-disc storage, management and input and output (I/O), and hard disk
Development speed is the slowest.First piece of hard disk in the world is released in JIUYUE, 1956 from IBM Corporation
So far, the rotating speed of main flow enterprise-level hard disk is 15, and 000RPM (turning every point) has stopped many years, except
Outside capacity can also be continuously increased, other various aspects of performance cannot obtain more effectively raising always, it may be said that hard
The performance of dish limits the lifting of systematic entirety energy the most to a certain extent.Although enterprise-level hard disk is permissible
The data storage and access efficiency of entirety is promoted by hard disk array, but due to the data read/write of hard disk
Can cannot be greatly improved, data transfer latency is oversize, causes system data read/write performance demand and supply
Between there is bigger contradiction.Currently employed hard disk array technology is also difficult to meet the backgrounds such as big data, cloud computing
Under system application demand.
Fundamentally seeing, the performance of storage system depends on its memory device relied on.HDD remains mesh
Front topmost Large Copacity, low cost memory device.The research of various new technology such as perpendicular recording and application
The capacity making hard disk adds 100 in 30 years of past, 000 times, but owing to being limited by the mobile speed of magnetic head
Degree, its access delay only improve only 2 times.And further improve the rotation speed of hard disk can bring energy resource consumption and
Temperature problem, therefore the contradiction between capacity and the performance of hard disk is the most prominent.
Along with the development of big data age, the big data service of movement will have the biggest development.Local hard drive
Storage also will be applied on a large scale.But in most of the cases, locally stored effect based on server
Rate is the highest, and the especially I/O disposal ability of big capacity hard disk is more not enough.Although using hard disk array energy
Solve some storage efficiency problems to a certain extent, but there is cost, physical space management, energy consumption simultaneously
Etc. the shortcoming of aspect.
In recent years, the fast development of solid state hard disc (SSD, Solid State Disk/Drive) is for promoting storage
Systematic function provides new major opportunity.SSD sampling solid-state electronic storage chip array, has random
The advantages such as reading speed is fast, low in energy consumption, shock resistance is good.Although SSD there is also on the high side at present, capacity
Less, and there is restrictions such as writing front erasing, durability, complementarity good for SSD with HDD remains as
Design Large Copacity, high-performance, the mixing storage system of low cost provide brand-new opportunity.
But, in the mixing storage system of SSD and HDD, SSD writes front erasing and writes life-span restriction
Bigger to the write operation performance impact of system.Further, owing to attribute intrinsic for SSD limits, SSD there is also
(inplace updating) can not be updated on the spot, the problem such as front erasing, erasing life-span need to be write.These problems are led
Cause that existing mixing storage system exists write operation and face awkward predicament: for the write-in policy of data, if
Frequently write into SSD, then can affect the service life of SSD, if but frequently write into HDD, can affect again
The write performance of system.
Summary of the invention
For solving the technical problem of existing existence, embodiment of the present invention expectation provide a kind of data storage device and
Method, it is possible to increase data writing operation and the operating characteristics of reading data, and drop while guaranteeing data security
Low carrying cost.
The technical scheme of the embodiment of the present invention is achieved in that
A kind of date storage method that the embodiment of the present invention provides, including:
After receiving write data requests, write data into the first data Layer or the second data according to data writing operation degree
In Ceng；
After receiving read data request, inquire about the second data Layer or the first data Layer according to reading data operational readiness, and
Return Query Result.
In such scheme, described first data Layer includes the first nonvolatile memory, described second data Layer
Including the second nonvolatile memory.
In such scheme, described method farther includes, and uses frequency algorithm and prediction hit rate according to data
Algorithm instant analysis, determines write data operational readiness or described reading data operational readiness.
In such scheme, described in write data into the first data Layer or the second data Layer includes: use write-back
The first nonvolatile memory or the second nonvolatile memory is write data into described in pattern Write-Back
In.
In such scheme, described in write data into the first data Layer or the second data Layer includes: by described number
Data block is unit according to this, is mapped by directly mapping, full associative mapping or set associative and data block writes institute
State in the first nonvolatile memory or described second nonvolatile memory.
In such scheme, described method farther includes: before inquiry the second data Layer or the first data Layer,
According to reading data operational readiness inquiry volatile memory；The application memory space of described volatile memory is set
Accelerate memory space with read/write, and the read/write that described volatile memory is set to accelerate memory space maximum and
Little available threshold.
A kind of data storage device that the embodiment of the present invention provides, including: processor, the first data Layer and the
Two data Layers,
Described processor is used for: after receiving write data requests, write data into first according to data writing operation degree
In data Layer or the second data Layer；After receiving read data request, inquire about the second data according to reading data operational readiness
Layer or the first data Layer, and return Query Result；
Described first data Layer and the second data Layer are used for storing described data.
In such scheme, described first data Layer includes the first nonvolatile memory, described second data Layer
Including the second nonvolatile memory.
In such scheme, described processor is further used for using frequency algorithm and prediction hit rate according to data
Algorithm instant analysis determines write data operational readiness or described reading data operational readiness.
In such scheme, described device farther includes: volatile memory and described volatile memory
Physics mirror image；
Described processor is further used for before inquiry the second data Layer or the first data Layer, according to reading data
Operational readiness inquires about described volatile memory, and the read/write arranging described volatile memory accelerates memory space
Minimum and maximum available threshold.
In such scheme, described processor is further used for by described data in units of data block, by directly
Connect mapping, full associative mapping or set associative mapping and data block is write described first nonvolatile memory or institute
State in the second nonvolatile memory.
In such scheme, described first nonvolatile memory includes hard disk；Described second non-volatile memories
Device includes solid state hard disc and/or nonvolatile random access memory；Described volatile memory includes depositing at random
Access to memory.
The date storage method that the embodiment of the present invention provides, writes data into the first number according to data writing operation degree
According in layer or the second data Layer, and inquire about the first data Layer or the second data Layer according to reading data operational readiness, from
And achieve data writing operation and the shunting reading data manipulation, improve data writing operation and reading the most simultaneously
According to operating characteristics, and enhance the reliability of data storage.
The embodiment of the present invention provide data storage device, arrange the first nonvolatile memory, second non-easily
The property lost memorizer and volatile memory are respectively used to store data, so, no according to writing/reading operational readiness
Only considered various storage media types, data payload, storage system architecture, memory capacity, performance,
The impact of the many factors such as reliability, cost and energy consumption, and by cache hierarchy by SSD and/
Or NVRAM is as the caching of HDD, it is achieved that data writing operation and the shunting of reading data manipulation, not only
Carrying cost can be reduced, additionally it is possible to improve data writing operation simultaneously and read data manipulation performance.
The embodiment of the present invention is by arranging minimum and maximum available threshold for RAM disc (Ramdisk)
Value ensures minimum and maximum available quantity, thus ensure that application memory space is unaffected.
The embodiment of the present invention, by arranging the physics mirror image of described volatile memory, is avoided that because volatibility is deposited
Reservoir damages and causes accessing data.
It is right during performing data writing operation that the embodiment of the present invention passes through write-back (Write-Back) pattern
Ramdisk and HDD locks, thus ensure that the safety of data.
Accompanying drawing explanation
The schematic flow sheet of the date storage method that Fig. 1 provides for the embodiment of the present invention；
The cache hierarchy schematic diagram that Fig. 2 provides for the embodiment of the present invention；
The schematic flow sheet of the data writing operation that Fig. 3 provides for the embodiment of the present invention；
The data that Fig. 4 provides for the embodiment of the present invention map schematic diagram；
The composition structural representation of the data storage device that Fig. 5 provides for the embodiment of the present invention.
Detailed description of the invention
In various embodiments disclosed by the invention, describe a kind of date storage method, write number including reception
After request, write data in the first data Layer or the second data Layer according to data writing operation degree；Receive and read
After request of data, inquire about the second data Layer or the first data Layer according to reading data operational readiness, and return inquiry knot
Really；So, can realize reading data manipulation and the shunting of data writing operation, can improve simultaneously read data manipulation with
Data writing operation performance.
The present invention is described in further detail to combine accompanying drawing below by specific embodiment.
In one embodiment of the invention as shown in Figure 1, disclosed date storage method includes following step
Rapid:
Step S101: receive write data requests or read data request；For write data requests, perform step 102,
For read data request, perform step 103；
Step S102: write data in the first data Layer or the second data Layer according to data writing operation degree；
Wherein, above-mentioned first data Layer can include that the first nonvolatile memory, the second data Layer can wrap
Include the second nonvolatile memory；
Further, write data in the first data Layer or the second data Layer can be write data into first non-easily
In the property lost memorizer or the second nonvolatile memory；
Above-mentioned first nonvolatile memory or the second nonvolatile memory can be according to various storage mediums
Characteristic selects.Table 1 lists the Character Comparison of various storage medium:
Table 1
Consider the various storage media types listed by table 1, data payload, storage system architecture, deposit
The impact of the many factors such as storage capacity, performance, reliability, cost and energy consumption, above-mentioned first non-volatile
Property memorizer can include that HDD, described second nonvolatile memory can include SSD and/or non-volatile
Property random access storage device (NVRAM, Non-Volatile Random Access Memory).
In step s 102, it is also possible to use frequency algorithm and prediction hit rate algorithm i.e. time-division according to data
Analysis, determines write data operational readiness or described reading data operational readiness.
Wherein, data writing operation degree is the unit time interior number of times HotW performing data writing operation, during unit
Between can be hour, day or week；Reading data operational readiness is that interior execution of unit memory space reads the secondary of data manipulation
Number HotR.
Wherein, the formula of frequency algorithm is:
Parameter i is data block identifier, and its span is 1～k, and k is the number of the memory block of data storage device
Amount；Parameter t represents data collection cycle, and its value can be hour, day or week；Parameter Wblock represents
Each data block performs the number of times of data writing operation in data collection cycle, and its span is 0～1000,
And can be hot spot data by the value of the Wblock data block identifier more than 1000；
Wherein, it was predicted that the formula of hit rate algorithm is: P=Sn/N；
P is prediction hit rate；Parameter Sn represents historical heat data acquisition hot spot data set, its value
Scope is the most data block of number of times that top n performs data writing operation, i.e. Sn=takes top n RANK
(HotWi)；
Frequency algorithm and prediction hit rate algorithm instant analysis is used to determine write data operational readiness according to data
Computing formula be: M=α P+ β HotW；
Parameter M represents the data that needs cache, and parameter alpha, the span of β are 0～1, and meet alpha+beta=
1。
It is possible to further data writing operation degree to be met the frequent hot spot data write described the of M feature
Two data Layers.
Above-mentioned writing data into may include that employing Write-Back mould in the first data Layer or the second data Layer
Write data into described in formula in the first nonvolatile memory or the second nonvolatile memory.
Above-mentioned writing data in the first data Layer or the second data Layer can also include: by described data with number
It is unit according to block, is mapped by directly mapping, full associative mapping or set associative and data block is write described first
In nonvolatile memory or described second nonvolatile memory.
Step S103: inquire about the second data Layer or the first data Layer according to reading data operational readiness；
Wherein it is possible to use frequency algorithm and prediction hit rate algorithm instant analysis according to data, determine described
Read data operational readiness；It is possible to further arrange the frequency statistics meeting M feature according to reading data operational readiness
Strategy, first inquires about the second data Layer and inquires about the first data Layer again；It is further possible to according to reading data behaviour
The height of degree of work, first inquires about volatile memory, if not having result to return, then inquiry second is non-volatile deposits
Reservoir, if still not having result to return, then inquires about the first nonvolatile memory.
Above-mentioned volatile memory can arrange application memory space and read/write accelerates memory space, wherein, easily
The property lost memorizer can be random access memory (RAM, Random Access Memory), and read/write adds
Speed memory space can be further arranged to Ramdisk, and it is minimum and maximum to be that Ramdisk is arranged
Available threshold is to ensure minimum and maximum available quantity such that it is able to ensure that application memory space is unaffected.
And, it is also possible to the physics mirror image of described volatile memory is set further such that it is able to avoid because of
Volatile memory is damaged and is caused accessing data.
Step S104: return Query Result.
In one embodiment of the invention as shown in Figure 2, it is provided that cache hierarchy, by logical data
Data A-E of layer are respectively stored in the first data Layer and the second data Layer, the logical address of Logical Data Layer with
First data Layer physical address one_to_one corresponding.
Wherein, the first data Layer can be made up of HDD equipment, the second data Layer can by SSD and/or
NVRAM device is constituted.SSD and/or NVRAM is as the caching of HDD, after receiving read data request,
First in the second data Layer, inquire about these data the most buffered, if the most buffered, then access SSD and/
Or NVRAM device；Otherwise, HDD equipment is accessed.The cache hierarchy that the embodiment of the present invention provides,
By introducing second data Layer bigger than random access memory capacity, it is possible to caching more data, improve
Read data manipulation performance.
Further, receive after write data requests, can first by the SSD of data pre-write the second data Layer and/or
In NVRAM device, batch write HDD equipment the most again.The caching layering that the embodiment of the present invention provides
Structure, by introducing second data Layer more higher than HDD equipment random writing performance, it is possible to increase write number
According to operating characteristics.
In one embodiment of the invention, cache hierarchy can use flash cache further
Flashcache structure.
The cache hierarchy that the embodiment of the present invention provides, has considered various storage media types, data
The many factors such as load, storage system architecture, memory capacity, performance, reliability, cost and energy consumption
Impact, by using SSD and/or NVRAM as the caching of HDD, it is achieved that data writing operation and reading
The shunting of data manipulation, improves data writing operation the most simultaneously and reads data manipulation performance, and enhancing
The reliability of data storage.
In one embodiment of the invention as shown in Figure 3, data writing operation uses Write-Back pattern,
Comprise the following steps:
Step S301: if data to be write have been stored in volatile memory and the first non-volatile memories
In device；
Here, described storage can be to have stored in Ramdisk and HDD, then, the most right
Ramdisk and HDD locks (Lock)；
Step S302: write data in the second nonvolatile memory；
Here, described second nonvolatile memory can be in SSD and/or NVRAM；
Step S303: the corresponding data in the second nonvolatile memory is write in Ramdisk, release
Lock on RAMDISK；
Step S304: the corresponding data in the second nonvolatile memory is write in HDD, release HDD
On Lock.
In one embodiment of the invention as shown in Figure 4, data are unit storage with data block (block)
In the storage device.Below in conjunction with the Fig. 4 logical address to the data storage device that the embodiment of the present invention provides
It is described in further detail with physical address map.
For writing/reading operation requests, require consideration for how the logical address of data is mapped to corresponding thing
Reason address, i.e. mapping ruler.Wherein, mapping ruler includes mapping granule, i.e. storage device is capable of identify that
The most basic unit of data block, can include one or more data block.
It is the mapping that mapping granule is the thinnest that block level maps, can be using the data block of each 4KB as base
This map unit.All writing/reading operation requests are according to start physical address, such as sector number dbn, and press
According to modulo operation, Hash operation scheduling algorithm, the logical address of data is mapped to the physics of data storage device
Address.
Block level maps can use cache hierarchy, by directly mapping, full associative mapping or group phase
Connection maps and data block is mapped in the corresponding physical address of data storage device.
Embodiment shown in Fig. 4 uses set associative to map, and wherein, first passes through equation below and calculates target group:
Target group=(start sector number/(data block size * group size)) mod (quantity of group)
Such as, after being defined as target group 3, in target group 3 traversal search idle or be labeled as invalid
(invalid) sector, such as: first free time or be labeled as the sector of invalid such as figure in target group 3
Shown in 4 bends are filled.In Fig. 4, the dark sector table filled is shown as busy or is labeled as effectively (valid),
If not finding respective free in target group or being labeled as the sector of invalid, then according to minimum
Algorithm (Least Recently Used, LRU) is used to replace the respective sectors in target group.
In one embodiment of the invention as shown in Figure 5, described data storage device includes: processor,
First data Layer and the second data Layer.
Described processor is used for: after receiving write data requests, write data into first according to data writing operation degree
In data Layer or the second data Layer；After receiving read data request, inquire about the second data according to reading data operational readiness
Layer or the first data Layer, and return Query Result；
Described first data Layer and the second data Layer are used for storing described data.
Above-mentioned first data Layer can include that the first nonvolatile memory, the second data Layer can include second
Nonvolatile memory.
Above-mentioned processor can be further used for using frequency algorithm and prediction hit rate algorithm instant according to data
Analyze, determine write data operational readiness or described reading data operational readiness.
Said apparatus may further include volatile memory, is used for storing described data；Above-mentioned processor
Can be further used for before inquiry the second data Layer or the first data Layer, according to reading the inquiry of data operational readiness
Described volatile memory.
Above-mentioned data storage device may further include the physics mirror image of described volatile memory；Described place
Reason device is further used for arranging the minimum and maximum available threshold of described volatile memory.
Above-mentioned processor can be further used for by described data in units of data block, by directly mapping,
Full associative mapping or set associative map and data block write described first nonvolatile memory or described second non-
In volatile memory.
The above, only presently preferred embodiments of the present invention, it is not intended to limit the protection model of the present invention
Enclose.
Claims (12)
1. a date storage method, it is characterised in that the method includes:
After receiving write data requests, write data into the first data Layer or the second data according to data writing operation degree
In Ceng；
After receiving read data request, inquire about the second data Layer or the first data Layer according to reading data operational readiness, and
Return Query Result.
Method the most according to claim 1, it is characterised in that described first data Layer includes that first is non-
Volatile memory, described second data Layer includes the second nonvolatile memory.
Method the most according to claim 1, it is characterised in that described method farther includes, according to
Data use frequency algorithm and prediction hit rate algorithm instant analysis, determine write data operational readiness or described
Read data operational readiness.
Method the most according to claim 2, it is characterised in that described in write data into the first data Layer
Or second data Layer include: use that to write data into first described in write-back mode Write-Back non-volatile
In memorizer or the second nonvolatile memory.
Method the most according to claim 2, it is characterised in that described in write data into the first data Layer
Or second data Layer include: by described data in units of data block, by directly mapping, complete association is reflected
Penetrate or set associative maps and data block writes described first nonvolatile memory or described second non-volatile deposits
In reservoir.
6. according to the method described in any one of claim 1 to 5, it is characterised in that described method is further
Including: before inquiry the second data Layer or the first data Layer, deposit according to reading data operational readiness inquiry volatibility
Reservoir；The application memory space and the read/write that arrange described volatile memory accelerate memory space, and arrange institute
The read/write stating volatile memory accelerates the minimum and maximum available threshold of memory space.
7. a data storage device, including: processor, the first data Layer and the second data Layer, its feature
It is,
Described processor is used for: after receiving write data requests, write data into first according to data writing operation degree
In data Layer or the second data Layer；After receiving read data request, inquire about the second data according to reading data operational readiness
Layer or the first data Layer, and return Query Result；
Described first data Layer and the second data Layer are used for storing described data.
Data storage device the most according to claim 7, it is characterised in that described first data Layer bag
Including the first nonvolatile memory, described second data Layer includes the second nonvolatile memory.
Data storage device the most according to claim 7, it is characterised in that described processor is further
For using frequency algorithm and prediction hit rate algorithm instant analysis to determine write data operational readiness according to data
Or described reading data operational readiness.
Data storage device the most according to claim 8, it is characterised in that described device is further
Including: the physics mirror image of volatile memory and described volatile memory；
Described processor is further used for before inquiry the second data Layer or the first data Layer, according to reading data
Operational readiness inquires about described volatile memory, and the read/write arranging described volatile memory accelerates memory space
Minimum and maximum available threshold.
11. data storage devices according to claim 8, it is characterised in that described processor enters one
Step is used for by described data in units of data block, is mapped by directly mapping, full associative mapping or set associative
Data block is write in described first nonvolatile memory or described second nonvolatile memory.
12. according to Claim 8 to the data storage device described in 11 any one, it is characterised in that described
First nonvolatile memory includes hard disk；Described second nonvolatile memory includes solid state hard disc and/or non-
Volatile random access memory；Described volatile memory includes random access memory.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510067474.2A CN105988720B (en) | 2015-02-09 | 2015-02-09 | Data storage device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510067474.2A CN105988720B (en) | 2015-02-09 | 2015-02-09 | Data storage device and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105988720A true CN105988720A (en) | 2016-10-05 |
| CN105988720B CN105988720B (en) | 2019-07-02 |
Family
ID=57040959
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510067474.2A Active CN105988720B (en) | 2015-02-09 | 2015-02-09 | Data storage device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105988720B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106844052A (en) * | 2017-01-22 | 2017-06-13 | 郑州云海信息技术有限公司 | A kind of method and device that fusion cluster is built based on Windows Server |
| CN106873903A (en) * | 2016-12-30 | 2017-06-20 | 北京联想核芯科技有限公司 | Date storage method and device |
| CN109491616A (en) * | 2018-11-14 | 2019-03-19 | 三星（中国）半导体有限公司 | The storage method and equipment of data |
| CN113656204A (en) * | 2020-05-12 | 2021-11-16 | 中国移动通信集团浙江有限公司 | Solid state disk management method and device and computing equipment |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101533335A (en) * | 2009-04-09 | 2009-09-16 | 杭州华三通信技术有限公司 | Disk storage system and method for controlling disk storage and device thereof |
| CN101552032A (en) * | 2008-12-12 | 2009-10-07 | 深圳市晶凯电子技术有限公司 | Method and device for constructing a high-speed solid state memory disc by using higher-capacity DRAM to join in flash memory medium management |
| CN102521260A (en) * | 2011-11-18 | 2012-06-27 | 华为技术有限公司 | Data preheating method and device |
| CN102609360A (en) * | 2012-01-12 | 2012-07-25 | 华为技术有限公司 | Data processing method, data processing device and data processing system |
| CN102662459A (en) * | 2012-04-22 | 2012-09-12 | 复旦大学 | Method for reducing energy consumption of server by using mixed storage of solid-state drive and mechanical hard disk |
| WO2014015409A1 (en) * | 2012-07-24 | 2014-01-30 | Institute Of Computer Science Of The Foundation For Research And Technology - Hellas (Ics-Forth) | System and method for implementing ssd-based i/o caches |
| CN104268099A (en) * | 2014-08-29 | 2015-01-07 | 浪潮(北京)电子信息产业有限公司 | Management data reading and writing method and device |
| CN104285214A (en) * | 2012-03-07 | 2015-01-14 | 网络装置公司 | Hybrid storage aggregate block tracking |
-
2015
- 2015-02-09 CN CN201510067474.2A patent/CN105988720B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101552032A (en) * | 2008-12-12 | 2009-10-07 | 深圳市晶凯电子技术有限公司 | Method and device for constructing a high-speed solid state memory disc by using higher-capacity DRAM to join in flash memory medium management |
| CN101552032B (en) * | 2008-12-12 | 2012-01-18 | 深圳市晶凯电子技术有限公司 | Method and device for constructing a high-speed solid state memory disc by using higher-capacity DRAM to join in flash memory medium management |
| CN101533335A (en) * | 2009-04-09 | 2009-09-16 | 杭州华三通信技术有限公司 | Disk storage system and method for controlling disk storage and device thereof |
| CN102521260A (en) * | 2011-11-18 | 2012-06-27 | 华为技术有限公司 | Data preheating method and device |
| CN102609360A (en) * | 2012-01-12 | 2012-07-25 | 华为技术有限公司 | Data processing method, data processing device and data processing system |
| CN104285214A (en) * | 2012-03-07 | 2015-01-14 | 网络装置公司 | Hybrid storage aggregate block tracking |
| CN102662459A (en) * | 2012-04-22 | 2012-09-12 | 复旦大学 | Method for reducing energy consumption of server by using mixed storage of solid-state drive and mechanical hard disk |
| WO2014015409A1 (en) * | 2012-07-24 | 2014-01-30 | Institute Of Computer Science Of The Foundation For Research And Technology - Hellas (Ics-Forth) | System and method for implementing ssd-based i/o caches |
| CN104268099A (en) * | 2014-08-29 | 2015-01-07 | 浪潮(北京)电子信息产业有限公司 | Management data reading and writing method and device |
Non-Patent Citations (1)
| Title |
|---|
| 关玉英 等: "《汇编语言与微机原理教程》", 31 August 2007, 中国计划出版社 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106873903A (en) * | 2016-12-30 | 2017-06-20 | 北京联想核芯科技有限公司 | Date storage method and device |
| CN106873903B (en) * | 2016-12-30 | 2020-02-18 | 深圳忆联信息系统有限公司 | Data storage method and device |
| CN106844052A (en) * | 2017-01-22 | 2017-06-13 | 郑州云海信息技术有限公司 | A kind of method and device that fusion cluster is built based on Windows Server |
| CN109491616A (en) * | 2018-11-14 | 2019-03-19 | 三星（中国）半导体有限公司 | The storage method and equipment of data |
| CN113656204A (en) * | 2020-05-12 | 2021-11-16 | 中国移动通信集团浙江有限公司 | Solid state disk management method and device and computing equipment |
| CN113656204B (en) * | 2020-05-12 | 2023-11-21 | 中国移动通信集团浙江有限公司 | Solid state disk management method and device and computing equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105988720B (en) | 2019-07-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10430084B2 (en) | Multi-tiered memory with different metadata levels | |
| Debnath et al. | FlashStore: High throughput persistent key-value store | |
| US8838895B2 (en) | Solid-state disk caching the top-K hard-disk blocks selected as a function of access frequency and a logarithmic system time | |
| US9767032B2 (en) | Systems and methods for cache endurance | |
| Jiang et al. | S-FTL: An efficient address translation for flash memory by exploiting spatial locality | |
| Li et al. | Tree indexing on solid state drives | |
| CN104115133B (en) | For method, system and the equipment of the Data Migration for being combined non-volatile memory device | |
| CN103942157B (en) | Method and system for calculating the data processing in storage environment | |
| CN107066393A (en) | The method for improving map information density in address mapping table | |
| Liu et al. | PCM-based durable write cache for fast disk I/O | |
| 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 | |
| CN103019958A (en) | Method for managing data in solid state memory through data attribute | |
| WO2013071882A1 (en) | Storage system and management method used for metadata of cluster file system | |
| Wu et al. | APP-LRU: A new page replacement method for PCM/DRAM-based hybrid memory systems | |
| CN110888600A (en) | Buffer area management method for NAND flash memory | |
| Na et al. | Dynamic In-Page Logging for B⁺-tree Index | |
| CN105988720A (en) | Data storage device and method | |
| CN106055679A (en) | Multi-level cache sensitive indexing method | |
| Kargar et al. | Challenges and future directions for energy, latency, and lifetime improvements in NVMs | |
| Huang et al. | SSDs Striking Back: The Storage Jungle and Its Implications to Persistent Indexes. | |
| Fevgas et al. | A study of R-Tree performance in hybrid flash/3DXPoint storage | |
| KR20160121819A (en) | Apparatus for data management based on hybrid memory | |
| Kim et al. | IPL-P: In-page logging with PCRAM | |
| Koo et al. | All-Flash Array Key-Value Cache for Large Objects | |
| JP2020013318A (en) | Database management system, memory management device, database management method, and program |
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 |