CN107066208A - The asymmetric reading/writing method and NVM external equipments of a kind of external equipment - Google Patents
The asymmetric reading/writing method and NVM external equipments of a kind of external equipment Download PDFInfo
- Publication number
- CN107066208A CN107066208A CN201710255785.0A CN201710255785A CN107066208A CN 107066208 A CN107066208 A CN 107066208A CN 201710255785 A CN201710255785 A CN 201710255785A CN 107066208 A CN107066208 A CN 107066208A
- Authority
- CN
- China
- Prior art keywords
- write
- pos
- data block
- len
- dirty
- 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
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
- G06F3/061—Improving I/O performance
-
- 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
- G06F3/0614—Improving the reliability of storage systems
- G06F3/0616—Improving the reliability of storage systems in relation to life time, e.g. increasing Mean Time Between Failures [MTBF]
-
- 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
- G06F3/0638—Organizing or formatting or addressing of data
- G06F3/064—Management of blocks
-
- 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
-
- 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)
- Techniques For Improving Reliability Of Storages (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of asymmetric reading/writing method of external equipment and new NVM external equipments, by separating read-write operation and using asymmetric reading/writing method come work；The asymmetric reading/writing method of external equipment includes the interior external memory mapping method of the reading/writing method, the write method of Dynamic Granulation and many granularities of separate type；Read-write operation is separated first, then write operation is completed by Dynamic Granulation write method, and many granularities for completing interior external memory using external memory mapping method in many granularities map.The present invention, which is avoided that, writes scale-up problem, reduces the writing data quantity of PCIe mouthfuls of NVM storage devices, improves the service life and I/O performances of PCIe mouthfuls of NVM storage devices.
Description
Technical field
The invention belongs to technical field of memory, and in particular to the asymmetric reading/writing method and new PCIe of a kind of external equipment
Mouth NVM storage devices.It is mainly used in avoiding writing scale-up problem, reduces the writing data quantity of PCIe mouthfuls of NVM storage devices, improve PCIe
The service life and I/O performances of mouth NVM storage devices.
Background technology
NVM storage devices are to solve the problems, such as that computer system stores the important technical of wall.DIMM and PCIe mouthfuls of NVM is deposited
Storage equipment is two main at present class NVM storage devices, and compared to DIMM mouthfuls NVM storage devices, PCIe mouthfuls of NVM storage devices have
Larger limitation, in order to obtain higher efficiency of transmission PCIe interface to support based on block access mode, although can change
The size transmitted every time, but far below DIMM interfaces in terms of efficiency and flexibility.There are some researches show access PCIe mouthfuls of NVM
Storage system software expense also accounts for more than the 63% of storage system access delay during storage device, therefore improves related system software
Efficiency be a key.The current storage device accessed by PCIe interface is generally used as block device and managed, in management and
Block granularity is used during read-write, it is impossible to the advantage of byte granularity read-write is supported using NVM storage devices；And writing using block granularity
Mode can be brought and write scale-up problem, and this can seriously reduce the service life of PCIe mouthfuls of NVM storage devices；Simultaneously because adding reality
Border writes the data volume of NVM storage devices, so as to reduce the I/O performances of PCIe mouthfuls of NVM storage devices.And it is former based on locality
Reason, can effectively improve the reading performance of storage device, while read operation does not interfere with NVM storage devices yet by the method pre-read
Service life, therefore it is equally the effective means for improving PCIe mouthfuls of NVM storage device reading performances to choose the read mode of larger granularity.
It is therefore desirable to change management and access PCIe mouthful NVM storage devices mode, distinguish different access characteristic, while acquisition compared with
High readwrite performance.
The content of the invention
It is an object of the invention to provide a kind of asymmetric reading/writing method of external equipment and new NVM storage devices, with
The writing data quantity of PCIe mouthfuls of NVM storage devices is reduced, the service life and I/O performances of PCIe mouthfuls of NVM storage devices is improved.
Use of the present invention for the writing data quantity, PCIe mouthfuls of NVM storage devices of raising of PCIe mouthfuls of NVM storage devices of reduction
Life-span and the consideration of I/O performances, the design separation read-write operation, then write operation is completed by Dynamic Granulation write method, and use
External memory mapping method completes many granularities mapping of interior external memory in many granularities.
The asymmetric reading/writing method of a kind of external equipment, it is characterised in that by separating read-write operation and being read using asymmetric
Write method carrys out work；The asymmetric reading/writing method of the external equipment includes the reading/writing method of separate type, the side of writing of Dynamic Granulation
The interior external memory mapping method of method and many granularities；Specifically include following steps：Read-write operation is separated first, then is write by Dynamic Granulation
Method completes write operation, and many granularities for completing interior external memory using external memory mapping method in many granularities map.
The characteristic of PCIe mouthfuls of NVM storage devices can be directed to by separating read-write operation, kept using data block as unit read operation
While, change the way to manage of write operation, thereby using the different characteristic of read-write operation, improve the I/O performances of external equipment.
The reading/writing method of the separate type specifically includes the following steps：
Step 1, the request of file system access external equipment is received, analysis operation type is that read operation goes to step 2,
Step 3 is gone to for write operation；
Step 2, external equipment completes read operation using the reading method based on data block, and by the data feedback of reading to text
Part system；
Step 3, external equipment completes write operation using the write method of Dynamic Granulation, and result is fed back into file system.
The write method of Dynamic Granulation, to need the actual size of writing data blocks as according to transmission and write-in data, it is to avoid
The problem of writing amplification, solves to there are additional transmissions when performing write operation in units of data block and write PCIe mouthfuls of NVM storages to set
The problem of standby data, the execution efficiency of write operation is improved, extend the life-span of PCIe mouthfuls of NVM storage devices.
The write method of described Dynamic Granulation comprises the following steps：
Step 1, the initial address write_pos that data block need to write is write in the write operation that Study document system is sent, acquisition
With length write_len, and compare the size for writing data block and system data block, turn if the small block with system data of data block is write
To step 2, step 3 is otherwise gone to；
Step 2, set according to write_len and be transferred to the data volume of external equipment, by write-in data, write_pos and
Write_len builds write-in packet and is transferred to external equipment, and will according to write_pos and write_len by external equipment
Write-in packet write-in relevant position is needed, and goes to step 8；
Step 3, according to be to the maximum system data block size by write data be divided into several packets, with actual size structure
Last packet is built, and first packet is gone into step 4 as need to currently write packet；
Step 4, if the size that need to currently write packet is equal to system data block, step 5 is gone to, step is otherwise gone to
6；
Step 5, write_pos value is set to -1, used using data block as unit write method, and will by external equipment
Packet goes to step 7 as a data block write device；
Step 6, write_pos and write_len information is obtained from need to currently write packet, setting is transferred to outer
The data volume of equipment is deposited, and sends write-in packet, write_pos and write_len, last external equipment is according to write_
Pos and write_len need to write packet write-in relevant position, go to step 8；
Step 7, judge whether also next unfinished write-in packet, if any step 4 is then gone to, otherwise go to step
Rapid 8；
Step 8, terminate write operation, and successful information is write to file system feedback.
The interior external memory mapping method of many granularities, can not be split by data block and be limited, and neatly correspond to write-in data
Map in item, preserve to be actually needed in each mapping item and be written to the initial address and length of outer deposit data, it is to avoid including
Scale-up problem is write when external memory maps, it is ensured that file system can obtain the number for being actually needed PCIe mouthfuls of NVM storage devices of write-in
According to initial address and length so that Dynamic Granulation writes data when algorithm is avoided that transmission and write, and is deposited to improve PCIe mouthfuls of NVM
The service life and I/O performances for storing up equipment provide support.
The interior external memory mapping method of described many granularities comprises the following steps：
Step 1, increase dirty_pos and dirty_len in each mapping of interior external memory mapping table, preserve interior external memory
Between many granularities map information, wherein dirty_pos represents to need the starting updated to external memory data field in the data block
Address, dirty_len represents the length for needing to update to external memory data field in the data block, and dirty_pos value represents for -1
The data block does not use many granularity mapping methods, S (S ∈ N) and L (L ∈ N) represent to write respectively data initial logical address with
Length, goes to step 2；
Step 2, the initial address phase of mapping corresponding in interior external memory mapping table, such as S and the mapping is searched according to S
With and L value it is identical with data block size, then the dirty_pos of the mapping is set to -1, it is otherwise that S is corresponding physically
Location is saved in dirty_pos, goes to step 3；
Step 3, compare L value whether beyond the mapping counterlogic block length, such as without departing from then being set using L
Dirty_len value is put, and L value is reset, step 4 is gone to；Otherwise calculated according to data block size and dirty_pos
Dirty_len value, and S and L are updated, go to step 4；
Step 4, if L value does not go to step 2 for 0, whole operation is otherwise terminated.
New NVM external equipments, include writing module, the interior external memory mapping of many granularities of read and write abruption module, Dynamic Granulation
Module and with PCIe mouthfuls of NVM storage devices, it is characterised in that：Using read and write abruption module, separate file system is to PCIe
The read-write operation of mouth NVM storage devices；Using the writing module of Dynamic Granulation, the actual size for writing data block, dynamic adjustment are obtained
PCIe mouthfuls of transmission granularity, is only written the data for being actually needed write-in, it is to avoid write amplifieroperation；Reflected using the interior external memory of many granularities
Module is penetrated, is not split by data block and is limited, is neatly corresponded to write-in data in mapping item, it is to avoid when interior external memory maps
Write scale-up problem；Using with PCIe mouthfuls of NVM storage devices, received file system read-write operation is completed.
The present invention has beneficial effect.Different qualities and PCIe mouthfuls of NVM that the present invention is operated for file system read-write
The characteristics of storage device, by avoiding writing amplification, reduce the data for completing that PCIe mouthfuls of NVM storage devices are actually written into during write operation
Amount, improves I/O performances, the delay service life of PCIe mouthfuls of NVM storage devices.
Brief description of the drawings
Fig. 1 is overall structure figure of the present invention；
Fig. 2 is the test result of Fio order write performances when accessing different size file；
Fig. 3 is the test result of Fio order reading performances when accessing different size file；
Fig. 4 is the test result of Fio order write performances when setting different system data block size；
Fig. 5 is the I/O results of property of Filebench tests；
When Fig. 6 is that write-in data volume is less than system data block size, the write method schematic diagram of the Dynamic Granulation used；
When Fig. 7 is that write-in data volume is more than system data block size, the write method schematic diagram of the Dynamic Granulation used；
Fig. 8 is the interior external memory mapping method using many granularities, obtains write operation and calculates mapping item dirty_pos and dirty_
Len schematic diagram；
Fig. 9 is the schematic diagram of read operation.
Embodiment
Below in conjunction with the accompanying drawings and specific embodiment, technical scheme is described in further details.
Embodiment 1
As shown in figure 1, the new NVM external equipments include read and write abruption module, the writing module of Dynamic Granulation, many granularities
Interior external memory mapping block and with PCIe mouthfuls of NVM storage devices.The explanation of each functional module is as shown in table 1.
Each functional module in the new NVM external equipments of table 1
Embodiment 2
In order to realize new NVM external equipments, because currently without practical PCIe mouth NVM storage devices, we use
PMBD is simulated, and using default configuration, increases 85ns read latency and 500ns write delay simulation NVM media；Change simultaneously
Linux kernel, 10G kernel spacing and carry PMBD are reserved in kernel address afterbody；
PMBD, increase read and write abruption module and Dynamic Granulation writing module are changed, NVM bytes are utilized when performing write operation
Characteristic is addressed, actual increased data are only written.We change the mapping buffer_head of external memory in linux kernel simultaneously
Structure, increases the interior external memory mapping block of many granularities, and increase dirty_pos and dirty_len preservation needs are actually needed and write back
The data positional information of external memory, realizes the interior external memory mapping algorithm of many granularities；Linux kernel is changed simultaneously to interact with device drives
Bio structures, increase write_pos and write_len preserve the dirty_pos that obtains from buffer_head structures and
Dirty_len information, so as to will realize that write back data positional information is delivered to PCIe mouthfuls of NVM storage devices；Build new NVM
The prototype of external equipment, we are named as APMBD.
In order to reduce influence of the network service to test I/O performances, test environment is built using a machine, is furnished with
Intel (R) Xeon (R) E5606 2.13GHz processors and 20G internal memories.Testing tool is used as using Fio and Filebench.
Using the libaio in Fio, APMBD I/O performances are tested, and carry out with Ext4on PMBD and Ext2on PMBD
Contrast.It is 4KB to set and access block size, and it is respectively 8KB, 32KB, 128KB and 512KB file that testing sequence, which reads and writes 30,000 sizes,
When performance, Fig. 2 is the test result of order write performance, and Fig. 3 is the test result of order reading performance；System data is set in addition
The size of block is respectively 4KB, 8KB, 16KB and 32KB, and testing sequence writes the performance of 30,000 512KB files, as a result such as Fig. 4 institutes
Show.
Carry out the access situation of simulation application server using Filebench simultaneously, have selected Copyfile, Webserver
With Fileserver three types load testing I/O performances, setting file size is 32KB, and quantity of documents is 50,000, with
Ext4on PMBD are contrasted, and test result is as shown in Figure 5.
Embodiment 3
When realizing the prototype of new NVM external equipments, mainly include read and write abruption thread, Dynamic Granulation and write thread and many
External memory maps thread in granularity, and their job step difference is as follows.
Read and write abruption thread：
Step 1, the request of file system access external equipment is received, analysis operation type is that read operation goes to step 2,
Step 3 is gone to for write operation；
Step 2, external equipment completes read operation using the reading method based on data block, and by the data feedback of reading to text
Part system；
Step 2, external equipment completes write operation using the write method of Dynamic Granulation, and result is fed back into file system.
Dynamic Granulation writes thread：
Step 1, the initial address write_pos that data block need to write is write in the write operation that Study document system is sent, acquisition
(write_pos ∈ N) and length write_len (write_len ∈ N), and compare the size for writing data block and system data block,
It is small with going to step 2 if system data block if writing data block, otherwise go to step 3；
Step 2, set according to write_len and be transferred to the data volume of external equipment, by write-in data, write_pos and
Write_len builds write-in packet and is transferred to external equipment, and will according to write_pos and write_len by external equipment
Write-in packet write-in relevant position is needed, and goes to step 8；
Step 3, according to be to the maximum system data block size by write data be divided into several packets, with actual size structure
Last packet is built, and first packet is gone into step 4 as need to currently write packet；
Step 4, if the size that need to currently write packet is equal to system data block, step 5 is gone to, step is otherwise gone to
6；Step 5, write_pos value is set to -1, used using data block as unit write method, and by external equipment by packet
As a data block write device, step 7 is gone to；
Step 6, write_pos and write_len information is obtained from need to currently write packet, setting is transferred to outer
The data volume of equipment is deposited, and sends write-in packet, write_pos and write_len, last external equipment is according to write_
Pos and write_len need to write packet write-in relevant position, go to step 8；
Step 7, judge whether also next unfinished write-in packet, if any step 4 is then gone to, otherwise go to step
Rapid 8；
Step 8, terminate write operation, and successful information is write to file system feedback.
External memory maps thread in many granularities：
Step 1, the increase dirty_pos (dirty_pos ∈ N) and dirty_ in each mapping of interior external memory mapping table
Len (diry_len ∈ N), preserves the map information of many granularities between interior external memory, and wherein dirty_pos is represented in the data block
The initial address updated to external memory data field is needed, dirty_len represents to need to update to external memory data field in the data block
Length, dirty_pos value represents that the data block does not use many granularity mapping methods, S (S ∈ N) and L (L ∈ N) difference tables for -1
Show data initial logical address to be write and length, go to step 2；
Step 2, the initial address phase of mapping corresponding in interior external memory mapping table, such as S and the mapping is searched according to S
With and L value it is identical with data block size, then the dirty_pos of the mapping is set to -1, it is otherwise that S is corresponding physically
Location is saved in dirty_pos, goes to step 3；
Step 3, compare L value whether beyond the mapping counterlogic block length, such as without departing from then being set using L
Dirty_len value is put, and L value is reset, step 4 is gone to；Otherwise calculated according to data block size and dirty_pos
Dirty_len value, and S and L are updated, go to step 4；
Step 4, if L value does not go to step 2 for 0, whole operation is otherwise terminated.
When Fig. 6 gives write-in data volume less than system data block size, the write method schematic diagram of the Dynamic Granulation used.
When Fig. 7 gives write-in data volume more than system data block size, the write method schematic diagram of the Dynamic Granulation used.
Fig. 8 gives the interior external memory mapping method using many granularities, obtain write operation calculate mapping item dirty_pos and
Dirty_len schematic diagram.
Fig. 9 gives the schematic diagram of read operation.
Claims (3)
1. a kind of asymmetric reading/writing method of external equipment, it is characterised in that：By separating read-write operation and being read using asymmetric
Write method carrys out work；The asymmetric reading/writing method of the external equipment includes the reading/writing method of separate type, the side of writing of Dynamic Granulation
The interior external memory mapping method of method and many granularities；Specifically include following steps：Read-write operation is separated first, then is write by Dynamic Granulation
Method completes write operation, and many granularities for completing interior external memory using external memory mapping method in many granularities map；
The reading/writing method of the separate type specifically includes the following steps：
Step 1.1, the request of file system access external equipment is received, analysis operation type is that read operation goes to step 1.2,
Step 1.3 is gone to for write operation；
Step 1.2, external equipment completes read operation using the reading method based on data block, and by the data feedback of reading to file
System；
Step 1.3, external equipment completes write operation using the write method of Dynamic Granulation, and result is fed back into file system.
2. a kind of asymmetric reading/writing method of external equipment according to claim 1, it is characterised in that described dynamic grain
The write method of degree comprises the following steps：
Step 2.1, the write operation that Study document system is sent, obtain write initial address write_pos that data block need to write and
Length write_len, wherein write_pos ∈ N, write_len ∈ N, and compare the size for writing data block and system data block,
It is small with going to step 2.2 if system data block if writing data block, otherwise go to step 2.3；
Step 2.2, set according to write_len and be transferred to the data volume of external equipment, by write-in data, write_pos and
Write_len builds write-in packet and is transferred to external equipment, and will according to write_pos and write_len by external equipment
Write-in packet write-in relevant position is needed, and goes to step 2.8；
Step 2.3, according to be to the maximum system data block size by write data be divided into several packets, built with actual size
Last packet, and first packet is gone into step 2.4 as need to currently write packet；
Step 2.4, if the size that need to currently write packet is equal to system data block, step 2.5 is gone to, step is otherwise gone to
2.6；
Step 2.5, write_pos value is set to -1, used using data block as unit write method, and by external equipment by number
According to bag as a data block write device, step 2.7 is gone to；
Step 2.6, write_pos and write_len information is obtained from need to currently write packet, setting is transferred to external memory
The data volume of equipment, and write-in packet, write_pos and write_len are sent, last external equipment is according to write_pos
Packet write-in relevant position need to be write with write_len, step 2.8 is gone to；
Step 2.7, judge whether also next unfinished write-in packet, if any step 2.4 is then gone to, otherwise go to step
Rapid 2.8；
Step 2.8, terminate write operation, and successful information is write to file system feedback.
3. a kind of asymmetric reading/writing method of external equipment according to claim 1, it is characterised in that described Quito grain
The interior external memory mapping method of degree comprises the following steps：
Step 3.1, dirty_pos and dirty_len, wherein dirty_ are increased in each mapping of interior external memory mapping table
Pos ∈ N, diry_len ∈ N, preserve the map information of many granularities between interior external memory, and wherein dirty_pos is represented in the data block
The middle initial address for needing to update to external memory data field, dirty_len represents to need to update in the data block to arrive external memory data field
Length, dirty_pos value do not use many granularity mapping methods for -1 expression data block, and S (S ∈ N) and L (L ∈ N) distinguish
Expression will write data initial logical address and length, go to step 3.2；
Step 3.2, mapping corresponding in interior external memory mapping table is searched according to S, such as S is identical with the initial address of the mapping
And L value is identical with data block size, then the dirty_pos by the mapping is set to -1, otherwise by the corresponding physical address of S
It is saved in dirty_pos, goes to step 3.3；
Step 3.3, compare L value whether beyond the mapping counterlogic block length, such as without departing from then being set using L
Dirty_len value, and L value is reset, go to step 3.4；Otherwise calculated according to data block size and dirty_pos
Dirty_len value, and S and L are updated, go to step 3.4；
Step 3.4, if L value does not go to step 3.2 for 0, whole operation is otherwise terminated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710255785.0A CN107066208B (en) | 2017-04-19 | 2017-04-19 | Asymmetric read-write method of external memory device and NVM (non-volatile memory) external memory device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710255785.0A CN107066208B (en) | 2017-04-19 | 2017-04-19 | Asymmetric read-write method of external memory device and NVM (non-volatile memory) external memory device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107066208A true CN107066208A (en) | 2017-08-18 |
| CN107066208B CN107066208B (en) | 2020-04-03 |
Family
ID=59600355
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710255785.0A Active CN107066208B (en) | 2017-04-19 | 2017-04-19 | Asymmetric read-write method of external memory device and NVM (non-volatile memory) external memory device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107066208B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111341374A (en) * | 2018-12-18 | 2020-06-26 | 炬芯（珠海）科技有限公司 | Memory test method and device and readable memory |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103268292A (en) * | 2013-06-13 | 2013-08-28 | 江苏大学 | Method for prolonging life of non-volatile external memory and high-speed long-life external memory system |
| CN103942159A (en) * | 2014-03-19 | 2014-07-23 | 华中科技大学 | Data read-write method and device based on mixed storage device |
| CN105335100A (en) * | 2015-09-29 | 2016-02-17 | 华为技术有限公司 | Method and device for processing data, and flash memory device |
-
2017
- 2017-04-19 CN CN201710255785.0A patent/CN107066208B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103268292A (en) * | 2013-06-13 | 2013-08-28 | 江苏大学 | Method for prolonging life of non-volatile external memory and high-speed long-life external memory system |
| CN103942159A (en) * | 2014-03-19 | 2014-07-23 | 华中科技大学 | Data read-write method and device based on mixed storage device |
| CN105335100A (en) * | 2015-09-29 | 2016-02-17 | 华为技术有限公司 | Method and device for processing data, and flash memory device |
Non-Patent Citations (2)
| Title |
|---|
| CAI TAO ETC.: "NVMCFS: Complex File System for Hybrid NVM", 《2016 IEEE 22ND INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED SYSTEMS》 * |
| 何耀等: "面向NVM的混合粒度文件系统", 《软件导刊》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111341374A (en) * | 2018-12-18 | 2020-06-26 | 炬芯（珠海）科技有限公司 | Memory test method and device and readable memory |
| CN111341374B (en) * | 2018-12-18 | 2022-03-15 | 炬芯科技股份有限公司 | Memory test method and device and readable memory |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107066208B (en) | 2020-04-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11029853B2 (en) | Dynamic segment allocation for write requests by a storage system | |
| CN105830059B (en) | File access method, device and storage equipment | |
| JP5347061B2 (en) | Method and apparatus for storing data in a flash memory data storage device | |
| US9135190B1 (en) | Multi-profile memory controller for computing devices | |
| CN103425602B (en) | A kind of method, device and the host computer system of data of flash memory storage equipment read-write | |
| US10031674B2 (en) | DIMM SSD addressing performance techniques | |
| US10430329B2 (en) | Quality of service aware storage class memory/NAND flash hybrid solid state drive | |
| GB2507410A (en) | Storage class memory having low power, low latency, and high capacity | |
| US20170003911A1 (en) | Information processing device | |
| CN102999441A (en) | Fine granularity memory access method | |
| WO2015058641A1 (en) | Hybrid hard disk implementation method and device | |
| US10503657B2 (en) | DIMM SSD Addressing performance techniques | |
| KR102320044B1 (en) | Pci device, interface system including same, and computing system including same | |
| CN107066208A (en) | The asymmetric reading/writing method and NVM external equipments of a kind of external equipment | |
| JP6100927B2 (en) | Information processing device | |
| US8504764B2 (en) | Method and apparatus to manage object-based tiers | |
| US10437478B1 (en) | Replication based on a differential multiple write command | |
| WO2020055534A1 (en) | Hybrid memory system interface | |
| TWI805937B (en) | Data processing method and memory controller utilizing the same | |
| CN100405777C (en) | Cashe method based on target device internal storage in ether net storage regional network | |
| CN104424124A (en) | Memory device, electronic equipment and method for controlling memory device | |
| US10671322B1 (en) | Method for performing storage space management, associated data storage device, and controller thereof | |
| CN117806570B (en) | Online memory expansion method, device, equipment and storage medium | |
| US20230359394A1 (en) | Operating method of memory device for managing map data of each of plurality of storage devices, computing system including memory device, and operating method of computing system | |
| US9164681B1 (en) | Method and apparatus for dynamic path-selection for improving I/O performance in virtual provisioned storage arrays with data striping |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |