CN103473185B - Method, buffer storage and the storage system of caching write - Google Patents
Method, buffer storage and the storage system of caching write Download PDFInfo
- Publication number
- CN103473185B CN103473185B CN201310402988.XA CN201310402988A CN103473185B CN 103473185 B CN103473185 B CN 103473185B CN 201310402988 A CN201310402988 A CN 201310402988A CN 103473185 B CN103473185 B CN 103473185B
- Authority
- CN
- China
- Prior art keywords
- metadata
- write
- data block
- group
- flash
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Techniques For Improving Reliability Of Storages (AREA)
Abstract
Embodiments providing and a kind of cache the method for write, buffer storage and storage system, the method includes: update at least one metadata in internal memory；Obtain described at least one metadata group belonging at least one metadata；If the write state of the first metadata group at least one metadata group described is multiple write I/O Request orders of the pending metadata write operation such as existence under write-not state and described first metadata group, then the plurality of write I/O Request order is moved to the merging write metadata queue of described first metadata group；The metadata of described first metadata group is written to described in described Flash in the meta data block corresponding to the first metadata group；Described first metadata group is merged the multiple write I/O Request orders under write metadata queue and returns described first metadata group write the most successfully state.The methods, devices and systems of the embodiment of the present invention can reduce the write number of times of the metadata to Flash to a certain extent.
Description
Technical field
The present embodiments relate to field of data storage, and cache write more particularly, to a kind of
Method, buffer storage and storage system.
Background technology
Industry uses prior art now, due to the storage medium of caching (Cache) be based on
NVRAM or NVDIMM, these storage medium characteristics can well meet the little IO's of system
Performance, especially for metadata.
Being typically of size of between 30B-64B of a usual metadata.When the storage medium of Cache uses
During Flash, due to the safety of data to be ensured, write request is required for the IO of 2 times each time.And Flash
Life-span with the number of times write directly related, if each metadata is according to the direct IO of the size of metadata,
Then can not give full play to the performance of Flash, if IO under the size alignd according to 4K of metadata, then may be used
The problem writing amplification can be there is, the life-span of whole Flash can greatly be affected.
Summary of the invention
The embodiment of the present invention provides a kind of and caches the method for write, buffer storage and storage system, it is possible to one
Determine in degree, to reduce the write number of times of the metadata to Flash, improve the service life of Flash.
First aspect, it is provided that a kind of method caching Cache write, the flash memory Flash of this Cache
Memory area include multiple chunk, this chunk includes a chunk meta-data region and one
Chunk data field, this chunk meta-data region includes at least one meta data block, this chunk
Data field includes at least one data field corresponding with this at least one meta data block, and this meta data block includes
Multiple metadata, this data field includes multiple data block, this meta data block and this data field one_to_one corresponding,
This metadata and this data block one_to_one corresponding, a metadata group in internal memory corresponds to this metadata
Block, this metadata group comprises multiple metadata, the metadata in this metadata group and this yuan of number
According to the metadata one_to_one corresponding in the meta data block that group is corresponding, the method includes: in renewal internal memory at least
One metadata, this at least one metadata is corresponding to being used in this Flash storing write I/O Request order
At least one data block of the data of request write；Obtain at least one belonging to this at least one metadata
Metadata group；If the write state of the first metadata group in this at least one metadata group is write-not shape
Multiple write I/O Request life of the pending metadata write operation such as existence under state and this first metadata group
Order, then move to the merging write metadata of this first metadata group by the plurality of write I/O Request order
Queue；The metadata of this first metadata group is written in this Flash corresponding to this first metadata group
Meta data block in；This first metadata group is merged the multiple write IO under write metadata queue
Request command returns this first metadata group write the most successfully state.
In conjunction with first aspect, in the implementation that the first is possible, update at least one unit in internal memory
Data are implemented as: if the status indication information of one of this at least one metadata indicate this at least one
One of individual metadata is clean Clean state, then update one of this at least one metadata, wherein, be somebody's turn to do
Clean state is for indicating one of this at least one metadata in metadata corresponding for this Flash with this extremely
One of few metadata content is identical.
In conjunction with the first possible implementation of first aspect or first aspect, in the reality that the second is possible
In existing mode, it is implemented as: the size of this meta data block page-size equal to this Flash.
In conjunction with the implementation that the second of first aspect is possible, in the implementation that the third is possible,
It is implemented as: the size of this metadata is 32B, 64B or 128B.
The third possible reality of the implementation possible in conjunction with the second of first aspect or first aspect
Existing mode, in the 4th kind of possible implementation, is implemented as: this chunk meta-data region big
Little in the same size with this data block.
The third possible reality of the implementation possible in conjunction with the second of first aspect or first aspect
Existing mode, in the 5th kind of possible implementation, is implemented as: the size of this data block is
512B, 1K, 2K, 4K, 8K, 16K, 32K or 64K.
Second aspect, it is proposed that a kind of buffer storage, the memory area of the flash memory Flash of this buffer storage
A chunk meta-data region and chunk data are included including multiple chunk, this chunk
District, this chunk meta-data region includes at least one meta data block, this chunk data field bag
Including at least one data field corresponding with this at least one meta data block, this meta data block includes multiple units number
According to, this data field includes multiple data block, this meta data block and this data field one_to_one corresponding, this metadata
With this data block one_to_one corresponding, a metadata group in internal memory correspond to this meta data block, one
This metadata group comprises multiple metadata, and the metadata in this metadata group is corresponding with this metadata group
Meta data block in metadata one_to_one corresponding, this buffer storage includes: updating block, in being used for updating
At least one metadata in depositing, this at least one metadata is used for storing write corresponding in this Flash
At least one data block of the data of I/O Request command request write；Acquiring unit, is used for obtaining this extremely
The first metadata group of few at least one meta data block belonging to a metadata, in this metadata group
Metadata corresponding to the metadata in a meta data block in this Flash；Queue mobile unit, uses
If the write state of the first metadata group in this at least one metadata group is write-not state and is somebody's turn to do
Multiple write I/O Request orders of the pending metadata write operation such as existence under first metadata group, then
The plurality of write I/O Request order is moved to the merging write metadata queue of this first metadata group；
Writing unit, for being written to this first metadata group in this Flash corresponding to this first metadata group
Meta data block；Feedback unit, for merging this first metadata group under write metadata queue
Multiple write I/O Request orders return this first metadata group write the most successfully state.
In conjunction with second aspect, in the implementation that the first is possible, this updating block is implemented as:
If the status indication information of one of this at least one metadata indicates one of this at least one metadata to be
Clean Clean state, then update one of this at least one metadata, and wherein, this Clean state is used for
Indicate one of this at least one metadata metadata corresponding for this Flash and this at least one metadata it
One content is identical.
In conjunction with the first possible implementation of second aspect or second aspect, in the reality that the second is possible
In existing mode, it is implemented as: the size of this meta data block page-size equal to this Flash.
The first possible implementation or the second of second aspect in conjunction with second aspect or second aspect
Plant possible implementation, in the implementation that the third is possible, be implemented as: this chunk unit
The size of data field is in the same size with this data block.
The third aspect, it is proposed that a kind of storage system, including internal memory, caching Cache and processor, should
Cache includes flash memory Flash, and the memory area of this Flash includes multiple chunk, this chunk
Including a chunk meta-data region and a chunk data field, this chunk meta-data region includes
At least one meta data block, this chunk data field includes corresponding with this at least one meta data block
At least one data field, this meta data block includes multiple metadata, and this data field includes multiple data block,
This meta data block and this data field one_to_one corresponding, this metadata and this data block one_to_one corresponding, in internal memory
One metadata group is corresponding to this meta data block, and this metadata group comprises multiple metadata, and one
Metadata in the meta data block that metadata in this metadata group individual is corresponding with this metadata group one a pair
Should, wherein, this processor is used for: update at least one metadata in this internal memory, this at least one unit
Data corresponding in this Flash for storing at least one of the data of write I/O Request command request write
Data block；Obtain the first metadata group of at least one meta data block belonging to this at least one metadata,
Metadata in one this metadata group is corresponding to the metadata in a meta data block in this Flash；
If the write state of the first metadata group in this at least one metadata group be write-not state and this
Multiple write I/O Request orders of the pending metadata write operation such as existence under unitary data set, then will
The plurality of write I/O Request order moves to the merging write metadata queue of this first metadata group；Will
This first metadata group is written in this Flash this meta data block corresponding to the first metadata group；To this
First metadata group merge write metadata queue under multiple write I/O Request orders return this first
Metadata group write the most successfully state.
In conjunction with the third aspect, in the implementation that the first is possible, it is being used for updating in this internal memory extremely
A few metadata, this processor is implemented as: if the state of one of this at least one metadata refers to
Show that information indicates one of this at least one metadata for clean Clean state, then update this at least one unit
One of data, wherein, this Clean state is used for indicating one of this at least one metadata at this Flash
One of corresponding at least one metadata of metadata and this content is identical.
In conjunction with the first possible implementation of the third aspect or the third aspect, in the reality that the second is possible
In existing mode, it is implemented as: the size of this meta data block page-size equal to this Flash.
The first possible implementation or the second of the third aspect in conjunction with the third aspect or the third aspect
Plant possible implementation, in the implementation that the third is possible, be implemented as: this chunk unit
The size of data field is in the same size with this data block.
Based on above technical scheme, method, buffer storage and the storage of the caching write of the embodiment of the present invention
System, by merging multiple write I/O Request orders write operation to same meta data block, Ke Yiyi
Determine reduce the write number of times of Flash in degree and avoid IO to write the problem of amplification to a certain extent, it is thus possible to
Enough improve the service life of Flash.
Accompanying drawing explanation
In order to be illustrated more clearly that the technical scheme of the embodiment of the present invention, below will be to embodiment or existing skill
In art description, the required accompanying drawing used is briefly described, it should be apparent that, the accompanying drawing in describing below
It is only some embodiments of the present invention, for those of ordinary skill in the art, is not paying creation
Property work on the premise of, it is also possible to obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is a kind of data layout schematic diagram of embodiment of the present invention Flash.
Fig. 2 is the method flow diagram of a kind of Cache of embodiment of the present invention write.
Fig. 3 is the another kind of data layout schematic diagram of embodiment of the present invention Flash.
Fig. 4 is another data layout's schematic diagram of embodiment of the present invention Flash.
Fig. 5 is another data layout's schematic diagram of embodiment of the present invention Flash.
Fig. 6 is embodiment of the present invention metadata position of actual persistence in position and Flash in internal memory
A kind of relation schematic diagram.
Fig. 7 is that embodiment of the present invention buffer storage merges the particular flow sheet of write to metadata.
Fig. 8 is the structural representation of embodiment of the present invention buffer storage 800.
Fig. 9 is the structural representation of embodiment of the present invention storage system 900.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out
Clearly and completely describe, it is clear that described embodiment be a part of embodiment of the present invention rather than
Whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making wound
The every other embodiment obtained under the property made work premise, broadly falls into the scope of protection of the invention.
Fig. 1 is a kind of data layout schematic diagram of embodiment of the present invention flash memory (Flash).As it is shown in figure 1,
The memory area of Flash can include superblock, meta-data region and data field three part.
Superblock, the global data in memory buffers (Cache), mainly for system is delayed after machine
The data needed during recovery.Wherein, this global data can include the size of meta data block, data block
Size etc..
Metadata (metadata), safeguards the data of data block in storage system, is used for describing and organizing
The relation of data block, its size determines according to the data structure size defined in system.
Meta-data region, refers to the region for storing metadata.
Data field, in memory buffers (Cache), real data, can include multiple data under it
Block.
Fig. 3 is the another kind of data layout schematic diagram of embodiment of the present invention Flash.As it is shown on figure 3, Flash
Under include superblock and multiple memory block (chunk), each chunk includes a chunk meta-data region
With a chunk data field, each chunk meta-data region includes one or more meta data block, each
Chunk data field includes the data field corresponding with these one or more meta data block.Each chunk data
Meta data block one_to_one corresponding under the chunk meta-data region that data field under district is corresponding.One first number
Safeguard a corresponding data field according to block, specifically, include multiple metadata under each meta-data region, often
Individual metadata is corresponding to a data block under data field.
In Fig. 3, although 1 chunk meta-data region contains 4 meta data block, it is not intended that
1 chunk meta-data region can only comprise 4 meta data block, it is also possible to is 1,2 or other number
Value.Same, the number of the chunk shown in Fig. 3, the number of data field, unit in chunk data field
In data field, in the number of metadata, data field, the number of data block is not limited to the numerical value shown in Fig. 3.
Each data field corresponds to a different meta data block, and the metadata institute in a meta data block is right
One data field of data block composition answered.It is to say, a data field comprises a meta data block size
The data corresponding to metadata, the size of a data field is by the number of metadata in a meta data block
Determine with the size of a data block.
In Flash, the size of a page write is 4K.Therefore, a kind of preferably scheme, one
The size of meta data block is 4K.The size that might as well assume a metadata is 32B, a data block
Size is 16K, then the size of a data field is 2M(16K*(4K/32B)).
Fig. 2 is the method flow diagram of a kind of Cache of embodiment of the present invention write.The method of Fig. 2 is by caching
Device performs.Wherein, the memory area of the Flash of buffer storage includes multiple chunk, this chunk
Including a chunk meta-data region and a chunk data field, this chunk meta-data region includes
At least one meta data block, this chunk data field includes corresponding with this at least one meta data block
At least one data field, this meta data block includes multiple metadata, and this data field includes multiple data block,
This meta data block and this data field one_to_one corresponding, this metadata and this data block one_to_one corresponding, in internal memory
One metadata group is corresponding to this meta data block, and this metadata group comprises multiple metadata, and one
Metadata in the meta data block that metadata in this metadata group individual is corresponding with this metadata group one a pair
Should.The method includes:
201, update at least one metadata in internal memory.
Wherein, this at least one metadata please corresponding to being used in this Flash storing write I/O Request order
Seek at least one data block of the data of write.
In other words, this at least one metadata is corresponding at least one data block in this Flash, this Flash
In at least one data block for store write I/O Request command request write data.
202, obtain at least one the metadata group belonging to this at least one metadata.
During the present invention implements, obtain a metadata group, can be by obtaining the mark of metadata group, internal memory
The mode such as address or variable realizes, and implements and refers to prior art, and the embodiment of the present invention is at this no longer
Repeat.
Metadata in internal memory, can be grouped according to the size of the meta data block in Flash.Such as,
In Flash, a meta data block has 128 metadata, then can 128 metadata be compiled in internal memory is 1
Group.Meanwhile, the metadata of a metadata group and the metadata of a meta data block in Flash in internal memory
One_to_one corresponding.
203, if the write state of the first metadata group in this at least one metadata group is write-not shape
Multiple write I/O Request life of the pending metadata write operation such as existence under state and this first metadata group
Order, then move to the merging write metadata of this first metadata group by the plurality of write I/O Request order
Queue.
The first metadata group that in the present invention, arbitrary place mentions, all refers in this at least one metadata group
Any piece of metadata group, comprises any piece of metadata group of more new metadata in other words.
In the embodiment of the present invention, the write state of the first metadata group is used for marking whether to first yuan of number
Write operation is carried out in the meta data block corresponding to Flash according to group.
If the first metadata group is carrying out write operation is in write state in other words, then complete at write operation
The write I/O Request order needs relating to the first metadata group write operation before one-tenth wait.
If the first metadata group write operation is complete or is not carried out write operation, it is in write-not shape in other words
State, then can be by multiple write IO of the pending metadata write operations such as existence under the first metadata group
Request command moves in the merging write metadata queue of the first metadata group.Now, the first metadata
Group is in write state, can perform write operation.
204, this first metadata group is written in this Flash the first number corresponding to this first metadata group
According to block.
205, this first metadata group is merged the multiple write I/O Request life under write metadata queue
Order returns this first metadata group write the most successfully state.
In the embodiment of the present invention, when writing metadata to the Flash of Cache, merge multiple write IO
The request command write operation to same meta data block, can reduce the write time of Flash to a certain extent
Number also avoids IO to write the problem of amplification to a certain extent, improves the service life of Flash.
Alternatively, as an embodiment, step 201 specifically can be embodied as: if this at least one unit
The status indication information of one of data indicate one of this at least one metadata for clean Clean state, then
Update one of this at least one metadata.Wherein, this Clean state is used for indicating this at least one unit's number
According to one of identical in one of at least one metadata of metadata and this corresponding for this Flash content.The present invention
In embodiment, to having put dirty metadata, no longer carry out write operation, reduce the most to a certain extent
The write number of times of Flash, improves the service life of Flash.
Alternatively, after step 205, the method also includes: if the first metadata group there is also
The write I/O Request order of pending metadata write operation, then will wait pending metadata write operation
Write I/O Request order move to this first metadata group merging write metadata queue and to this
Unitary data set performs metadata write operation.Write if there is no pending metadata write operation
I/O Request order, the most now can be set to write-not state by the write state of the first metadata group.
Alternatively, the size of this meta data block page-size equal to this Flash.Meta data block
Size is consistent with the page-size of Flash, it is possible to improves the life-span of Flash to a certain extent, and improves
Flash write performance.In prior art, the page-size of Flash is 4K.
Alternatively, the size of this metadata is 32B, 64B or 128B.
Alternatively, as an embodiment, the size of this chunk meta-data region and this data block
In the same size.The size of Chunk meta-data region is in the same size with a data block so that buffer storage
When loading Flash or Flash being initialized, memory area can be carried out according to the size of data block and draw
Point.
Alternatively, the size of this data block can be 512B, 1K, 2K, 4K, 8K, 16K,
32K or 64K.
Below, in conjunction with specific example, the method for the embodiment of the present invention is described.
As it is shown on figure 3, include superblock and multiple chunk under Flash, each chunk includes one
Chunk meta-data region and a chunk data field, each chunk meta-data region includes one or more
Meta data block, each chunk data field includes the data field corresponding with these one or more meta data block.
Meta data block under the chunk meta-data region that data field under each chunk data field is corresponding is one by one
Corresponding.
Fig. 4 is another data layout's schematic diagram of embodiment of the present invention Flash.The embodiment of the present invention
A kind of layout type as shown in Figure 4, comprises 1 meta data block in a chunk meta-data region, corresponding
Ground, a chunk data field comprises 1 data field.
Fig. 5 is another data layout's schematic diagram of embodiment of the present invention Flash.The embodiment of the present invention
A kind of layout type is as it is shown in figure 5, comprise 4 meta data block in a chunk meta-data region, corresponding
Ground, a chunk data field comprises 4 data fields.
From Fig. 4 and Fig. 5, a chunk meta-data region potentially includes one or more meta data block,
The size of one chunk meta-data region is variable.A kind of preferably scheme, can be by a data block
Size is as the size of a chunk meta-data region, so, can make when Flash loads and initializes,
According to the size of data block, the memory area of Flash is divided.Such as, the size of a data block
For 4K, then the size of a chunk meta-data region is also 4K, and the number of its meta data block is 1
(4K/4K), as shown in Figure 4.The most such as, the size of a data block is 16K, then a chunk
The size of meta-data region is also 16K, and the number of its meta data block is 4(16K/4K).The size of data block
Can be by operating system or storage device configuration, its size can be 4K, 8K, 16K, 32K or 64K
Etc., this is not restricted for the embodiment of the present invention.
Certainly, the size of data block can also be with the varying in size of chunk meta-data region.Such as, data
The size of block can also is that 512B, 1K, 2K etc., and at least includes 1 unit in chunk meta-data region
Data block, it is clear that bigger than a data block.The most such as, it is 8K, 16K, 32K when the size of data block
Or during 64K etc., the size of chunk meta-data region can also be less than the size of data block, such as chunk
Meta-data region only includes 1 meta data block.
Fig. 6 is embodiment of the present invention metadata position of actual persistence in position and Flash in internal memory
A kind of relation schematic diagram.
As shown in Figure 6, a data structure p_head array, the data of storage can be designed in internal memory
The metadata that in district, each data block is corresponding.The size of metadata required for each data block might as well be assumed
It is 64 bytes, then p_head array can be divided into one according to 64 (4K/64B=64) metadata
Group, a metadata group is corresponding to a meta data block in Flash.
In internal memory, metadata is corresponding to an element of p_head array, its data structure
FlashCache_wbhead and implication can be as follows:
As shown in Figure 6, also can design a data structure metadata_head array in internal memory, dimension
Protect in the metadata of data in EMS memory and these metadata reality in actual persistence equipment (Flash)
The relation of storage position (i.e. corresponding meta data block).
The metadata group that a meta data block in Flash is corresponding in internal memory, corresponds to
An element in metadata_head array, its data structure metadata_head and implication can be as follows
Shown in:
As shown in Figure 6, internal memory is store the metadata of a data block by flashCache_wbhead
Information, is store again the information of same group metadata (i.e. one meta data block) by metadata_head,
Write according to meta data block when write to facilitate.
By introducing the associated maintenance information of metadata group corresponding to meta data block in internal memory, for caching dress
Put realize meta data block rank multiple write I/O Request Command combination write provide possibility.
Fig. 7 is that embodiment of the present invention buffer storage merges the particular flow sheet of write to metadata.
Below, the method for Fig. 7 is described by the data structure in conjunction with Fig. 6 and metadata, meta data block.
Certainly, the method for Fig. 7 applies also for other Flash data layout type comprising multiple meta-data region,
This is not restricted for the embodiment of the present invention.
701, receive write I/O Request order.
Buffer storage can receive write I/O Request order.
702, store data into Flash.
Buffer storage can be according to write I/O Request order, and the data being written into store buffer storage
In at least one data block of Flash.
Step 701,702 implement refer to prior art realize, the embodiment of the present invention is at this not
Repeat again.
703, check that metadata is the need of renewal.
Buffer storage can according to the position of at least one data block of storage in Flash, obtain this at least one
The metadata that individual data block is corresponding in internal memory.
Buffer storage can be according to the amendment condition adjudgement metadata of current meta data the need of renewal.This yuan
The amendment state of data, it may include dirty (Dirty) state and clean (Clean) state.Wherein, Dirty
State can represent that this metadata is revised relative to metadata corresponding in Flash, Clean state representation
This metadata is revised relative to metadata corresponding in Flash.
Buffer storage can judge that the metadata corresponding to data block is the need of renewal one by one, it is possible to batch is sentenced
Disconnected metadata corresponding to data block is the need of renewal.
If metadata corresponding to this at least one data block has put Dirty, after the most again putting Dirty
Effect is the same, it is not necessary to again put Dirty.Now, step 704 is performed.
If being divided into Clean state in the middle part of the metadata that this at least one data block is corresponding, then to this part
Metadata for Clean state performs step 705.
To the amendment condition adjudgement of metadata to decide whether to be updated, internal memory can be reduced to a certain extent
The middle metadata needing to update, and then reduce the write number of times of metadata in Flash, improve Flash's
Life-span.
Certainly, buffer storage also can not judge that this at least one data block metadata of correspondence in internal memory is worked as
Front amendment state is write direct, and does not affect the execution of embodiment of the present invention Cache wiring method.
704, write I/O Request order is returned and is write as merit, and to revise metadata write state be non-write shape
State.
Owing to metadata is made without updating, after data write Flash, this I/O Request can terminate.
Buffer storage can return to this write I/O Request order and be write as merit, and labelling metadata write state is non-
Write state.
705, update the metadata in internal memory.
If being divided into Clean state in the middle part of the metadata that this at least one data block is corresponding, buffer storage can
The metadata that this amendment state is Clean state is put Dirty.
Certainly, also it is not excluded for buffer storage and the metadata that amendment state is Dirty state is put Dirty.Example
As, in the middle part of multiple continuous print metadata, it being divided into Clean state, part is Dirty state, can be to this
Multiple continuous print metadata put Dirty simultaneously.
706, according at least one unit at the metadata updated position acquisition metadata place in internal memory
Data set.
The metadata place that buffer storage can update according to the metadata updated position acquisition in internal memory
Metadata group.
Such as, in figure 6, according to metadata position (index) in internal memory, buffer storage can obtain
Take the meta_head of metadata group corresponding to this metadata.The unit required for each data block might as well be assumed
The size of data is 64 bytes, and a metadata group (meta_head) just has 64 (4K/64B=64)
Metadata.If the metadata of amendment is 33-120, then the metadata group (meta_head) of its correspondence
It is the 0th, 1 totally 2 metadata groups.
707, the first metadata group that write I/O Request order is linked in this at least one metadata group
In metadata queue to be combined.
After getting this at least one the metadata group needing to perform metadata write operation, buffer storage
Write I/O Request order is linked into the unit to be combined number of the first metadata group in this at least one metadata group
According to queue.
Wherein, any piece of metadata group during this first metadata group refers to this at least one metadata group.Example
If, this first metadata group can be the 0th metadata group of step 706 or the 1st metadata group.
708, check the first metadata group the most just at write operation.
What buffer storage judged the first metadata group the most carries out write operation.
As shown in Figure 6, buffer storage can judge according to the metadata_head that the first metadata group is corresponding
Whether the write state of the first metadata group is write-not state.
The write state of the first metadata group is used for marking whether right in Flash institute to the first metadata group
The meta data block answered carries out write operation.
If the first metadata group is carrying out write operation is in write state in other words, then complete at write operation
The write I/O Request order needs relating to the first metadata group write operation before one-tenth wait.
If the first metadata group write operation is complete or is not carried out write operation, it is in write-not shape in other words
State, then can be by multiple write IO of the pending metadata write operations such as existence under the first metadata group
Request command moves in the merging write metadata queue of the first metadata group.Now, the first metadata
Group is in write state, can perform write operation.
It is to say, if write state, then perform step 709；If write-not state, then
Perform step 710.
709, return, wait that write operation completes.
710, the write I/O Request order in the metadata queue to be combined of the first metadata group is moved on to
The merging write metadata queue of unitary data set.
After once writing in the first metadata group, buffer storage is to be combined by the first metadata group
Multiple write I/O Request orders in metadata queue move on to the merging write metadata of the first metadata group
Queue.
711, the first metadata is submitted to write request operation.
Buffer storage closes and operates the first metadata group asynchronous submission write request, the first metadata group write
In the meta data block that Flash is corresponding.
712, whether metadata writes successfully.
Buffer storage judges whether metadata writes successfully.
If it does not, perform step 713.
If it is, perform step 714.
713, the write I/O Request life merged in write metadata queue of circular treatment the first metadata group
Order, returns the state writing failure respectively, terminates.
714, the write I/O Request life merged in write metadata queue of circular treatment the first metadata group
Order, returns respectively and writes successful state.
715, check whether the metadata queue to be combined of the first metadata group is empty.
If it does not, perform step 710.
If it is, perform step 716.
716, the write state revising the first metadata group is non-write state.
It is non-write state that buffer storage revises the write state of the first metadata group.
Fig. 8 is the structural representation of embodiment of the present invention buffer storage 800.Wherein, this buffer storage 800
The memory area of Flash include that multiple chunk, this chunk include a chunk meta-data region
With a chunk data field, this chunk meta-data region includes at least one meta data block, and this is one years old
Individual chunk data field includes at least one data field corresponding with this at least one meta data block, this yuan of number
Including multiple metadata according to block, this data field includes multiple data block, this meta data block and this data field one
One is corresponding, this metadata and this data block one_to_one corresponding, and a metadata group in internal memory corresponds to one
This meta data block, this metadata group comprises multiple metadata, the metadata in this metadata group
Metadata one_to_one corresponding in the meta data block corresponding with this metadata group.This buffer storage 800 can include
Updating block 801, acquiring unit 802, queue mobile unit 803, writing unit 804 and feedback unit
805。
Updating block 801, for updating at least one metadata in internal memory.
Wherein, this at least one metadata please corresponding to being used in this Flash storing write I/O Request order
Seek at least one data block of the data of write.
In other words, this at least one metadata is corresponding at least one data block in this Flash, this Flash
In at least one data block for store write I/O Request command request write data.
Acquiring unit 802, for obtaining at least one the metadata group belonging to this at least one metadata.
Metadata in internal memory, can be grouped according to the size of the meta data block in Flash.Such as,
In Flash, a meta data block has 128 metadata, then can 128 metadata be compiled in internal memory is 1
Group.Meanwhile, the metadata of a metadata group and the metadata of a meta data block in Flash in internal memory
One_to_one corresponding.
Queue mobile unit 803, if the first metadata group in this at least one metadata group
Write state is the pending metadata write operation such as existence under write-not state and this first metadata group
Multiple write I/O Request orders, then move to this first metadata group by the plurality of write I/O Request order
Merging write metadata queue.
The first metadata group that in the present invention, arbitrary place mentions, all refers in this at least one metadata group
Any piece of metadata group, comprises any piece of metadata group of more new metadata in other words.
In the embodiment of the present invention, the write state of the first metadata group is used for marking whether to first yuan of number
Write operation is carried out in the meta data block corresponding to Flash according to group.
If carrying out write operation to be in write state in other words, then before write operation completes, relate to
The write I/O Request order of unitary data set write operation needs to wait.
If write operation is complete or is not carried out write operation, be in write-not state in other words, then can be by
Under unitary data set, multiple write I/O Request orders of the pending metadata write operation such as existence move to
In the merging write metadata queue of the first metadata group.Now, the first metadata group is in write state,
Write operation can be performed.
Writing unit 804, for being written in this Flash this first metadata by this first metadata group
Meta data block corresponding to group.
Feedback unit 805, for merging write under metadata queue multiple to this first metadata group
Write I/O Request order returns this first metadata group write the most successfully state.
In the embodiment of the present invention, when writing metadata to the Flash of buffer storage 800, merge multiple
The write I/O Request order write operation to same meta data block, can reduce Flash to a certain extent
Write number of times and avoid IO to write the problem of amplification to a certain extent, the service life of raising Flash.
Alternatively, as an embodiment, updating block 801 specifically for: if this at least one unit
The status indication information of one of data indicate one of this at least one metadata for clean Clean state, then
Updating one of this at least one metadata, wherein, this Clean state is used for indicating this at least one unit's number
According to one of identical in one of at least one metadata of metadata and this corresponding for this Flash content.The present invention
In embodiment, buffer storage 800, to having put dirty metadata, no longer carries out write operation, the most necessarily
Reduce the write number of times of Flash in degree, improve the service life of Flash.
Alternatively, alternatively, the size of this meta data block page-size equal to this Flash.Unit
The size of data block is consistent with the page-size of Flash, it is possible to improve the life-span of Flash to a certain extent,
And improve Flash write performance.In prior art, the page-size of Flash is 4K.
Alternatively, the size of this metadata is 32B, 64B or 128B.
Alternatively, as an embodiment, the size of this chunk meta-data region and this data block
In the same size.The size of Chunk meta-data region is in the same size with a data block so that buffer storage
800, when loading Flash or initializing Flash, can carry out memory area according to the size of data block
Divide.
Alternatively, the size of this data block can be 512B, 1K, 2K, 4K, 8K, 16K,
32K or 64K.
It addition, the method that buffer storage 800 can also carry out Fig. 2, and realize buffer storage shown in Fig. 7
The function of embodiment, the embodiment of the present invention does not repeats them here.
Fig. 9 is the structural representation of embodiment of the present invention storage system 900.This storage system 900 can be wrapped
Include: memorizer 901 and processor 902.Memorizer 901 can include internal memory 903, caching 904 etc..
Memorizer 901 and processor 902 are connected with each other by bus 905 system；Bus 904 can be
Isa bus, pci bus or eisa bus etc..Described bus can be divided into address bus, data total
Line, control bus etc..For ease of representing, Fig. 9 only represents by a two-way block arrow, but not
Represent and only have a bus or a type of bus.
Caching 904, the memory area including Flash, this Flash includes multiple chunk, this chunk
Including a chunk meta-data region and a chunk data field, this chunk meta-data region includes
At least one meta data block, this chunk data field includes corresponding with this at least one meta data block
At least one data field, this meta data block includes multiple metadata, and this data field includes multiple data block,
This meta data block and this data field one_to_one corresponding, this metadata and this data block one_to_one corresponding, in internal memory
One metadata group is corresponding to this meta data block, and this metadata group comprises multiple metadata, and one
Metadata in the meta data block that metadata in this metadata group individual is corresponding with this metadata group one a pair
Should.
Certainly, memorizer 901 may also comprise other high-speed RAM memorizer, it is also possible to also includes non-
Volatile memory (non-volatile memory), for example, at least one disk memory.Memorizer 901
It is additionally operable to provide instruction and data to processor 902.
Memorizer 901 stores following element, executable module or data structure, or they
Subset, or their superset:
Operational order: include various operational order, is used for realizing various operation.
Operating system: include various system program, is used for realizing various basic business and processing based on firmly
The task of part.
In embodiments of the present invention, processor 902 (should by calling the operational order of memorizer 901 storage
Operational order is storable in operating system), perform to operate as follows:
Updating at least one metadata in internal memory, wherein, this at least one metadata corresponds to this Flash
In for storing at least one data block of data of write I/O Request command request write；For obtaining
At least one metadata group belonging to this at least one metadata；If in this at least one metadata group
The write state of the first metadata group is the pending unit such as existence under write-not state and this first metadata group
Multiple write I/O Request orders of data write operation, then move to the plurality of write I/O Request order
The merging write metadata queue of this first metadata group；This first metadata group is written to this Flash
In this meta data block corresponding to the first metadata group；Merging to this first metadata group writes metadata
Multiple write I/O Request orders under queue return this first metadata group write the most successfully state.
Metadata in internal memory, can be grouped according to the size of the meta data block in Flash.Such as,
In Flash, a meta data block has 128 metadata, then can 128 metadata be compiled in internal memory is 1
Group.Meanwhile, the metadata of a metadata group and the metadata of a meta data block in Flash in internal memory
One_to_one corresponding.
In the embodiment of the present invention, storage system 800, when writing metadata to Flash, merges multiple writing
Enter the I/O Request order write operation to same meta data block, writing of Flash can be reduced to a certain extent
Indegree also avoids IO to write the problem of amplification to a certain extent, improves the service life of Flash.
Alternatively, as an embodiment, when for updating at least one metadata in internal memory, place
Reason device 902 specifically for: if the status indication information of one of this at least one metadata indicates this at least
One of one metadata is clean Clean state, then update one of this at least one metadata, wherein,
This Clean state and is somebody's turn to do in metadata corresponding for this Flash for indicating one of this at least one metadata
One of at least one metadata content is identical.In the embodiment of the present invention, buffer storage 900 is dirty to having put
Metadata, no longer carry out write operation, reduce the write number of times of Flash the most to a certain extent, improve
The service life of Flash.
Alternatively, alternatively, the size of this meta data block page-size equal to this Flash.Unit
The size of data block is consistent with the page-size of Flash, it is possible to improve the life-span of Flash to a certain extent,
And improve Flash write performance.In prior art, the page-size of Flash is 4K.
Alternatively, the size of this metadata is 32B, 64B or 128B.
Alternatively, as an embodiment, the size of this chunk meta-data region and this data block
In the same size.The size of Chunk meta-data region is in the same size with a data block so that buffer storage
900, when loading Flash or initializing Flash, can carry out memory area according to the size of data block
Divide.
Alternatively, the size of this data block can be 512B, 1K, 2K, 4K, 8K, 16K,
32K or 64K.
It addition, the method that buffer storage 900 can also carry out Fig. 2, and realize buffer storage shown in Fig. 7
The function of embodiment, the embodiment of the present invention does not repeats them here.
Those of ordinary skill in the art are it is to be appreciated that combine each of the embodiments described herein description
The unit of example and algorithm steps, it is possible to electronic hardware or computer software and the knot of electronic hardware
Incompatible realization.These functions perform with hardware or software mode actually, depend on the spy of technical scheme
Fixed application and design constraint.Professional and technical personnel can use not Tongfang to each specifically should being used for
Method realizes described function, but this realization is it is not considered that beyond the scope of this invention.
Those skilled in the art is it can be understood that arrive, and for convenience and simplicity of description, above-mentioned retouches
The specific works process of system, device and the unit stated, is referred to the correspondence in preceding method embodiment
Process, does not repeats them here.
In several embodiments provided herein, it should be understood that disclosed system, device and
Method, can realize by another way.Such as, device embodiment described above is only shown
Meaning property, such as, the division of described unit, be only a kind of logic function and divide, actual can when realizing
There to be other dividing mode, the most multiple unit or assembly can in conjunction with or be desirably integrated into another
System, or some features can ignore, or do not perform.Another point, shown or discussed each other
Coupling direct-coupling or communication connection can be the INDIRECT COUPLING by some interfaces, device or unit
Or communication connection, can be electrical, machinery or other form.
The described unit illustrated as separating component can be or may not be physically separate, makees
The parts shown for unit can be or may not be physical location, i.e. may be located at a place,
Or can also be distributed on multiple NE.Can select according to the actual needs part therein or
The whole unit of person realizes the purpose of the present embodiment scheme.
It addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit
In, it is also possible to it is that unit is individually physically present, it is also possible to two or more unit are integrated in one
In individual unit.
If described function realizes using the form of SFU software functional unit and as independent production marketing or make
Used time, can be stored in a computer read/write memory medium.Based on such understanding, the present invention
The part that the most in other words prior art contributed of technical scheme or the portion of this technical scheme
Dividing and can embody with the form of software product, this computer software product is stored in a storage medium
In, including some instructions with so that computer equipment (can be personal computer, server,
Or the network equipment etc.) perform all or part of step of method described in each embodiment of the present invention.And it is front
The storage medium stated includes: USB flash disk, portable hard drive, read only memory (ROM, Read-Only Memory),
Random access memory (RAM, Random Access Memory), magnetic disc or CD etc. are various can
To store the medium of program code.
The above, the only detailed description of the invention of the present invention, but protection scope of the present invention is not limited to
In this, any those familiar with the art, can be easily in the technical scope that the invention discloses
Expect change or replace, all should contain within protection scope of the present invention.Therefore, the protection of the present invention
Scope should described be as the criterion with scope of the claims.
Claims (14)
1. the method caching Cache write, it is characterised in that the flash memory Flash of described Cache
Memory area include that multiple memory block chunk, a described chunk include a chunk metadata
District and a chunk data field, one chunk meta-data region includes at least one meta data block,
One chunk data field includes at least one data corresponding with at least one meta data block described
District, described meta data block includes that multiple metadata, described data field include multiple data block, described unit number
According to block and described data field one_to_one corresponding, described metadata and described data block one_to_one corresponding, in internal memory
One metadata group comprises multiple units number corresponding to a described meta data block, a described metadata group
According to, the first number in the meta data block that metadata in a described metadata group is corresponding with described metadata group
According to one_to_one corresponding, including:
Updating at least one metadata in internal memory, at least one metadata described corresponds to described Flash
In for storing at least one data block of data of write I/O Request command request write；
Obtain described at least one metadata group belonging at least one metadata；
If the write state of the first metadata group at least one metadata group described is write-not state
And multiple write I/O Request life of the pending metadata write operation such as existence under described first metadata group
Order, then the merging write that the plurality of write I/O Request order moves to described first metadata group is first
Data queue；
The metadata of described first metadata group is written to the first metadata group institute described in described Flash
In corresponding meta data block；
Described first metadata group is merged the multiple write I/O Request orders under write metadata queue
Return described first metadata group write the most successfully state.
2. the method for claim 1, it is characterised in that at least in described renewal internal memory
Individual metadata includes:
If the status indication information instruction of one of at least one metadata described at least one unit's number described
According to one of for clean Clean state, then update one of at least one metadata described, wherein, described Clean
State for indicate one of at least one metadata described metadata corresponding for described Flash with described extremely
One of few metadata content is identical.
3. method as claimed in claim 1 or 2, it is characterised in that a described meta data block
The size page-size equal to described Flash.
4. method as claimed in claim 3, it is characterised in that the size of a described metadata is
32B, 64B or 128B.
5. the method as described in claim 3 or 4, it is characterised in that described chunk meta-data region
Size and a described data block in the same size.
6. the method as described in claim 3 or 4, it is characterised in that a described data block big
Little for 512B, 1K, 2K, 4K, 8K, 16K, 32K or 64K.
7. a buffer storage, it is characterised in that the memory area of the flash memory Flash of described buffer storage
Including multiple memory blocks chunk, described chunk includes a chunk meta-data region and one
Chunk data field, one chunk meta-data region includes at least one meta data block, one
Chunk data field includes at least one data field corresponding with at least one meta data block described, described unit
Data block includes that multiple metadata, described data field include multiple data block, and described meta data block is with described
Data field one_to_one corresponding, described metadata and described data block one_to_one corresponding, a metadata in internal memory
Group is corresponding to a described meta data block, and a described metadata group comprises multiple metadata, described in one
Metadata one_to_one corresponding in the meta data block that metadata in metadata group is corresponding with described metadata group,
Including:
Updating block, for updating at least one metadata in internal memory, at least one metadata pair described
For storing at least one data of the data of write I/O Request command request write in Flash described in Ying Yu
Block；
Acquiring unit, for obtaining described at least one metadata group belonging at least one metadata, one
Metadata in individual described metadata group is corresponding to the first number in a meta data block in described Flash
According to；
Queue mobile unit, if the writing of the first metadata group at least one metadata group described
State is the pending metadata write operation such as existence under write-not state and described first metadata group
Multiple write I/O Request orders, then move to described first yuan of number by the plurality of write I/O Request order
Merging write metadata queue according to group；
Writing unit, for being written to first yuan of number described in described Flash by described first metadata group
According to the meta data block corresponding to group；
Feedback unit, writes for described first metadata group merges write under metadata queue multiple
Enter I/O Request order and return described first metadata group write the most successfully state.
8. buffer storage as claimed in claim 7, it is characterised in that described updating block is specifically used
In:
If the status indication information instruction of one of at least one metadata described at least one unit's number described
According to one of for clean Clean state, then update one of at least one metadata described, wherein, described Clean
State for indicate one of at least one metadata described metadata corresponding for described Flash with described extremely
One of few metadata content is identical.
9. buffer storage as claimed in claim 7 or 8, it is characterised in that a described metadata
The size of the block page-size equal to described Flash.
10. the buffer storage as described in any one of claim 7 to 9, it is characterised in that described chunk
The size of meta-data region is in the same size with a described data block.
11. 1 kinds of storage systems, it is characterised in that include memorizer and and processor, described memorizer
Flash memory Flash, the memory area of described Flash is included including internal memory, caching Cache, described Cache
Including multiple memory blocks chunk, described chunk includes a chunk meta-data region and one
Chunk data field, one chunk meta-data region includes at least one meta data block, one
Chunk data field includes at least one data field corresponding with at least one meta data block described, described unit
Data block includes that multiple metadata, described data field include multiple data block, and described meta data block is with described
Data field one_to_one corresponding, described metadata and described data block one_to_one corresponding, a metadata in internal memory
Group is corresponding to a described meta data block, and a described metadata group comprises multiple metadata, described in one
Metadata one_to_one corresponding in the meta data block that metadata in metadata group is corresponding with described metadata group,
Wherein, described processor is used for:
Updating at least one metadata in described internal memory, at least one metadata described is corresponding to described
For storing at least one data block of the data of write I/O Request command request write in Flash；
Obtain described at least one metadata group belonging at least one metadata, a described metadata group
In metadata corresponding to the metadata in a meta data block in described Flash；
If the write state of the first metadata group at least one metadata group described is write-not state
And multiple write I/O Request life of the pending metadata write operation such as existence under described first metadata group
Order, then the merging write that the plurality of write I/O Request order moves to described first metadata group is first
Data queue；
Described first metadata group is written to the unit corresponding to the first metadata group described in described Flash
Data block；
Described first metadata group is merged the multiple write I/O Request orders under write metadata queue
Return described first metadata group write the most successfully state.
12. store system as claimed in claim 11, it is characterised in that be used for updating described in
At least one metadata in depositing, described processor specifically for:
If the status indication information instruction of one of at least one metadata described at least one unit's number described
According to one of for clean Clean state, then update one of at least one metadata described, wherein, described Clean
State for indicate one of at least one metadata described metadata corresponding for described Flash with described extremely
One of few metadata content is identical.
The 13. storage systems as described in claim 11 or 12, the size etc. of a described meta data block
Page-size in described Flash.
The 14. storage systems as described in any one of claim 11 to 13, it is characterised in that described chunk
The size of meta-data region is in the same size with a described data block.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310402988.XA CN103473185B (en) | 2013-09-06 | 2013-09-06 | Method, buffer storage and the storage system of caching write |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310402988.XA CN103473185B (en) | 2013-09-06 | 2013-09-06 | Method, buffer storage and the storage system of caching write |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103473185A CN103473185A (en) | 2013-12-25 |
| CN103473185B true CN103473185B (en) | 2016-08-10 |
Family
ID=49798047
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310402988.XA Active CN103473185B (en) | 2013-09-06 | 2013-09-06 | Method, buffer storage and the storage system of caching write |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103473185B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104238962B (en) * | 2014-09-16 | 2018-02-06 | 华为技术有限公司 | The method and device of data is write into caching |
| CN104484285B (en) * | 2014-12-09 | 2017-11-17 | 杭州华为数字技术有限公司 | A kind of memory management method and device |
| CN105812891B (en) * | 2014-12-30 | 2019-03-19 | Tcl集团股份有限公司 | A kind of method and system of module TV screen end plate data storage |
| CN104503710B (en) * | 2015-01-23 | 2018-02-13 | 福州瑞芯微电子股份有限公司 | A kind of method and apparatus for improving flash memory writing speed |
| CN105117351B (en) * | 2015-09-08 | 2018-07-03 | 华为技术有限公司 | To the method and device of buffering write data |
| CN109213693B (en) * | 2017-06-30 | 2023-05-05 | 伊姆西Ip控股有限责任公司 | Storage management method, storage system and computer program product |
| CN114995748A (en) * | 2022-05-09 | 2022-09-02 | 阿里巴巴（中国）有限公司 | Request processing method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8407437B1 (en) * | 2012-03-14 | 2013-03-26 | Tegile Systems, Inc. | Scalable metadata acceleration with datapath metadata backup |
| CN102999428A (en) * | 2012-11-01 | 2013-03-27 | 华中科技大学 | Four-stage addressing method for tile recording disk |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8402205B2 (en) * | 2010-03-18 | 2013-03-19 | Seagate Technology Llc | Multi-tiered metadata scheme for a data storage array |
| US9274937B2 (en) * | 2011-12-22 | 2016-03-01 | Longitude Enterprise Flash S.A.R.L. | Systems, methods, and interfaces for vector input/output operations |
-
2013
- 2013-09-06 CN CN201310402988.XA patent/CN103473185B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8407437B1 (en) * | 2012-03-14 | 2013-03-26 | Tegile Systems, Inc. | Scalable metadata acceleration with datapath metadata backup |
| CN102999428A (en) * | 2012-11-01 | 2013-03-27 | 华中科技大学 | Four-stage addressing method for tile recording disk |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103473185A (en) | 2013-12-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103473185B (en) | Method, buffer storage and the storage system of caching write | |
| CN108139902A (en) | The method and apparatus of SSD drive are accessed for providing mixed mode | |
| US7529880B2 (en) | Address mapping table and method of storing mapping data in the same | |
| CN102779096B (en) | Page, block and face-based three-dimensional flash memory address mapping method | |
| CN103207840B (en) | For imperfect record to be degraded to the system and method for the second buffer memory from the first buffer memory | |
| CN106874211A (en) | The control method of accumulator system and nonvolatile memory | |
| CN104731886B (en) | A kind of processing method and system of mass small documents | |
| CN104246721A (en) | Storage system, storage controller, and method, for managing mapping between logical addresses and physical addresses | |
| CN109144886A (en) | Data storage device | |
| KR20140111588A (en) | System, method and computer-readable medium for managing a cache store to achieve improved cache ramp-up across system reboots | |
| CN104714893A (en) | Method and system for garbage collection scaling | |
| CN104809237A (en) | LSM-tree (The Log-Structured Merge-Tree) index optimization method and LSM-tree index optimization system | |
| CN102023809A (en) | Storage system, method for reading data from storage system and method for writing data to storage system | |
| CN103810237A (en) | Data management method and system | |
| CN104346290A (en) | Storage device, computer system and methods of operating same | |
| CN105117351A (en) | Method and apparatus for writing data into cache | |
| CN103383672A (en) | Cache control to reduce transaction roll back | |
| CN101968795A (en) | Cache method for file system with changeable data block length | |
| CN107688436A (en) | Memory module and its method of control | |
| CN109783396A (en) | The method of access flash memory module and relevant flash controller and electronic device | |
| CN104063330A (en) | Data prefetching method and device | |
| CN110442529A (en) | Configurable storage system and configuration and the method using the storage system | |
| CN103176914B (en) | The caching method of a kind of low miss rate, low disappearance punishment and device | |
| CN112148217B (en) | Method, device and medium for caching deduplication metadata of full flash memory system | |
| CN103049393B (en) | Memory headroom management method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |