CN111831228A - SSD (solid State disk) cold and hot data separation method based on out-of-band interaction - Google Patents
SSD (solid State disk) cold and hot data separation method based on out-of-band interaction Download PDFInfo
- Publication number
- CN111831228A CN111831228A CN202010644277.3A CN202010644277A CN111831228A CN 111831228 A CN111831228 A CN 111831228A CN 202010644277 A CN202010644277 A CN 202010644277A CN 111831228 A CN111831228 A CN 111831228A
- Authority
- CN
- China
- Prior art keywords
- lba
- data
- ssd
- cold
- host
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000926 separation method Methods 0.000 title claims abstract description 35
- 230000003993 interaction Effects 0.000 title claims abstract description 17
- 239000007787 solid Substances 0.000 title description 4
- 238000012005 ligant binding assay Methods 0.000 claims abstract description 119
- 230000005540 biological transmission Effects 0.000 claims abstract description 9
- 238000000034 method Methods 0.000 claims abstract description 8
- 238000004422 calculation algorithm Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 3
- 238000007726 management method Methods 0.000 description 18
- 230000006870 function Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 4
- 238000007635 classification algorithm Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000035939 shock Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
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
- G06F3/0607—Improving or facilitating administration, e.g. storage management by facilitating the process of upgrading existing storage systems, e.g. for improving compatibility between host and storage device
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
-
- 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
- G06F3/0656—Data buffering arrangements
-
- 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
- G06F3/0683—Plurality of storage devices
- G06F3/0688—Non-volatile semiconductor memory arrays
Abstract
The invention discloses an SSD cold and hot data separation method based on out-of-band interaction, which comprises the following steps: s01), establishing an out-of-band management path between the host and the SSD device independent of the host transmission path; s02), the SSD device sends the LBA access information to the host through the out-of-band management channel, wherein the LBA access information comprises the LBA value and the write operation execution time; s03), the host completes cold and hot classification of all LBAs based on the LBA access information, and sends the classification result to the SSD device through an out-of-band management channel; s04), the SSD device completes the physical separation of cold and hot data according to the classification result. The method is used for solving the problem of wear balance of the Nand Flash storage medium in the SSD device.
Description
Technical Field
The invention relates to the field of flash memory storage control, in particular to an SSD (solid State disk) cold and hot data separation method based on out-of-band interaction.
Background
Solid State Disk (SSD) mainly uses Flash memory (Nand Flash) as its permanent storage medium, and records 0 and 1 through potential height or phase State difference, so it has the characteristics of fast read-write speed, small volume, light weight, low power consumption, and good shock resistance, but because there is a frequency limit in Nand Flash programming and erasing operation (PE), it also needs to be subjected to wear leveling operation, and it ensures the PE frequency balance of all physical blocks in Nand Flash. One of the implementation manners of wear leveling is to perform cold-hot separation on all blocks, that is, to store data (hot data) with higher execution frequency of write operation into blocks with lower PE times, and to store data with lower frequency into blocks with higher PE times, so that the key point is how to distinguish cold data from hot data. The current mainstream cold and hot data distinguishing mode is to record the write operation times of all logical pages (equal to LBA), perform cold and hot data division according to the write operation times, and finally move the cold and hot data to corresponding physical blocks respectively. The method is simple to execute, saves the time expenditure of an SSD system, can effectively distinguish cold data and hot data to a certain extent, and has the following defects: the running storage space of the SSD system is wasted, for example, an SSD with a capacity of nominally 1.92TB contains about 470,000,000 LBAs in units of 4KB, the number of write operations of each LBA needs to be recorded by using one DWord (4Byte), and 1,875,374,424Byte (about 1875GB) of running storage space is consumed in total; only the number of write operations is taken as a cold and hot determination basis, without considering a time factor, if the number of write operations of a certain LBA is high, but the number of write operations executed in a unit time is low or the time of the latest write operation is far from the current time, the LBA cannot be simply determined as hot data, and a time factor comprehensive analysis needs to be added.
Out-of-Band Management (Out-of-Band Management) is a Management method based on hardware, and usually uses a dedicated hardware module to provide a Management interface independent of a main logic hardware module, and a Host can monitor and maintain a device through a dedicated data channel. Some SSD devices currently implement out-of-band management interfaces. Most SSD devices which realize the out-of-band management technology currently perform out-of-band interaction with a Host based on SMBus/IIC interfaces or PCIe interfaces, and the interaction mode is mainly divided into the following two modes;
1. the Host reads the data of the fixed address of the SSD equipment through the out-of-band management interface to acquire the internal state information of the SSD equipment;
2. the Host sends the request message to the SSD device through the out-of-band management interface, and the SSD processes the received request message and sends the completion message to the Host through the out-of-band management interface.
The mode 1 is a simplex data mode, data can only be transmitted to the Host by the SSD device, and the Host cannot request the SSD device to execute specific operations by sending command data. The mode 2 is a half-duplex mode, data can be bidirectionally transmitted in different time periods, the Host can send instruction data to the SSD device, the SSD device executes related operations after receiving the instructions and sends execution results to the Host, but the SSD device cannot actively transmit the instructions and the data to the Host in the mode.
Disclosure of Invention
The invention aims to provide an SSD cold and hot data separation method based on out-of-band interaction, which is used for solving the problem of wear balance of Nand Flash storage media in SSD equipment.
In order to solve the technical problem, the technical scheme adopted by the invention is as follows: an SSD cold and hot data separation method based on out-of-band interaction is characterized in that: the method comprises the following steps:
s01), establishing an out-of-band management path between the host and the SSD device independent of the host transmission path;
s02), the SSD device sends the LBA access information to the host through the out-of-band management channel, wherein the LBA access information comprises the LBA value and the write operation execution time;
s03), the host completes cold and hot classification of all LBAs based on the LBA access information, and sends the classification result to the SSD device through an out-of-band management channel;
s04), the SSD device completes the physical separation of cold and hot data according to the classification result.
Further, the host machine uses a k-nearest neighbor clustering algorithm to divide hot, warm and cold data types of the LBA access information, and data classification is completed based on the similarity between samples.
Further, the specific process of the host using the k-nearest neighbor clustering algorithm to divide the hot, warm and cold data types of the LBA access information is as follows:
s31), after receiving the LBA access information data uploaded by the SSD device, the host updates the related LBA access information in the local cache, wherein the related LBA access information comprises the LBA write operation times NwrLBA earliest write operation time T1stTime T of latest write operation of LBAlastLBA write frequency Fwr；
S32), when the number of times of uploading data of the SSD device reaches the threshold or the number of times of writing a certain LBA reaches the threshold, the host uses all LBAs as the sample set S ═ i1,i2,i3,...,inWhere N denotes the total number of LBAs, in Nwr、Tlast、FwrCreating and initializing a hot data cluster C as a sample attributeHTemperature data cluster CWAnd cold data cluster CC；
S33), traversing the sample set and respectively finding out Nwr、Tlast、FwrMaximum value of (N) maxwr)、max(Tlast)、max(Fwr)；
S34), traversing the sample set, calculating the data temperature of all LBAs based on the following formula:
and selecting a maximum temperature sample and a minimum temperature sample from the following samples: i ismax＝{Nwr_max,Tlast_max,Fwr_max}，Imin＝{Nwr_min,Tlast_min,Fwr_min}；
S35), further calculating an intermediate temperature sample based on the maximum temperature sample and the minimum temperature sample:
s36), by Imax、Imid、IminAs CH、CW、CCInitial mean vector mu ofH、μW、μC；
S37), calculating each sample and mu in the sample setH、μW、μCEuclidean distance between:
s38), putting each sample in the sample set into the data cluster closest to the sample set, and counting the sample data n contained in each data clusterjCalculate a new mean vector μ'H、μ'W、μ'C：
S39), check μH、μW、μCMu's'H、μ'W、μ'CIf so, finishing the cold and hot data classification operation, and taking the current result as the final cold and hot data classification result; if not, the mu is processedH、μW、μCAre respectively replaced by mu'H、μ'W、μ'CReturning to step S37.
Further, a cache for storing all the LBA access information is allocated in the host, and before the host receives the SSD device host access information, the host initializes the LBA access information cache, and records the number N of write operations of each LBA using the LBA sequence number as an indexwrEarliest, the firstWrite operation time T1stLatest write operation time TlastFrequency of writing Fwr(ii) a The host end receives the LBA access information data sent by the SSD equipment through an out-of-band channel, and executes the following steps on each record in the received data:
A. reading the serial number i of the LBA in the record, and accessing the N of the ith LBA in the information cache for the local LBAwrExecuting an adding 1 operation;
B. reading the time stamp T in the record, and checking the T of the iLBA in the local LBA access information cache1stIf it is 0, if so, then T is1stSetting t, if not, not processing;
C. caching the I-th LBA T in the local LBA access informationlastSetting as t;
D. f for calculating ith LBA in local LBA access information cachewr：
Further, only a buffer with limited capacity is allocated to the LBA access information in the SSD device memory; when the host executes write operation on a certain LBA, the LBA value and the current timestamp are recorded in the SSD device cache, and when the data volume of the LBA access information in the cache reaches a threshold value, all data in the cache is uploaded to the host end through an out-of-band channel.
Further, the out-of-band management path is in a full-duplex data mode, and is realized based on an SMBus/IIC interface, and the specific implementation mode is as follows: the host and the SSD device are both provided with 2 SMBus/IIC interfaces, one SMBus/IIC interface serves as a Master end and sends request messages and data to the opposite-end device, and the other SMBus/IIC interface serves as a Slave end and receives completion messages sent by the opposite-end device.
Further, when the host side sends a cold and hot data separation request, the host selects two types with small data volume from the three types of data clusters, sends the LBA and the temperature type contained in the two types to the SSD side, and completes physical separation of the cold and hot data at the SSD side.
The invention has the beneficial effects that: the SSD cold and hot data separation technology based on out-of-band interaction is mainly used for solving the problem of wear balance of Nand Flash storage media in SSD equipment. The time attribute is added on the basis of the LBA write operation times as a measurement parameter of the data temperature, and meanwhile, the data are divided into hot, warm and cold types by using a k neighbor clustering algorithm, so that the data classification precision can be effectively improved; the hot and cold data classification operation is distributed to the Host end to be executed, so that the non-service resource consumption of the SSD equipment can be effectively saved; data transmission between the SSD device and the Host is realized when the cold and hot data separation operation is performed through the out-of-band channel, so that the transmission bandwidth of service data can be effectively saved.
Drawings
FIG. 1 is a diagram of an out-of-band communications hardware architecture;
FIG. 2 is a flow chart of a cold-hot data separation operation;
FIG. 3 is a flow chart of a cold-hot data classification algorithm.
Detailed Description
Example 1
The embodiment discloses an SSD cold and hot data separation method based on out-of-band interaction. In order to implement the out-of-band management technology, the Host and the SSD device need to implement a dedicated data interface independent of the main transmission path, and implement a related driver and service program based on the data interface.
In view of the defects of the existing simplex data mode and half-duplex data mode, the embodiment provides an out-of-band management path of a full-duplex data mode, and implements the software and hardware architecture of Host and SSD devices based on SMBus/IIC interfaces, as shown in fig. 1, the implementation is specifically: the Host and the SSD equipment respectively need to realize 2 SMBus/IIC interfaces, wherein one SMBus/IIC interface is used as a Master end and sends request information and data to opposite-end equipment; and the other SMBus/IIC interface is used as a Slave end and receives a completion message sent by the opposite end equipment. Through the full-duplex data mode, the function that the SSD device actively transmits the internal LBA access statistical information data to the Host through the out-of-band management interface can be realized, and the function that the Host transmits the cold and hot data separation result to the SSD device is realized.
After the out-of-band management technology is realized, the LBA access information uploading function of the SSD end and the hot and cold data separation request issuing function of the Host end need to be realized based on the interactive mode, and a hot and cold data classification algorithm is realized at the Host end.
When the SSD end LBA access information uploading function is realized, in order to save the storage space in the SSD during operation and simplify the LBA access information updating mode, only a buffer with limited capacity is allocated to the LBA access information in the SSD operation, and a buffer for storing all the LBA access information is allocated in the Host. When Host executes write operation on a certain LBA, recording the LBA value and the current timestamp in a cache; and when the data quantity of the LBA access information in the cache reaches a threshold value, uploading all the data in the cache to a Host end through an out-of-band channel.
FIG. 2 is a flow chart of the cold-hot data separation operation of the present invention. When the Host writes data into the SSD device, the SSD device simultaneously records the LBA for executing the write operation and the current time, and stores the LBA and the current time into the LBA access information cache of the storage memory during the operation. And if the LBA access information cache space of the SSD equipment is full, sending the LBA access information data to the Host end through the out-of-band interface. And the Host executes preprocessing operation on the data, namely updating the write operation times, the earliest write operation time, the latest write operation time and the write operation execution frequency of the LBA, and storing the updated information into the local LBA access information cache. And if the write operation times of a certain LBA reach a threshold value, the Host executes the cold and hot data classification operation based on the clustering algorithm on all the LBAs, and sends the classification result to the SSD end through an out-of-band interface. After receiving the cold and hot data classification result, the SSD executes physical separation of the cold and hot data.
Fig. 3 is a flowchart of a cold and hot data classification algorithm according to the present invention, in this embodiment, a specific implementation process of cold and hot data classification is as follows:
s1), the host side initializes the LBA access information buffer, takes the LBA serial number as the index, records the writing operation times N of each LBAwrEarliest write operation time T1stLatest write operation time TlastFrequency of writing Fwr；
S2), the host receives the LBA access information data sent by the SSD device through the out-of-band channel, and executes the following steps for each record in the received data:
A. reading the serial number i of the LBA in the record, and accessing the N of the ith LBA in the information cache for the local LBAwrExecuting an adding 1 operation;
B. reading the time stamp T in the record, and checking the T of the iLBA in the local LBA access information cache1stIf it is 0, if so, then T is1stSetting t, if not, not processing;
C. caching the I-th LBA T in the local LBA access informationlastSetting as t;
D. f for calculating ith LBA in local LBA access information cachewr：
S3), checking whether the write operation frequency in the LBA which is locally updated reaches the cold and hot data separation threshold or the SSD data uploading frequency reaches the threshold, if yes, proceeding to the step S4, and if not, returning to the step S2;
s4), with all LBAs as sample set S ═ i1,i2,i3,...,inWhere N denotes the total number of LBAs, in Nwr、Tlast、FwrCreating and initializing a hot data cluster C as a sample attributeHTemperature data cluster CWAnd cold data cluster CC；
S5), traversing the sample set and respectively finding out Nwr、Tlast、FwrMaximum value of (N) maxwr)、max(Tlast)、max(Fwr)；
S6), traversing the sample set, calculating the data temperature of all LBAs based on the following formula:
and selecting the maximum temperature sample ImaxAnd minimum temperature sample Imin：
Imax＝{Nwr_max,Tlast_max,Fwr_max}，
Imin＝{Nwr_min,Tlast_min,Fwr_min}；
S7), further calculating an intermediate temperature sample based on the maximum temperature sample and the minimum temperature sample:
s8), by Imax、Imid、IminAs CH、CW、CCInitial mean vector mu ofH、μW、μC；
S9), calculating each sample and mu in the sample setH、μW、μCEuclidean distance between:
s10), putting each sample in the sample set into the data cluster closest to the sample set, and counting the sample data n contained in each data clusterjCalculate a new mean vector μ'H、μ'W、μ'C：
S11), check μH、μW、μCMu's'H、μ'W、μ'CIf so, finishing the cold and hot data classification operation, and taking the current result as the final cold and hot data classification result; if not, the mu is processedH、μW、μCAre respectively replaced by mu'H、μ'W、μ'CReturning to step S9.
When the function of sending the hot and cold data separation request at the Host end is realized, the Host can select two types with less data volume from the three types of data clusters, the LBA and the temperature type contained in the two types of data clusters are sent to the SSD end, and the physical separation of the hot and cold data is completed at the SSD end.
According to the invention, the out-of-band management channel independent of the main transmission channel is constructed, the SSD device sends the LBA access information to the Host, the Host completes the cold and hot classification of all LBAs, and then sends the classification result to the SSD device, and finally completes the physical separation of cold and hot data. Compared with the prior art, the innovation is mainly as follows: the hot and cold data classification operation is executed at the Host end, the CPU resource and the run-time storage resource of the SSD device are effectively released, and the SSD device only needs to allocate a small amount of run-time cache and execute the physical separation of the hot and cold data; data transmission between the SSD device and the Host is completed through the out-of-band channel when the cold and hot data separation operation is performed, and service data transmission bandwidth does not need to be occupied; besides the number of times of write operation, the executable frequency of the write operation execution time is used as an attribute parameter of data temperature detection, so that the accuracy of cold and hot data classification is improved; and performing hot, warm and cold type division on the data by using a k nearest neighbor clustering algorithm, and finishing data classification based on the similarity among samples.
The foregoing description is only for the basic principle and the preferred embodiments of the present invention, and modifications and substitutions by those skilled in the art are included in the scope of the present invention.
Claims (7)
1. An SSD cold and hot data separation method based on out-of-band interaction is characterized in that: the method comprises the following steps:
s01), establishing an out-of-band management path between the host and the SSD device independent of the host transmission path;
s02), the SSD device sends the LBA access information to the host through the out-of-band management channel, wherein the LBA access information comprises the LBA value and the write operation execution time;
s03), the host completes cold and hot classification of all LBAs based on the LBA access information, and sends the classification result to the SSD device through an out-of-band management channel;
s04), the SSD device completes the physical separation of cold and hot data according to the classification result.
2. The SSD cold and hot data separation method based on out-of-band interaction of claim 1, wherein: in step S03, the host uses the k-nearest neighbor clustering algorithm to classify the LBA access information into hot, warm, and cold data types, and completes data classification based on the inter-sample similarity.
3. The SSD cold and hot data separation method based on out-of-band interaction of claim 2, wherein: the specific process of the host for dividing the hot, warm and cold data types of the LBA access information by using the k nearest neighbor clustering algorithm is as follows:
s31), after receiving the LBA access information data uploaded by the SSD device, the host updates the related LBA access information in the local cache, wherein the related LBA access information comprises the LBA write operation times NwrLBA earliest write operation time T1stTime T of latest write operation of LBAlastLBA write frequency Fwr；
S32), when the number of times of uploading data of the SSD device reaches the threshold or the number of times of writing a certain LBA reaches the threshold, the host uses all LBAs as the sample set S ═ i1,i2,i3,...,inWhere N denotes the total number of LBAs, in Nwr、Tlast、FwrCreating and initializing a hot data cluster C as a sample attributeHTemperature data cluster CWAnd cold data cluster CC；
S33), traversing the sample set and respectively finding out Nwr、Tlast、FwrMaximum value of (N) maxwr)、max(Tlast)、max(Fwr)；
S34), traversing the sample set, calculating the data temperature of all LBAs based on the following formula:
and selecting a maximum temperature sample and a minimum temperature sample from the following samples: i ismax＝{Nwr_max,Tlast_max,Fwr_max}，Imin＝{Nwr_min,Tlast_min,Fwr_min}；
S35), further calculating an intermediate temperature sample based on the maximum temperature sample and the minimum temperature sample:
s36), by Imax、Imid、IminAs CH、CW、CCInitial mean vector mu ofH、μW、μC；
S37), calculating each sample and mu in the sample setH、μW、μCEuclidean distance between:
s38), putting each sample in the sample set into the data cluster closest to the sample set, and counting the sample data n contained in each data clusterjCalculate a new mean vector μ'H、μ'W、μ'C：
S39), check μH、μW、μCMu's'H、μ'W、μ'CIf so, finishing the cold and hot data classification operation, and taking the current result as the final cold and hot data classification result; if not, the mu is processedH、μW、μCAre respectively replaced by mu'H、μ'W、μ'CReturning to step S37.
4. The SSD cold and hot data separation method based on out-of-band interaction of claim 3, wherein: allocating a cache for storing all LBA access information in the host, initializing the LBA access information cache before the host side receives the SSD equipment host access information, and recording the write operation times N of each LBA by taking the LBA serial number as an indexwrEarliest write operation timeT1stLatest write operation time TlastFrequency of writing Fwr；
The host end receives the LBA access information data sent by the SSD equipment through an out-of-band channel, and executes the following steps on each record in the received data:
A. reading the serial number i of the LBA in the record, and accessing the N of the ith LBA in the information cache for the local LBAwrExecuting an adding 1 operation;
B. reading the time stamp T in the record, and checking the T of the iLBA in the local LBA access information cache1stIf it is 0, if so, then T is1stSetting t, if not, not processing;
C. caching the I-th LBA T in the local LBA access informationlastSetting as t;
D. f for calculating ith LBA in local LBA access information cachewr：
5. The SSD cold and hot data separation method based on out-of-band interaction of claim 1, wherein: only distributing a buffer with limited capacity to the LBA access information in the SSD equipment; when the host executes write operation on a certain LBA, the LBA value and the current timestamp are recorded in the SSD device cache, and when the data volume of the LBA access information in the cache reaches a threshold value, all data in the cache is uploaded to the host end through an out-of-band channel.
6. The SSD cold and hot data separation method based on out-of-band interaction of claim 1, wherein: the out-of-band management path is in a full-duplex data mode and is realized based on an SMBus/IIC interface, and the specific realization mode is as follows: the host and the SSD device are both provided with 2 SMBus/IIC interfaces, one SMBus/IIC interface serves as a Master end and sends request messages and data to the opposite-end device, and the other SMBus/IIC interface serves as a Slave end and receives completion messages sent by the opposite-end device.
7. The SSD cold and hot data separation method based on out-of-band interaction of claim 3, wherein: when the host side sends a cold and hot data separation request, the host selects two types with small data quantity from the three types of data clusters, and sends the LBA and the temperature type contained in the two types of data clusters to the SSD side, so that the physical separation of the cold and hot data is completed at the SSD side.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010644277.3A CN111831228A (en) | 2020-07-06 | 2020-07-06 | SSD (solid State disk) cold and hot data separation method based on out-of-band interaction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010644277.3A CN111831228A (en) | 2020-07-06 | 2020-07-06 | SSD (solid State disk) cold and hot data separation method based on out-of-band interaction |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111831228A true CN111831228A (en) | 2020-10-27 |
Family
ID=72901211
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010644277.3A Pending CN111831228A (en) | 2020-07-06 | 2020-07-06 | SSD (solid State disk) cold and hot data separation method based on out-of-band interaction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111831228A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112346660A (en) * | 2020-11-13 | 2021-02-09 | 北京泽石科技有限公司 | Data storage method and device and data reading method and device |
| CN112346660B (en) * | 2020-11-13 | 2024-05-14 | 北京泽石科技有限公司 | Data storage method and device, and data reading method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105677245A (en) * | 2015-12-31 | 2016-06-15 | 记忆科技(深圳)有限公司 | Method for prolonging SSD service life based on WL multithreads |
| CN105677242A (en) * | 2015-12-31 | 2016-06-15 | 杭州华为数字技术有限公司 | Hot and cold data separation method and device |
| US9747202B1 (en) * | 2013-03-14 | 2017-08-29 | Sandisk Technologies Llc | Storage module and method for identifying hot and cold data |
| CN109189693A (en) * | 2018-07-18 | 2019-01-11 | 深圳大普微电子科技有限公司 | The method and SSD that a kind of pair of LBA information is predicted |
| CN110515550A (en) * | 2019-08-21 | 2019-11-29 | 深圳忆联信息系统有限公司 | A kind of method and device thereof of the cold and hot data separating of SATA solid state hard disk |
-
2020
- 2020-07-06 CN CN202010644277.3A patent/CN111831228A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9747202B1 (en) * | 2013-03-14 | 2017-08-29 | Sandisk Technologies Llc | Storage module and method for identifying hot and cold data |
| CN105677245A (en) * | 2015-12-31 | 2016-06-15 | 记忆科技(深圳)有限公司 | Method for prolonging SSD service life based on WL multithreads |
| CN105677242A (en) * | 2015-12-31 | 2016-06-15 | 杭州华为数字技术有限公司 | Hot and cold data separation method and device |
| CN109189693A (en) * | 2018-07-18 | 2019-01-11 | 深圳大普微电子科技有限公司 | The method and SSD that a kind of pair of LBA information is predicted |
| WO2020015550A1 (en) * | 2018-07-18 | 2020-01-23 | 深圳大普微电子科技有限公司 | Method for predicting lba information, and ssd |
| CN110515550A (en) * | 2019-08-21 | 2019-11-29 | 深圳忆联信息系统有限公司 | A kind of method and device thereof of the cold and hot data separating of SATA solid state hard disk |
Non-Patent Citations (1)
| Title |
|---|
| 赵泽亚: "《海量遥感数据的存储迁移策略研究》", 《信息工程大学学报 》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112346660A (en) * | 2020-11-13 | 2021-02-09 | 北京泽石科技有限公司 | Data storage method and device and data reading method and device |
| CN112346660B (en) * | 2020-11-13 | 2024-05-14 | 北京泽石科技有限公司 | Data storage method and device, and data reading method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11216185B2 (en) | Memory system and method of controlling memory system | |
| US11068165B2 (en) | Non-volatile memory data write management | |
| WO2017000658A1 (en) | Storage system, storage management device, storage device, hybrid storage device, and storage management method | |
| US20140337566A1 (en) | Solid state memory (ssm), computer system including an ssm, and method of operating an ssm | |
| US20130332693A1 (en) | Allocating storage memory based on future file size or use estimates | |
| US11042305B2 (en) | Memory system and method for controlling nonvolatile memory | |
| US10430329B2 (en) | Quality of service aware storage class memory/NAND flash hybrid solid state drive | |
| JP2020529679A (en) | Extending SSD lifespan | |
| CN111796765B (en) | Method for optimizing performance of data storage system | |
| US20160342364A1 (en) | Tiered storage system, storage controller and data location estimation method | |
| CN105354152A (en) | Nonvolatile memory and wear leveling method | |
| CN110377233A (en) | SSD reading performance optimization method, device, computer equipment and storage medium | |
| WO2021035555A1 (en) | Data storage method and apparatus for solid state disk and solid state disk (ssd) | |
| US11656778B2 (en) | Selection of block size for namespace management in non-volatile memory devices | |
| US9632950B2 (en) | Storage control device controlling access to a storage device and access controlling method of a storage device | |
| CN114253461A (en) | Mixed channel memory device | |
| US20240061782A1 (en) | Method and device for data caching | |
| CN111831228A (en) | SSD (solid State disk) cold and hot data separation method based on out-of-band interaction | |
| CN115993930A (en) | System, method and apparatus for in-order access to data in block modification memory | |
| CN114327272A (en) | Data processing method, solid state disk controller and solid state disk | |
| CN116235138A (en) | Data reading method and related device applied to solid state disk SSD | |
| CN110968520B (en) | Multi-stream storage device based on unified cache architecture | |
| CN117992360A (en) | Storage system and storage method | |
| CN115989485A (en) | Data processing method, device and system | |
| KR20230172426A (en) | Method and device for data storage |
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 |