CN107092564A - A kind of data processing method and device - Google Patents
A kind of data processing method and device Download PDFInfo
- Publication number
- CN107092564A CN107092564A CN201710266779.5A CN201710266779A CN107092564A CN 107092564 A CN107092564 A CN 107092564A CN 201710266779 A CN201710266779 A CN 201710266779A CN 107092564 A CN107092564 A CN 107092564A
- Authority
- CN
- China
- Prior art keywords
- chained list
- metadata
- data
- shadow
- priority
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/0223—User address space allocation, e.g. contiguous or non contiguous base addressing
- G06F12/0292—User address space allocation, e.g. contiguous or non contiguous base addressing using tables or multilevel address translation means
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Computer And Data Communications (AREA)
Abstract
The invention provides a kind of data processing method and device, applied to distributed memory system, the distributed memory system includes multiple chained lists, and priority and access times and the other chained lists of the data corresponding to metadata that each chained list is stored are differed, the treating method comprises:Receive the request for accessing data, and the metadata according to corresponding to the request determines the data that needs are accessed;Determine the target chained list where the metadata;The position of the metadata is adjusted according to the priority and access times of data in the target chained list.It can be seen that, the present invention is provided with multiple chained lists according to priority and access times, because the priority of key business data is higher, and access times are also relatively more, therefore, when running into scan-type read-write, storage location of the corresponding metadata of key business data in chained list can't preferentially be eliminated, scan-type read-write can be resisted, the performance of distributed memory system is improved.
Description
Technical field
The present invention relates to technical field of data processing, more specifically to a kind of data processing method and device.
Background technology
Super fusion architecture be it is a kind of will calculate, facility is integrated based on the resource such as network and storage, can be with
Selection combination is carried out and self-defined according to specific operation system demand, data center is conveniently and efficiently carried out and builds and operation system
A kind of Technical Architecture of deployment.
Distributed memory system is one of core component of super fusion architecture, usually using Bedding storage.Layering is deposited
Storage refers to the mixing storage architecture of a kind of solid-state disk and mechanical disk, and the performance of solid-state disk is higher by much again than mechanical disk, and price
Also many times are higher by, therefore, using solid-state disk as performance layer, mechanical disk is placed on as capacity layer using the data frequently accessed
Quick solid-state disk, the data of less access are placed on mechanical disk at a slow speed.Determine which data is saved using hierarchical algorithm
In solid-state disk, which data is stored in mechanical disk.
Wherein, conventional hierarchical algorithm is not to be accessed algorithm (LRU) at most.Lru algorithm is by the data in solid-state disk
Metadata organization is into chained list, and by the sequencing sequence of data the last time accessed time, and the accessed time is more early, then
Corresponding metadata is stored in chained list afterbody, and the accessed time is more late, closer to the current time, then corresponding metadata
It is stored in the head of chained list.So, as long as data are accessed, then its corresponding metadata will be moved to chained list head,
And metadata then is extractd from chained list afterbody when needing to eliminate cold data, and then by the corresponding data of the metadata picked-offing from admittedly
Mechanical disk is moved in state disk.
But, when run into such as antivirus, Backup Data need the scan-type read-write operation of ergodic data when, calculated using LRU
The processing method of method can upset original list structure completely, so that when needing to eliminate data, it is likely that eliminate key business
Data, it is impossible to ensure that key business data is preferably stored in solid-state disk, have had a strong impact on the performance of distributed memory system.
The content of the invention
In view of this, the invention provides a kind of data processing method and device, for solving not supporting in the prior art
Imperial scan-type read-write causes the problem of performance of distributed memory system is relatively low.
Technical scheme includes:
A kind of data processing method, applied to distributed memory system, the distributed memory system includes multiple chained lists,
The chained list is used for the corresponding metadata of data for storing the distributed memory system, the metadata institute that each chained list is stored
The priority and access times of corresponding data are differed with other chained lists, the treating method comprises:
Receive the request for accessing data, and the metadata according to corresponding to the request determines the data that needs are accessed;
Determine target chained list of the metadata in the multiple chained list;
The storage location of the metadata is adjusted according to the priority and access times of data in the target chained list
It is whole.
It is preferred that, the distributed memory system also includes at least one shadow chained list, and the processing method also includes:
If it is determined that the metadata does not judge whether the metadata is stored in the shadow in the multiple chained list, then
In chained list;
In the case where the metadata is stored in the shadow chained list, according to the priority of the data and access time
Number determines first object chained list of the metadata in the multiple chained list;
The metadata is stored in the first object chained list.
It is preferred that, in addition to:
In the case where the metadata is not stored in the multiple chained list and the shadow chained list, then according to described
The priority of data determines the second target chained list of the metadata, and the second target chained list is the multiple chained list or shadow
Any one in chained list;
The metadata is stored in the second target chained list.
It is preferred that, the chained list includes:First chained list, the second chained list, the 3rd chained list and the 4th chained list;
Wherein, it is common and senior and only accessed 1 time data that first chained list, which is used for the priority of data storage,
Corresponding metadata, and the metadata removed from the 3rd chained list;
The priority that second chained list is used for data storage is common corresponding with senior and accessed at least 2 times data
Metadata, and the metadata removed from the 4th chained list;
It is the high and only accessed 1 time corresponding first number of data that 3rd chained list, which is used for the priority of data storage,
According to;
It is the high and at least accessed 2 times corresponding first number of data that 4th chained list, which is used for the priority of data storage,
According to.
It is preferred that, the storage according to the priority and access times of data in the target chained list to the metadata
Position be adjusted including:
If it is determined that the metadata is in first chained list, then whether the priority for judging the data is high, if
It is that then the metadata is moved in the 4th chained list;Otherwise, the metadata is moved in second chained list;
If it is determined that the metadata is in second chained list, then whether the priority for judging the data is high, if
It is that then the metadata is moved in the 4th chained list;Otherwise, the metadata is placed in the head of second chained list;
If it is determined that the metadata is in the 3rd chained list, then the metadata is moved in the 4th chained list;
If it is determined that the metadata is in the 4th chained list, the metadata is placed in the head of the 4th chained list.
It is preferred that, the shadow chained list includes the first shadow chained list, the second shadow chained list and special shadow chained list;
Wherein, the first shadow chained list is used to store the metadata eliminated from first chained list;
The second shadow chained list is used to store the metadata eliminated from second chained list;
It is first number corresponding to data that are common and being only accessed 1 time that the special shadow chained list, which is used to store priority,
According to.
It is preferred that, it is described to determine in shadow chained list described in the metadata, then according to the priority of the data and access
Number of times determines that storage location of the metadata in the multiple chained list includes:
If it is determined that the metadata is in the first shadow chained list, then whether the priority for judging the data is high, if
It is that then the metadata in the first shadow chained list is moved in the 4th chained list, otherwise, the metadata is moved
Into second chained list;
If it is determined that the metadata is in the second shadow chained list, then whether the priority for judging the data is high, if
It is that then the metadata in the second shadow chained list is moved in the 4th chained list, otherwise, the metadata is moved
Into second chained list;
If it is determined that the metadata is in the special shadow chained list, then by first number in the special shadow chained list
According to being moved in second chained list.
Another aspect of the present invention discloses a kind of data processing equipment, applied to distributed memory system, the distribution
Storage system includes multiple chained lists, and the chained list is used for the corresponding metadata of data for storing the distributed memory system, often
Priority and access times and the other chained lists for the data corresponding to metadata that individual chained list is stored are differed, the processing
Device includes:
Receiving module, the request of data is accessed for receiving, and determine that the data institute that needs are accessed is right according to the request
The metadata answered;
First determining module, for determining target chained list of the metadata in the multiple chained list;
First performing module, for according to the priority and access times of data in the target chained list to the metadata
Storage location be adjusted.
It is preferred that, the distributed memory system also includes at least one shadow chained list, and the processing unit also includes:
Second determining module, for if it is determined that the metadata does not judge the metadata in the multiple chained list, then
Whether it is stored in the shadow chained list;
Second performing module, in the case of being stored in the metadata in the shadow chained list, according to the number
According to priority and access times determine first object chained list of the metadata in the multiple chained list;By the metadata
It is stored in the first object chained list.
It is preferred that, in addition to:
3rd performing module, for being not stored in the metadata in the multiple chained list and the shadow chained list
In the case of, then the second target chained list of the metadata is determined according to the priority of the data, the second target chained list is
Any one in the multiple chained list or shadow chained list;The metadata is stored in the second target chained list.
The invention provides a kind of data processing method and device, applied to distributed memory system, the distribution is deposited
Storage system includes multiple chained lists, and the chained list is used for the corresponding metadata of data for storing the distributed memory system, each
Priority and access times and the other chained lists for the data corresponding to metadata that chained list is stored are differed, the processing side
Method includes:Receive the request for accessing data, and the metadata according to corresponding to the request determines the data that needs are accessed;It is determined that
Target chained list where the metadata;According to the priority and access times of data in the target chained list to the metadata
Position be adjusted.It can be seen that, the present invention is provided with multiple chained lists according to priority and access times, due to key business number
According to priority it is higher, and access times are also relatively more, therefore, when running into scan-type read-write, key business data correspondence
Storage location of the metadata in chained list can't preferentially be eliminated, it is ensured that key business data can preferentially be stored in solid
In state disk, scan-type read-write can be resisted, the performance of distributed memory system is improved.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the accompanying drawing used required in technology description to be briefly described, it should be apparent that, drawings in the following description are only this
The embodiment of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can also basis
The accompanying drawing of offer obtains other accompanying drawings.
Fig. 1 is a kind of flow chart of data processing method disclosed in the embodiment of the present invention;
Fig. 2 is the flow chart of another data processing method disclosed in the embodiment of the present invention;
Fig. 3 is the flow chart of another data processing method disclosed in the embodiment of the present invention;
Fig. 4 is a kind of data processing equipment structural representation disclosed in the embodiment of the present invention.
Embodiment
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 clear, complete
Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.It is based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of creative work is not made
Embodiment, belongs to the scope of protection of the invention.
First, the technical term used in the present invention is explained.
Super fusion architecture:Be it is a kind of will calculate, facility is integrated based on the resource such as network and storage, can be with
Selection combination is carried out and self-defined according to specific operation system demand, data center is conveniently and efficiently carried out and builds and operation system
A kind of Technical Architecture of deployment.It is usually that software virtualization is incorporated in cell node (x86 servers) in specific implementation
Technology (including calculating, network, storage, safety etc. virtualization), and each cell node can be got up by network polymerization, it is real
Existing modular seamless (scale-out) extending transversely, builds unified resource pool.
Distributed memory system:Distributed memory system is one of core component of super fusion architecture, and data are divided
Dissipate and be stored in many independent storage devices, by the network interconnection and by transmitting message realization collaboration between storage device
Work.Distributed memory system uses expansible system architecture, shares storage load using many storage servers, utilizes position
Put server or uniformity Hash scheduling algorithm location data.
Quality of service guarantee (QoS):When business datum has exceeded solid-state disk capacity, the storage of key business is preferentially ensured
Energy.
Technical scheme is described in detail below.
Referring to Fig. 1, Fig. 1 is a kind of flow chart of data processing method disclosed in the embodiment of the present invention.
Processing method in the present invention is applied to distributed memory system, and the distributed memory system includes multiple chains
Table, the chained list is used for the corresponding metadata of data for storing the distributed memory system, first number that each chained list is stored
Differed according to the priority and access times and other chained lists of corresponding data.
It is understood that the distributed memory system applied in the present invention can be distributed layer storage system or
The layering subsystem of distributed memory system.
In the embodiment of the present invention, the data to distributed memory system pre-set priority.Specifically, by data
Priority be set to it is high, it is senior and common.Priority is that high data are usually key business data, to the performance of system
Influence is very big, it usually needs be stored in always in solid-state disk.
The corresponding metadata of data being stored with solid-state disk in multiple chained lists, is the priority according to data and access
The two parameters of number of times are come what is be configured, and putting in order for metadata is determined according to last access time.
The access times and priority of the data corresponding to metadata stored in each chained list are different.So set
The purpose put is exactly to be not easy to be eliminated in order that obtaining key business data, it is ensured that the performance of system.Multiple chained lists are preferentially stored
In the internal memory of distributed memory system.
As shown in figure 1, the processing method includes:
S101, the request for receiving access data, and first number according to corresponding to the request determines the data that needs are accessed
According to;
In the embodiment of the present invention, can be divided into by accessing the type of request of data by two kinds, and one kind is read request, it is another then
It is write request.
Receive after the request, according to the request data to be accessed, determine the corresponding metadata of data.
In this step, it can be performed by searching the Hash table originally set up in solid-state disk Central Plains, Hash table storage
There is the corresponding relation of data and metadata, can be by the corresponding metadata of data of this table lookup to the access request.
Wherein, the pointer of chained list, correspondence number where metadata includes physical address of the data on solid-state disk, metadata
According to priority.
S102, determine target chained list of the metadata in the multiple chained list;
S103, according to the priority and access times of data in the target chained list storage location of the metadata is entered
Row adjustment.
In the embodiment of the present invention, determine after metadata, judge whether the metadata is stored in multiple chained lists, if any
In one chained list, then the chained list is regard as target chained list.
Obtain after target chained list, the storage of metadata in multiple chained lists is determined according to the priority and access times of data
Position.Being ranked up according to priority and access times can just avoid key business from being preferentially put into the position that chained list needs to eliminate
In putting, when needing to eliminate data, the data of key business will not be eliminated to machinery disk.
Therefore, when running into scan-type read-write, the corresponding metadata of key business data can't be influenceed in chained list
Position, and due to the priority setting to data, it is ensured that key business data preferentially can be stored in solid-state disk.
Fig. 2 is a kind of another flow chart for data processing method that the present invention is provided.
In the embodiment of the present invention, the metadata that is also stored with is additionally provided with least one shadow chained list, shadow chained list, these
Metadata is that the metadata eliminated from multiple chained lists either priority is corresponding for the data for accessing commonly and only 1 time
Metadata, the data corresponding to metadata stored in shadow chained list are stored in mechanical disk.Below pair if it is determined that institute
Situation of the metadata not in the multiple chained list is stated to be introduced.
Referring to Fig. 2, the invention provides a kind of data processing method, the distributed memory system also includes at least one
Shadow chained list, the treating method comprises:
S201, the request for receiving access data, and first number according to corresponding to the request determines the data that needs are accessed
According to;
S202, determine target chained list of the metadata in the multiple chained list;
S203, according to the priority and access times of data in the target chained list storage location of the metadata is entered
Row adjustment;
Step S201- steps S203 is identical with step S101- steps S103, here, without repeating.
S204, if it is determined that the metadata does not judge whether the metadata is stored in institute in the multiple chained list, then
State in shadow chained list;
In the embodiment of the present invention, chained list and shadow chained list are two distinct types of tables, if the metadata is not multiple
In any one in chained list, then judge whether the metadata is stored in shadow chained list.To judge to need the data accessed to be
The no priority for being the data being just eliminated or being accessed 1 time is common data.
S205, in the case where the metadata is stored in the shadow chained list, according to the priority of the data and
Access times determine first object chained list of the metadata in the multiple chained list;
If metadata is in shadow chained list, mean the data be probably the data eliminated from chained list either
Priority is data that are common and being accessed 1 time.If the data being eliminated from chained list, it is meant that needed before being
The data in solid-state disk are put into, if priority is data that are common and being accessed 1 time, then meaning can be right
Whether the data optimize operation, i.e., be put into solid-state disk.
Accordingly, it is determined that the first object chained list of the metadata, first object chained list be multiple chained lists one of those.
S206, the metadata is stored in the first object chained list.
The metadata is moved to from shadow chained list in the first object chained list so that data are moved to from mechanical disk
In solid-state disk.
In the embodiment of the present invention, there is provided shadow chained list, can enter for being actually needed to the metadata in shadow chained list
Row optimization.For example it is directly stored in certain chained list so that data are moved to solid-state disk from mechanical disk.And due to normal priority
Data need it is accessed can just be placed into solid-state disk twice, therefore, can preferentially ensure the data Cun Chudao solid-states of key business
In disk, it is ensured that the performance of distributed memory system, that is, it ensure that the storage service quality of key business.
If neither meaning that this metadata is not deposited in multiple chained lists nor the metadata stored in shadow chained list
Storage describes in detail to secondary process below in chained list or shadow chained list.
Fig. 3 is a kind of another flow chart for data processing method that the present invention is provided.
Referring to Fig. 3, it the treating method comprises:
S301, the request for receiving access data, and first number according to corresponding to the request determines the data that needs are accessed
According to;
S302, determine target chained list of the metadata in the multiple chained list;
S303, according to the priority and access times of data in the target chained list storage location of the metadata is entered
Row adjustment;
S304, if it is determined that the metadata does not judge whether the metadata is stored in institute in the multiple chained list, then
State in shadow chained list;
S305, in the case where the metadata is stored in the shadow chained list, according to the priority of the data and
Access times determine first object chained list of the metadata in the multiple chained list;
S306, the metadata is stored in the first object chained list;
Step S301- steps S306 is identical with step S201- steps S206, here, without repeating.
S307, in the case where the metadata is not stored in the multiple chained list and the shadow chained list, then press
The type of request and the priority of the data according to the access data determine the second target chained list of the metadata, institute
It is any one in the multiple chained list or shadow chained list to state the second target chained list;
S308, the metadata is stored in the second target chained list.
In the embodiment of the present invention, if it is determined that metadata can then be visited not in multiple chained lists or shadow chained list according to described
Ask that the type of the request of data and the priority of the data determine the metadata in the multiple chained list or shadow chained list
In the second target chained list.
When data that this needs is accessed are not to be accessed and when being not the data being just eliminated, it will be visited according to described
Ask metadata that the type of the request of data and the priority of data determine corresponding to data should be stored in chained list or
In shadow chained list.
For example, if the request for accessing data is read request, priority is common data, then by the corresponding member of the data
In data Cun Chudao shadow chained lists.If the request for accessing data is write request, priority is common data, then by the data
Corresponding metadata storage is into the first chained list.If the data of other priority, then the corresponding metadata of the data is deposited
Storage is in corresponding chained list.
Due to being the storage carried out according to the priority of data, therefore, key business data and other data can be distinguished,
Preferentially ensure that key business data is saved in solid-state disk, it is ensured that the performance of distributed memory system.
It refer to have multiple chained lists and at least one shadow chained list in the embodiment of the present invention in above-described embodiment, below to many
Individual chained list and shadow chained list describes in detail.
In the embodiment of the present invention, the chained list includes:First chained list, the second chained list, the 3rd chained list and the 4th chained list;
Wherein, priority is general when the type that first chained list is used to store the request of the access data is write request
Metadata corresponding to data that are logical and being only accessed 1 time, and the type of the request for accessing data are read request or write
Priority is the metadata corresponding to senior and only accessed 1 time data during request, and is removed from the 3rd chained list
Metadata;
The priority when type that second chained list is used to store the request of the access data is read request or write request
For the corresponding metadata of common and senior and accessed at least 2 times data, and the first number removed from the 4th chained list
According to;
Priority when 3rd chained list is used to store the type for stating the request for accessing data including read request or write request
For the high and only accessed 1 time corresponding metadata of data;
Priority when 4th chained list is used to store the type for stating the request for accessing data including read request or write request
For the high and at least accessed 2 times corresponding metadata of data.
The shadow chained list includes the first shadow chained list, the second shadow chained list and special shadow chained list;
Wherein, the first shadow chained list is used to store the metadata eliminated from first chained list;
The second shadow chained list is used to store the metadata eliminated from second chained list;
When the request that the special shadow chained list is used to store the access data is read request, priority is commonly and only
It is accessed the metadata corresponding to the data of 1 time.
In the embodiment of the present invention, 4 chained lists and 3 shadow chained lists are provided with, 4 chained lists are the first chained list T1, the second chain
Table T2, the 3rd chained list T3 and the 4th chained list T4.Shadow chained list is the first shadow chained list B1, the second shadow chained list B2 and special shadow
Child list Qs.
T1, T2, T3, T4 length are controlled by tri- parameters of P1, P2, P3.P1, P2, P3 are at the beginning of system
Beginningization is set constant, it is assumed that the capacity of solid-state disk is C, and T2 is preferably arranged to 20%C, and P1, P3 are then set to 0.Respectively
The length and parameter of individual chained list meet following relationship:
(1) T1+T2+T3+T4<=C;
(2) T3+T4<=P2;
(3) B1+B2+T1+T2+T3+T4<=2C;
(4) Qs<=0.5C;
(5) B1+T1+T3<C;
Wherein, P1 and P3 are that and then business carrys out adjust automatically, the length for controlling T1 and T3.Specific adjustment process exists
Followed by being discussed in detail.
In the embodiment of the present invention, the position for determining the metadata in the multiple chained list, and according to the number
According to priority and access times in the multiple chained list metadata rearrangement include:
If it is determined that the metadata is in first chained list, then whether the priority for judging the data is high, if
It is that then the metadata is moved in the 4th chained list;Otherwise, the metadata is moved in second chained list;
Metadata is determined in the first chained list it should be noted that working as, and the priority of data is high, by described first
While the metadata in chained list is moved in the 4th chained list, also increase P3 value, and if T3 length and
T4 length sum carries out eliminating processing more than P2, the then data to T3 or T4.
Processing is eliminated in actual use can use function DEGRADE ().Wherein, increase P3 refers to the base in original P3
Increase T4/T3, i.e. T4 divided by T3 on plinth.Need then to use formula during increase P3 in actual use:P3=P3+T4/T3, its is medium
P3 before number represents that the P3 behind the value of parameter P3 after increase, equal sign refers to the value of parameter P3 before increasing, T4 and T3 difference
The "/" identified in the length of the 3rd chained list and the 4th chained list, formula represents division arithmetic.
If it is determined that the metadata is in second chained list, then whether the priority for judging the data is high, if
It is that then the metadata is moved in the 4th chained list;Otherwise, the metadata is placed in the head of second chained list
Portion;
It should be noted that the head that the metadata is placed in into second chained list can also be according to preset rules pair
Metadata in second chained list re-starts sequence, for example, the priority of data is judged, if not high, then by the metadata
After the position for being arranged in the corresponding metadata of high data in second chained list.It is, of course, also possible to using other manner,
This is without repeating.
It should be noted that work as metadata in the second chained list, and the priority of data is high, then can reduce P3
Value, if T3 and T4 length and more than P2, the data to T3 or T4 are carried out eliminating processing, and letter can be used in actual use
Number DEGRADE ().
Wherein, P3 is reduced to refer to reduce T3/T4 on the basis of original P3.Need then to make when reducing P3 in actual use
Use formula:P3 before P3=P3-T3/T4, wherein equal sign represents that the P3 behind the value of parameter P3 after reducing, equal sign refers to subtracting
The "/" that small preceding parameter P3 value, T4 and T3 are identified in the length of the 3rd chained list and the 4th chained list, formula respectively represents that division is transported
Calculate.
Wherein, function DEGRADE () is specifically included:
Wherein, T3 value is more than P3 and refers to that T3 length is more than parameter P3, and being downgraded to T1 from T3 refers to wash in a pan from T3 end
Metadata is eliminated into T1, being downgraded to T2 from T4 refers to eliminate metadata into T1 from T4 end.
If it is determined that the metadata is in the 3rd chained list, then the metadata in the 3rd chained list is moved to
In 4th chained list;
If it is determined that the metadata is in the 4th chained list, the metadata is placed in the head of the 4th chained list.
By said process, the member in any one chained list in the processing procedure and chained list of metadata when metadata is completed
Process and the control to each chained list length that data are ranked up.
If described to determine shadow chain described in the metadata during any one of metadata in three shadow chained lists
In table, then storage location of the metadata in the multiple chained list is determined according to the priority and access times of the data
Including:
If it is determined that the metadata is in the first shadow chained list, then whether the priority for judging the data is high, if
It is that then the metadata in the first shadow chained list is moved in the 4th chained list, otherwise, the metadata is moved
Into second chained list;
In the embodiment of the present invention, however, it is determined that the metadata judges the priority of data in the first shadow chained list B1, if
It is high, the metadata in the first shadow chained list is moved in the 4th chained list, also, increase parameter P1
Value, and in the memory space inadequate of solid-state disk, then eliminates metadata to the first shadow chained list or by the from the first chained list
Metadata is eliminated in two chained lists to the second shadow chained list.Function REPLACE () is used in actual use.
If the priority of data is common and senior, the metadata in the first shadow chained list is moved to second
In chained list, and increase parameter P1 value, in the memory space inadequate of solid-state disk, then metadata is eliminated from the first chained list to the
One shadow chained list will eliminate metadata to the second shadow chained list in the second chained list.Function is used in actual use
REPLACE()。
Wherein, in actual use, increase P1 value uses formula P1=P1+T2/T1, the wherein P1 before equal sign to represent to increase
Plus after parameter P1 value, the P1 behind equal sign refers to the value of P3 parameters before increasing, T1 and T2 identify respectively the first chained list and
"/" in the length of second chained list, formula represents division arithmetic.
In actual use, function REPLACE () includes:
Wherein, eliminated from T1 to B1 and represent to eliminate metadata into the first shadow chained list from the end of the first chained list.T2 is washed in a pan
B2 is eliminated to represent to eliminate metadata into the second shadow chained list from the end of the second chained list.
If it is determined that the metadata is in the second shadow chained list, then whether the priority for judging the data is high, if
It is that then the metadata in the second shadow chained list is moved in the 4th chained list, otherwise, the metadata is moved
Into second chained list;
In the embodiment of the present invention, however, it is determined that the metadata then judges the excellent of the metadata in the second shadow chained list
Whether first level is common or senior, if so, then the metadata in the second shadow chained list is moved in the second chained list, and
And, reduce parameter P1 value.And in the insufficient space of solid-state disk, metadata is eliminated from the first chained list to the first shadow chained list
Or metadata will be eliminated in the second chained list to the second shadow chained list.In actual use using foregoing disclosed function
REPLACE()。
If the priority of data is high, the metadata in the second shadow chained list is moved in the 4th chained list,
And reduce parameter P1 value.In T3 and T4 length and during more than P2, then the data to T3 or T4 carry out eliminating processing, specifically
Performed using the function DEGRADE () in foregoing.And in the insufficient space of solid-state disk, member number is eliminated from the first chained list
According to metadata will be eliminated to the first shadow chained list or in the second chained list to the second shadow chained list.In actual use using foregoing
Disclosed function REPLACE ().
Wherein, in actual use, the value for reducing P1 uses formula P1=P1-T1/T2, the wherein P1 before equal sign to represent to subtract
The value of P1 parameters before P1 behind the value of parameter P1 after small, equal sign refers to reducing, T1 and T2 identify respectively the first chained list and
"/" in the length of second chained list, formula represents division arithmetic.
If it is determined that the metadata is in the special shadow chained list, then by first number in the special shadow chained list
According to being moved in second chained list.
In the embodiment of the present invention, however, it is determined that the metadata then means that metadata institute is right in special shadow chained list
The priority for the data answered is common grade and has been accessed before once that is, the data corresponding to the metadata is preferential
Level is that common grade and being requested before was read once.Therefore, the metadata is moved to from special shadow chained list
In two chained lists.
It should be noted that in the second chained list, priority has for the senior corresponding metadata of data additionally once eliminates
Chance.
It should be noted that being not stored in the situation in the multiple chained list and the shadow chained list in the metadata
Under, then determine the second of the metadata according to according to the type of request and the priority of the data of the access data
Target chained list, the second target chained list is any one in the multiple chained list or shadow chained list, and this process is specifically included:
If the type of the request for accessing data is read request or write request, the priority of the data be it is senior,
Then by metadata storage into the first chained list;
If the type of the request for accessing data is read request or write request, the priority of the data be it is high,
Then by metadata storage into the 3rd chained list.
If the priority that the type of the request for accessing data is data described in read request is common, by first number
The special shadow chained list is arrived according to storage;
If the priority that the type of the request for accessing data is data described in write request is common, by first number
According to storage into the first chained list.In the embodiment of the present invention, if metadata is not in the multiple chained list and the shadow chained list,
Then it is proved to be newly-increased metadata, then judges the priority of data.
If common, when accessing the type of request of data to read, then special shadow chained list, special shadow are arrived in storage
The length of chained list is fixed, is eliminated according to LRU policy, that is, eliminates at most not accessed metadata.Accessing asking for data
The type asked is when writing, then to store to the first chained list.
If senior, then by the meta-data preservation into the first chained list, and in the insufficient space of solid-state disk, from first
Metadata is eliminated in chained list to the first shadow chained list or metadata will be eliminated in the second chained list to the second shadow chained list.In reality
Foregoing disclosed function REPLACE () is used in use.If the insufficient space of the first shadow chained list or the second shadow chained list,
Then eliminate the metadata at end in the first shadow chained list or the second shadow chained list.Function REPLACE_B is used in actual use
()。
It should be noted that function REPLACE_B () is:
Wherein, while T1+T3+B1>=C, be judge the first chained list, the length of the 3rd chained list and the first shadow chained list and
Whether the capacity C of solid-state disk is more than or equal to.Superseded B1 refers to the metadata that end is eliminated from the first shadow chained list.while
T1+T2+T3+T4+B1+B2>=2C, is to judge the first chained list, the second chained list, the 3rd chained list, the 4th chained list and the first shadow
Chained list, the length of the second shadow chained list and the solid-state disk capacity C for whether being more than or equal to twice.Superseded B2 refers to from the second shadow
The metadata at end is eliminated in chained list.
It should be noted that the embodiment of the present application can also judge the class of the request after the request of data access is received
The regulation rule of type, read request and write request differs only in metadata not in chained list or shadow chained list, the priority of data
For it is common when, the storage location of metadata be in the first chained list still in shadow chained list.
Pass through the above-mentioned technical proposal of the present invention, it is possible to achieve following effect:
By setting different replacement policies for three kinds of different priorities, it is ensured that the storage service quality of key business;
By setting different anti-scanning strategies for three kinds of different priorities, it is ensured that the storage service quality of key business;
The degradation of high priority file and superseded flow, it is ensured that space availability ratio and key business service quality.
Another aspect of the present invention provides a kind of data processing equipment,
Referring to Fig. 4, Fig. 4 is a kind of structural representation for data processing equipment that invention is provided.
A kind of data processing equipment that the present invention is provided, applied to distributed memory system, the distributed memory system
Including multiple chained lists, the multiple chained list is used for the corresponding metadata of data for storing the distributed memory system, Mei Gelian
Priority and access times and the other chained lists for the data corresponding to metadata that table is stored are differed, the processing unit
Including:
Receiving module 401, the request of data, and the data institute for determining to need to access according to the request are accessed for receiving
Corresponding metadata;
First determining module 402, for determining target chained list of the metadata in the multiple chained list;
First performing module 403, for according to the priority and access times of data in the target chained list to the member
The storage location of data is adjusted.
It is preferred that, the distributed memory system also includes at least one shadow chained list, and the processing unit also includes:
Second determining module, for if it is determined that the metadata does not judge the metadata in the multiple chained list, then
Whether it is stored in the shadow chained list;
Second performing module, in the case of being stored in the metadata in the shadow chained list, according to the number
According to priority and access times determine first object chained list of the metadata in the multiple chained list;By the metadata
It is stored in the first object chained list.
It is preferred that, in addition to:
3rd performing module, for being not stored in the metadata in the multiple chained list and the shadow chained list
In the case of, then the second target chained list of the metadata is determined according to the priority of the data, the second target chained list is
Any one in the multiple chained list or shadow chained list;The metadata is stored in the second target chained list.
It is preferred that, the type of the request for accessing data includes read request or write request, and the multiple chained list includes:The
One chained list, the second chained list, the 3rd chained list and the 4th chained list;
Wherein, when the type that first chained list is used to store the request of the access data is write request, priority is
Metadata corresponding to data that are common and being only accessed 1 time, and the type of the request for accessing data for read request or
Priority is the metadata corresponding to senior and only accessed 1 time data during write request, and is moved from the 3rd chained list
The metadata removed;
When the type that second chained list is used to store the request of the access data is read request or write request, priority
For the corresponding metadata of common and senior and accessed at least 2 times data, and the first number removed from the 4th chained list
According to;
When 3rd chained list is used to store the type for stating the request for accessing data including read request or write request, priority
For the high and only accessed 1 time corresponding metadata of data;
When 4th chained list is used to store the type for stating the request for accessing data including read request or write request, priority
For the high and at least accessed 2 times corresponding metadata of data.
It is preferred that, first performing module includes:
First execution unit, for if it is determined that the metadata then judges the excellent of the data in first chained list
Whether first level is high, if so, then the metadata is moved in the 4th chained list;Otherwise, the metadata is moved to
In second chained list;
Second execution unit, for if it is determined that the metadata then judges the excellent of the data in second chained list
Whether first level is high, if so, then the metadata is moved in the 4th chained list;Otherwise, the metadata is placed in institute
State the head of the second chained list;
3rd execution unit, for if it is determined that the metadata then moves the metadata in the 3rd chained list
Into the 4th chained list;
4th execution unit, for if it is determined that the metadata is placed in institute by the metadata in the 4th chained list
State the head of the 4th chained list.
It is preferred that, the shadow chained list includes the first shadow chained list, the second shadow chained list and special shadow chained list;
Wherein, the first shadow chained list is used to store the metadata eliminated from first chained list;
The second shadow chained list is used to store the metadata eliminated from second chained list;
When the request that the special shadow chained list is used to store the access data is read request, priority is commonly and only
It is accessed the metadata corresponding to the data of 1 time.
It is preferred that, second performing module includes:
5th execution unit, for if it is determined that the metadata then judges the excellent of the data in the first shadow chained list
Whether first level is high, if so, then the metadata in the first shadow chained list is moved in the 4th chained list, it is no
Then, the metadata is moved in second chained list;
6th execution unit, for if it is determined that the metadata then judges the excellent of the data in the second shadow chained list
Whether first level is high, if so, then the metadata in the second shadow chained list is moved in the 4th chained list, it is no
Then, the metadata is moved in second chained list;
7th execution unit, for if it is determined that the metadata is in the special shadow chained list, then by the special shadow
The metadata in child list is moved in second chained list.
It should be noted that a kind of data processing equipment of the present embodiment can use one kind in above method embodiment
Data processing method, for realizing whole technical schemes in above method embodiment, the function of its modules can basis
Method in above method embodiment is implemented, and it implements process and can refer to associated description in above-described embodiment, this
Place is repeated no more.
The invention provides a kind of data processing equipment, the processing unit is applied to distributed memory system, described point
Cloth storage system includes multiple chained lists, and the chained list is used for the corresponding first number of data for storing the distributed memory system
According to priority and access times and the other chained lists of the data corresponding to metadata that each chained list is stored are differed, institute
State processing unit receiving module and receive the request for accessing data, and according to corresponding to the request determines the data that needs are accessed
Metadata;First determining module determines the target chained list where the metadata;First performing module is according to the target chained list
The priority and access times of middle data are adjusted to the position of the metadata.
It can be seen that, data processing equipment of the invention is provided with multiple chained lists according to priority and access times, due to closing
The priority of key business datum is higher, and access times are also relatively more, therefore, when running into scan-type read-write, key business
Storage location of the corresponding metadata of data in chained list can't preferentially be eliminated, it is ensured that key business data can be preferential
It is stored in solid-state disk, scan-type read-write can be resisted, the performance of distributed memory system is improved.
Finally, in addition it is also necessary to explanation, herein, such as first and second or the like relational terms be used merely to by
One entity or operation make a distinction with another entity or operation, and not necessarily require or imply these entities or operation
Between there is any this actual relation or order.Moreover, term " comprising ", "comprising" or its any other variant meaning
Covering including for nonexcludability, so that process, method, article or equipment including a series of key elements not only include that
A little key elements, but also other key elements including being not expressly set out, or also include be this process, method, article or
The intrinsic key element of equipment.In the absence of more restrictions, the key element limited by sentence "including a ...", is not arranged
Except also there is other identical element in the process including the key element, method, article or equipment.
The embodiment of each in this specification is described by the way of progressive, and what each embodiment was stressed is and other
Between the difference of embodiment, each embodiment identical similar portion mutually referring to.
The foregoing description of the disclosed embodiments, enables professional and technical personnel in the field to realize or using the present invention.
A variety of modifications to these embodiments will be apparent for those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, it is of the invention
The embodiments shown herein is not intended to be limited to, and is to fit to and principles disclosed herein and features of novelty phase one
The most wide scope caused.
Claims (14)
1. a kind of data processing method, applied to distributed memory system, it is characterised in that the distributed memory system includes
Multiple chained lists, the multiple chained list is used for the corresponding metadata of data for storing the distributed memory system, each chained list institute
The priority and access times of data corresponding to the metadata of storage are differed with other chained lists, the processing method bag
Include:
Receive the request for accessing data, and the metadata according to corresponding to the request determines the data that needs are accessed;
Determine target chained list of the metadata in the multiple chained list;
The storage location of the metadata is adjusted according to the priority and access times of data in the target chained list.
2. processing method according to claim 1, it is characterised in that the distributed memory system also includes at least one
Shadow chained list, the processing method also includes:
If it is determined that the metadata does not judge whether the metadata is stored in the shadow chained list in the multiple chained list, then
In;
It is true according to the priority and access times of the data in the case where the metadata is stored in the shadow chained list
Fixed first object chained list of the metadata in the multiple chained list;
The metadata is stored in the first object chained list.
3. processing method according to claim 2, it is characterised in that also include:
In the case where the metadata is not stored in the multiple chained list and the shadow chained list, then accessed according to described
The priority of the type of the request of data and the data determines the second target chained list of the metadata, second target
Chained list is any one in the multiple chained list or shadow chained list;
The metadata is stored in the second target chained list.
4. the processing method according to Claims 2 or 3, it is characterised in that the type of the request of the access data includes
Read request or write request, the multiple chained list include:First chained list, the second chained list, the 3rd chained list and the 4th chained list;
Wherein, when the type that first chained list is used to store the request of the access data is write request, priority is common
And the metadata corresponding to the data of 1 time, and the type of the request for accessing data are only accessed for read request or write please
Priority is the metadata corresponding to senior and only accessed 1 time data when asking, and removed from the 3rd chained list
Metadata;
When the type that second chained list is used to store the request of the access data is read request or write request, priority is general
Lead to metadata corresponding with senior and accessed at least 2 times data, and the metadata removed from the 4th chained list;
When 3rd chained list is used to store the type for stating the request for accessing data including read request or write request, priority is pole
The corresponding metadata of data that is high and being only accessed 1 time;
When 4th chained list is used to store the type for stating the request for accessing data including read request or write request, priority is pole
The corresponding metadata of data that is high and being at least accessed 2 times.
5. processing method according to claim 4, it is characterised in that it is described according in the target chained list data it is preferential
Level and access times the storage location of the metadata is adjusted including:
If it is determined that the metadata is in first chained list, then whether the priority for judging the data is high, if so, then
The metadata is moved in the 4th chained list;Otherwise, the metadata is moved in second chained list;
If it is determined that the metadata is in second chained list, then whether the priority for judging the data is high, if so, then
The metadata is moved in the 4th chained list;Otherwise, the metadata is placed in the head of second chained list;
If it is determined that the metadata is in the 3rd chained list, then the metadata is moved in the 4th chained list;
If it is determined that the metadata is in the 4th chained list, the metadata is placed in the head of the 4th chained list.
6. processing method according to claim 4, it is characterised in that the shadow chained list includes the first shadow chained list, the
Two shadow chained lists and special shadow chained list;
Wherein, the first shadow chained list is used to store the metadata eliminated from first chained list;
The second shadow chained list is used to store the metadata eliminated from second chained list;
When the request that the special shadow chained list is used to store the access data is read request, priority is common and is only interviewed
Asked the metadata corresponding to the data of 1 time.
7. processing method according to claim 6, it is characterised in that described to be stored in the shadow chain in the metadata
In the case of in table, of the metadata in the multiple chained list is determined according to the priority and access times of the data
One target chained list includes:
If it is determined that the metadata is in the first shadow chained list, then whether the priority for judging the data is high, if so, then
The metadata in the first shadow chained list is moved in the 4th chained list, otherwise, the metadata institute is moved to
State in the second chained list;
If it is determined that the metadata is in the second shadow chained list, then whether the priority for judging the data is high, if so, then
The metadata in the second shadow chained list is moved in the 4th chained list, otherwise, the metadata institute is moved to
State in the second chained list;
If it is determined that the metadata is in the special shadow chained list, then the metadata in the special shadow chained list is moved
Move in second chained list.
8. a kind of data processing equipment, applied to distributed memory system, it is characterised in that the distributed memory system includes
Multiple chained lists, the multiple chained list is used for the corresponding metadata of data for storing the distributed memory system, each chained list institute
The priority and access times of data corresponding to the metadata of storage are differed with other chained lists, the processing unit bag
Include:
Receiving module, the request of data is accessed for receiving, and according to corresponding to the request determines the data that needs are accessed
Metadata;
First determining module, for determining target chained list of the metadata in the multiple chained list;
First performing module, for being deposited according to the priority and access times of data in the target chained list to the metadata
Storage space, which is put, to be adjusted.
9. processing unit according to claim 8, it is characterised in that the distributed memory system also includes at least one
Shadow chained list, the processing unit also includes:
Second determining module, for if it is determined that whether the metadata does not judge the metadata in the multiple chained list, then
It is stored in the shadow chained list;
Second performing module, in the case of being stored in the metadata in the shadow chained list, according to the data
Priority and access times determine first object chained list of the metadata in the multiple chained list;The metadata is stored
In the first object chained list.
10. processing unit according to claim 9, it is characterised in that also include:
3rd performing module, for situation about being not stored in the metadata in the multiple chained list and the shadow chained list
Under, then determine the second target of the metadata according to the type of request and the priority of the data of the access data
Chained list, the second target chained list is any one in the multiple chained list or shadow chained list;The metadata is stored in
In the second target chained list.
11. the processing unit according to claim 9 or 10, it is characterised in that the type bag of the request of the access data
Read request or write request are included, the multiple chained list includes:First chained list, the second chained list, the 3rd chained list and the 4th chained list;
Wherein, when the type that first chained list is used to store the request of the access data is write request, priority is common
And the metadata corresponding to the data of 1 time, and the type of the request for accessing data are only accessed for read request or write please
Priority is the metadata corresponding to senior and only accessed 1 time data when asking, and removed from the 3rd chained list
Metadata;
When the type that second chained list is used to store the request of the access data is read request or write request, priority is general
Lead to metadata corresponding with senior and accessed at least 2 times data, and the metadata removed from the 4th chained list;
When 3rd chained list is used to store the type for stating the request for accessing data including read request or write request, priority is pole
The corresponding metadata of data that is high and being only accessed 1 time;
When 4th chained list is used to store the type for stating the request for accessing data including read request or write request, priority is pole
The corresponding metadata of data that is high and being at least accessed 2 times.
12. processing unit according to claim 11, it is characterised in that first performing module includes:
First execution unit, for if it is determined that the metadata then judges the priority of the data in first chained list
Whether it is high, if so, then the metadata is moved in the 4th chained list;Otherwise, the metadata is moved to described
In second chained list;
Second execution unit, for if it is determined that the metadata then judges the priority of the data in second chained list
Whether it is high, if so, then the metadata is moved in the 4th chained list;Otherwise, the metadata is placed in described
The head of two chained lists;
3rd execution unit, for if it is determined that the metadata is then moved to institute by the metadata in the 3rd chained list
State in the 4th chained list;
4th execution unit, for if it is determined that the metadata is in the 4th chained list, the metadata to be placed in into described
The head of four chained lists.
13. processing unit according to claim 11, it is characterised in that the shadow chained list include the first shadow chained list,
Second shadow chained list and special shadow chained list;
Wherein, the first shadow chained list is used to store the metadata eliminated from first chained list;
The second shadow chained list is used to store the metadata eliminated from second chained list;
When the request that the special shadow chained list is used to store the access data is read request, priority is common and is only interviewed
Asked the metadata corresponding to the data of 1 time.
14. processing unit according to claim 13, it is characterised in that second performing module includes:
5th execution unit, for if it is determined that the metadata then judges the priority of the data in the first shadow chained list
Whether it is high, if so, then the metadata in the first shadow chained list is moved in the 4th chained list, otherwise,
The metadata is moved in second chained list;
6th execution unit, for if it is determined that the metadata then judges the priority of the data in the second shadow chained list
Whether it is high, if so, then the metadata in the second shadow chained list is moved in the 4th chained list, otherwise,
The metadata is moved in second chained list;
7th execution unit, for if it is determined that the metadata is in the special shadow chained list, then by the special shadow chain
The metadata in table is moved in second chained list.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710266779.5A CN107092564A (en) | 2017-04-21 | 2017-04-21 | A kind of data processing method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710266779.5A CN107092564A (en) | 2017-04-21 | 2017-04-21 | A kind of data processing method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107092564A true CN107092564A (en) | 2017-08-25 |
Family
ID=59637899
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710266779.5A Pending CN107092564A (en) | 2017-04-21 | 2017-04-21 | A kind of data processing method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107092564A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107562873A (en) * | 2017-08-31 | 2018-01-09 | 北京小度互娱科技有限公司 | Push the method and device of waterfall flow data |
CN108829354A (en) * | 2018-06-21 | 2018-11-16 | 郑州云海信息技术有限公司 | A kind of solid-state hard disk SSD management method and SSD |
CN109274548A (en) * | 2018-08-22 | 2019-01-25 | 平安普惠企业管理有限公司 | A kind of method for monitoring application program, computer readable storage medium and terminal device |
CN111309796A (en) * | 2020-02-07 | 2020-06-19 | 腾讯科技(深圳)有限公司 | Data processing method and device and computer readable storage medium |
CN113590045A (en) * | 2021-08-03 | 2021-11-02 | 中国联合网络通信集团有限公司 | Data layered storage method, device and storage medium |
CN113778334A (en) * | 2021-08-30 | 2021-12-10 | 武汉市润普网络科技有限公司 | Distributed storage system based on electronic files |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103548004B (en) * | 2011-12-28 | 2016-03-30 | 华为技术有限公司 | The method and apparatus of dynamic data attemper is realized in file system |
US20160261694A1 (en) * | 2008-07-11 | 2016-09-08 | Avere Systems, Inc. | Method and Apparatus for Tiered Storage |
CN106527988A (en) * | 2016-11-04 | 2017-03-22 | 郑州云海信息技术有限公司 | SSD (Solid State Drive) data migration method and device |
-
2017
- 2017-04-21 CN CN201710266779.5A patent/CN107092564A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160261694A1 (en) * | 2008-07-11 | 2016-09-08 | Avere Systems, Inc. | Method and Apparatus for Tiered Storage |
CN103548004B (en) * | 2011-12-28 | 2016-03-30 | 华为技术有限公司 | The method and apparatus of dynamic data attemper is realized in file system |
CN106527988A (en) * | 2016-11-04 | 2017-03-22 | 郑州云海信息技术有限公司 | SSD (Solid State Drive) data migration method and device |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107562873A (en) * | 2017-08-31 | 2018-01-09 | 北京小度互娱科技有限公司 | Push the method and device of waterfall flow data |
CN107562873B (en) * | 2017-08-31 | 2021-02-02 | 北京小度互娱科技有限公司 | Method and device for pushing waterfall flow data |
CN108829354A (en) * | 2018-06-21 | 2018-11-16 | 郑州云海信息技术有限公司 | A kind of solid-state hard disk SSD management method and SSD |
CN109274548A (en) * | 2018-08-22 | 2019-01-25 | 平安普惠企业管理有限公司 | A kind of method for monitoring application program, computer readable storage medium and terminal device |
CN111309796A (en) * | 2020-02-07 | 2020-06-19 | 腾讯科技(深圳)有限公司 | Data processing method and device and computer readable storage medium |
CN111309796B (en) * | 2020-02-07 | 2023-09-26 | 腾讯科技(深圳)有限公司 | Data processing method, device and computer readable storage medium |
CN113590045A (en) * | 2021-08-03 | 2021-11-02 | 中国联合网络通信集团有限公司 | Data layered storage method, device and storage medium |
CN113590045B (en) * | 2021-08-03 | 2023-05-16 | 中国联合网络通信集团有限公司 | Data hierarchical storage method, device and storage medium |
CN113778334A (en) * | 2021-08-30 | 2021-12-10 | 武汉市润普网络科技有限公司 | Distributed storage system based on electronic files |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107092564A (en) | A kind of data processing method and device | |
CN104679594B (en) | A kind of middleware distributed computing method | |
CN102938790B (en) | Resource allocation methods in cloud computing system | |
US8914501B2 (en) | Method, apparatus, and system for scheduling distributed buffer resources | |
US20090138582A1 (en) | Multi-Level Load Balancer | |
CN107196865A (en) | A kind of adaptive threshold overload moving method of Load-aware | |
CN106302702A (en) | Burst storage method, the Apparatus and system of data | |
CN112214307A (en) | Execution of containerization procedures within the constraints of available host nodes | |
CN104978335A (en) | Data access control method and data access control device | |
CN102970379A (en) | Method for realizing load balance among multiple servers | |
CN104331253B (en) | Calculation method for object migration in object storage system | |
CN107729514A (en) | A kind of Replica placement node based on hadoop determines method and device | |
CN108810115A (en) | A kind of load-balancing method, device and server suitable for distributed data base | |
CN108268301A (en) | The virtual machine deployment method and device of virtual network function | |
CN106576260A (en) | Policy coordination method and apparatus in NFV system | |
CN106533961A (en) | Flow control method and device | |
CN104202349B (en) | The method of scheduling distributed buffer resources, Apparatus and system | |
CN104158902B (en) | A kind of Hbase data blocks distribution method and device based on number of request | |
CN108306912A (en) | Virtual network function management method and its device, network function virtualization system | |
CN107943615A (en) | Data processing method and system based on distributed type assemblies | |
CN105657009A (en) | Object gateway load balancing system and method, and cross-region object gateway storage system | |
CN105868002A (en) | Method for processing retransmission request in distributed calculation and device thereof | |
CN105373451B (en) | The method and device that a kind of virtual machine is placed | |
CN104683480A (en) | Distribution type calculation method based on applications | |
CN107688625A (en) | A kind of method and system of the storage small documents based on distributed memory system |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170825 |