CN110300138A - Picture business scheduling method, device, management node and computer readable storage medium - Google Patents
Picture business scheduling method, device, management node and computer readable storage medium Download PDFInfo
- Publication number
- CN110300138A CN110300138A CN201810242805.5A CN201810242805A CN110300138A CN 110300138 A CN110300138 A CN 110300138A CN 201810242805 A CN201810242805 A CN 201810242805A CN 110300138 A CN110300138 A CN 110300138A
- Authority
- CN
- China
- Prior art keywords
- picture
- client
- memory node
- token
- node
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/215—Flow control; Congestion control using token-bucket
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/02—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
- H04L67/61—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources taking into account QoS or priority requirements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Multimedia (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The embodiment of the present invention provides a kind of picture business scheduling method, device, management node and computer readable storage medium, is related to technical field of data storage.The method is applied to the management node with memory node and client communication, which comprises receives the IO information of the disk on the memory node that the memory node reports;The picture scheduling request that the client is sent is received, carries priority corresponding to the picture type of service of the client in the picture scheduling request;The frequency of picture is read from memory node according to client described in the IO information and two priority classes.Use picture business scheduling method, device, management node and the computer readable storage medium, it is ensured that the reliability of reading data.
Description
Technical field
The present invention relates to technical field of data storage, in particular to a kind of picture business scheduling method, device, pipe
Manage node and computer readable storage medium.
Background technique
It is more and more important for the reliable scheduling of picture business in many scenes, such as in monitoring scene, it is sent out through research
It is existing, the characteristics of due to picture business high concurrent, the case where often will appear multiple business type and deposit, for example, picture write-in and
Forwarding exists simultaneously or the business such as picture write-in, forwarding, synthesis exist simultaneously.It will form between different business to resource
It seizes, eventually lead to various businesses while being abnormal.
Summary of the invention
In view of this, the embodiment of the present invention is designed to provide a kind of picture business scheduling method, device, management node
And computer readable storage medium, to improve the reliability of picture traffic scheduling.
In a first aspect, being applied to and memory node and visitor the embodiment of the invention provides a kind of picture business scheduling method
The management node of family end communication, which comprises
Receive the IO information that memory node reports;
The picture scheduling request that client is sent is received, the picture scheduling request is parsed and obtains the picture of the client
Priority corresponding to type of service;
The frequency of picture is read from the memory node according to client described in the IO information and two priority classes.
Optionally, the management node is communicated with multiple memory nodes;
The frequency of picture is read from the memory node according to client described in the IO information and two priority classes
Step includes:
It parses the picture scheduling request and obtains the target storage node where picture to be read;
The client is controlled from the target according to the IO information that the priority and the target storage node report
The frequency of picture is read in memory node.
Optionally, the management node and multiple memory nodes and multiple client communications, it is described according to institute
State the step of client described in IO information and two priority classes reads the frequency of picture from the memory node, comprising:
Token bucket is respectively created for each memory node;
According to the IO information that each memory node reports, the token number in the token bucket of the memory node is determined;
It is corresponding excellent according to each client for reading the picture on the memory node for each memory node
Token in the token bucket of the memory node is distributed to each client by first grade, to control each client from described
The frequency of picture is read in memory node.
Optionally, the method also includes: be directed to each memory node, order in the token bucket of the memory node
When board number is 0, access of the client to the memory node is limited.
Optionally, the IO information reported according to each memory node, determines the order in the token bucket of the memory node
The step of board number, comprising:
For each memory node, token quantity corresponding with the IO information of the memory node is found out;
Judge to increase the token number after the token quantity in the token bucket of the memory node and whether be greater than and deposits in advance
Token number in the token bucket is remained the maximum value if more than the maximum value by the maximum value of storage;If being less than described
Maximum value increases the token of the token quantity in the token bucket of Xiang Suoshu memory node.
Optionally, for each memory node, according to each client for reading the picture on the memory node
Token in the token bucket of the memory node is distributed to each client by corresponding priority, to control each client
The step of holding the frequency that picture is read from the memory node, comprising:
For each memory node, the corresponding priority of each client for needing to access the memory node is obtained,
Obtain needing to access the classification of the corresponding priority of all clients of the memory node;
Inquiry obtains token number distribution weight corresponding with the priority of each classification;
According to the token number and token number distribution weight in the token bucket of the memory node, calculate
To distribution token quantity corresponding with the priority of each classification;
According to the quantity of the corresponding client of the priority of each classification, the priority with each classification that will be calculated
Corresponding distribution token quantity divides equally the token quantity that single client distribution is obtained to each client, controls each visitor
Family end carries out picture reading according to the token distributed within the setting period.
Optionally, the method also includes:
It detects whether there is the picture scheduling request being not carried out in the preset duration after distributing token, is not carried out if it exists
Picture scheduling request deletes the picture scheduling request, re-starts token distribution.
Second aspect, the embodiment of the invention provides a kind of picture traffic scheduling devices, are applied to and memory node and visitor
The management node of family end communication, the picture traffic scheduling device include:
Information receiving module, the IO information reported for receiving memory node;
Request receiving module parses the picture scheduling request and obtains for receiving the picture scheduling request of client transmission
Priority corresponding to picture type of service to the client;
Control module is read, for the client according to the IO information and two priority classes from the memory node
Read the frequency of picture.
The third aspect, the embodiment of the invention provides a kind of management nodes, comprising:
Memory;
Processor;And
Picture traffic scheduling device, the picture traffic scheduling device include:
Information receiving module, the IO information reported for receiving memory node;
Request receiving module parses the picture scheduling request and obtains for receiving the picture scheduling request of client transmission
Priority corresponding to picture type of service to the client;
Control module is read, for the client according to the IO information and two priority classes from the memory node
Read the frequency of picture.
Fourth aspect, the embodiment of the invention provides a kind of computer readable storage medium, the computer-readable storage
Medium includes computer program, and the computer program controls management node where the computer readable storage medium when running
Execute above-mentioned picture business scheduling method.
Picture business scheduling method, device, management node and computer readable storage medium provided in an embodiment of the present invention,
Management node is adjusted according to the IO information of the disk on memory node and the corresponding priority of the picture type of service of client
Different picture service customer ends are to the access frequency of the disk in memory node, thus being based on each memory node of IO information realization
Picture traffic balancing scheduling while, by priority division ensure that the higher scheduling business of priority can reliably be held
Row.
To enable the above objects, features and advantages of the present invention to be clearer and more comprehensible, preferred embodiment is cited below particularly, and cooperate
Appended attached drawing, is described in detail below.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair
The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 is a kind of block diagram of management node provided in an embodiment of the present invention.
Fig. 2 is a kind of flow chart of picture business scheduling method provided in an embodiment of the present invention.
Fig. 3 is the exemplary process diagram of step S13 in Fig. 2 provided in an embodiment of the present invention.
Fig. 4 is the exemplary process diagram of step S13 in Fig. 2 provided in an embodiment of the present invention.
Fig. 5 is the exemplary process diagram of step S135 in Fig. 4 provided in an embodiment of the present invention.
Fig. 6 is the exemplary block diagram of picture traffic scheduling device provided in an embodiment of the present invention.
Icon: 10- management node;11- memory;12- processor;13- network module;20- picture traffic scheduling device;
21- information receiving module;22- request receiving module;23- reads control module.
Specific embodiment
In many scenes, it is required to using to picture traffic scheduling scheme, for example, with the development of monitoring trade,
In current monitoring business, it is ensured that the reliable calling of picture business is more and more important.Picture type of business is more, can generally divide
Checked for picture, Intelligent data analysis, forwarding, write-in, picture synthesis, third party's client read etc. types of service.
The characteristics of due to picture business high concurrent, the case where often will appear multiple business type and deposit.For picture
Reading business, there are many implementations, for example, can pass through hypertext transfer protocol (Hyper Text in monitoring business
Transfer Protocol, http) image data is accessed, the server-side of http can be deployed in traffic exchange of media
On (Traffic Media Switch, TMS) server, different clients is sent on http message to TMS server, by
TMS server goes on memory node to read picture and returns to client, wherein TMS server can be responsible for connecing for image data
It receives, forward, storage and URL enter the work such as database.In the case where the client for sending picture scheduling request is more, if not
Picture business is limited, then will form between different business and resource is seized, eventually lead to various businesses while occurring
It is abnormal, it cannot achieve the reliable scheduling of picture business.
It has been investigated that in conjunction with the characteristics of picture business, multiple picture business are simultaneously in current monitoring cloud storage system
The performance bottleneck point of hair is the pressure of disk access, i.e., to the read-write operation of disk.To single memory node within the unit time
Access frequency with no restrictions in the case where, it is possible to occur to the very high situation of the access frequency of single memory node, if
Third party's client access memory node frequency is excessively high to will lead to picture forwarding, client in BS (Browser/Server) client
Check the business such as picture time-out.In order to solve this problem, if it is desired to which third party's client or certain service customer ends are reduced and visited
Ask frequency, clearly can not specifically limit specification in many cases, for example, there are the picture access of third party's client or
Intellectual analysis business, it is uncertain in third party's client, the corresponding number of clients of intellectual analysis business and access frequency
In the case of, the specification of concrete restriction can not be specified.If restriction rule is arranged in client-side, it is not general there is restriction rule
The problem of, if only going to be limited according to fixed rule, can not accomplish according to the practical networking in scene and combinations of services mode
It realizes Automatic dispatching, system resource cannot be made full use of.If by control by memory node overall data flow come into
Row access limitation, can not preferentially guarantee the normal use of important picture business.
Based on the studies above, the embodiment of the invention provides a kind of picture traffic scheduling schemes, to different picture business
Priority division is carried out, it is corresponding preferential according to the picture type of service of the IO information of the disk on memory node and client
Grade adjusts different picture service customer ends to the access frequency of the disk in memory node, to respectively deposit based on IO information realization
While storing up the picture traffic balancing scheduling of node, ensure that the higher scheduling business of priority can be by can by priority division
By executing.In the embodiment of the present invention, IO information can be the iowait information of disk.
For defect present in the prior art, be inventor being obtained after practicing and carefully studying as a result,
Therefore, the solution party that the discovery procedure of the above problem and the embodiment of the present invention are proposed for problem in the presence of the prior art
Case all should be the contribution that inventor makes the present invention in process of the present invention.
Below in conjunction with attached drawing in the embodiment of the present invention, technical solution in the embodiment of the present invention carries out clear, complete
Ground description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Usually exist
The component of the embodiment of the present invention described and illustrated in attached drawing can be arranged and be designed with a variety of different configurations herein.Cause
This, is not intended to limit claimed invention to the detailed description of the embodiment of the present invention provided in the accompanying drawings below
Range, but it is merely representative of selected embodiment of the invention.Based on the embodiment of the present invention, those skilled in the art are not doing
Every other embodiment obtained under the premise of creative work out, shall fall within the protection scope of the present invention.
It should also be noted that similar label and letter indicate similar terms in following attached drawing, therefore, once a certain Xiang Yi
It is defined in a attached drawing, does not then need that it is further defined and explained in subsequent attached drawing.
As shown in Figure 1, being a kind of block diagram of management node 10 provided in an embodiment of the present invention.The embodiment of the present invention
In management node 10 can equipment for metadata management equipment etc. with Data Analysis Services and forwarding capability.Such as Fig. 1 institute
Show, management node 10 includes: memory 11, processor 12, network module 13 and picture traffic scheduling device 20.
The memory 11, processor 12 and network module 13 are directly or indirectly electrically connected between each other, with reality
The transmission or interaction of existing data.For example, these elements can be realized between each other by one or more communication bus or signal wire
It is electrically connected.Picture traffic scheduling device 20 is stored in memory 11, the picture traffic scheduling device 20 includes at least one
A software function module that can be stored in the form of software or firmware (firmware) in the memory 11, the processor
12 software programs and module being stored in memory 11 by operation, such as the picture traffic scheduling dress in the embodiment of the present invention
20 are set, thereby executing various function application and data processing, i.e. picture business scheduling method in the realization embodiment of the present invention.
Wherein, the memory 11 may be, but not limited to, random access memory (Random Access Memory,
RAM), read-only memory (Read Only Memory, ROM), programmable read only memory (Programmable Read-Only
Memory, PROM), erasable read-only memory (Erasable Programmable Read-Only Memory, EPROM),
Electricallyerasable ROM (EEROM) (Electric Erasable Programmable Read-Only Memory, EEPROM) etc..
Wherein, memory 11 is for storing program, and the processor 12 executes described program after receiving and executing instruction.
The processor 12 may be a kind of IC chip, the processing capacity with data.Above-mentioned processor 12
It can be general processor, including central processing unit (Central Processing Unit, CPU), network processing unit
(Network Processor, NP) etc..It may be implemented or execute each method, step disclosed in the embodiment of the present invention and patrol
Collect block diagram.General processor can be microprocessor or the processor is also possible to any conventional processor etc..
Network module 13 is used to establish management node 10 and external communications terminals by network, such as client and memory node
Between communication connection, realize the transmitting-receiving operation of network signal and data.Above-mentioned network signal may include wireless signal or have
Line signal.
It is appreciated that structure shown in FIG. 1 is only to illustrate, management node 10 may also include it is more than shown in Fig. 1 or
Less component, or with the configuration different from shown in Fig. 1.Each component shown in Fig. 1 can using hardware, software or its
Combination is realized.
The embodiment of the present invention also provides a kind of computer readable storage medium, including computer program.The computer journey
Management node 10 where controlling the computer readable storage medium when sort run executes following picture business scheduling method.
Please refer to Fig. 2, business mainly read to picture in the embodiment of the present invention and is dispatched, management node 10 by with
Lower step realizes picture traffic scheduling.
Step S11 receives the IO information that memory node reports.
In the embodiment of the present invention, it is illustrated so that IO information is the iowait information of disk as an example.Memory node can lead to
Cross the iowait information for the disk that iostat order is obtained in real time on the memory node, and the iowait information that will be got in real time
Average value be periodically reported to management node.The average value of iowait information is reported to the mode of management node by memory node
There are many, it is reported in the interaction message with management node for example, the average value of iowait information can be carried, example
Such as, the average value of iowait information can be carried and is reported to management node in capability reporting message.In capability reporting message
It may include the information such as the loading condition of memory node.
Correspondingly, it after management node receives the capability reporting message that memory node periodically reports, parses in the ability
Message is reported, to obtain the iowait information that memory node reports.
Step S12 receives the picture scheduling request that client is sent, parses the picture scheduling request and obtain the client
Priority corresponding to the picture type of service at end.
, requirement of real-time difference different according to the action type of picture business, gives the corresponding client of different picture business
Distribute different priority.Priority is higher, and the resource use priority for carrying out picture traffic scheduling is higher.
It include that picture Web is checked, picture forwarding is read, picture synthesis is read, intellectual analysis is read with picture type of service
For reading picture with third party, the corresponding priority of different picture types of service and weight specific gravity can be as shown in table 1.
Table 1
Wherein, the picture type of service highest priority that allocation level is 1, the picture type of service that allocation level is 5 are excellent
First grade is minimum.
Each client is used for the different types of picture business of management node request scheduling, and client is sent to management node
When picture scheduling request, priority corresponding to the picture type of service of characterization client need to be carried in picture scheduling request
Content, for example, the rank field of characterization priority can be carried in picture scheduling request.
Correspondingly, management node can be obtained corresponding to the picture type of service of client by parsing picture scheduling request
Priority.
In the embodiment of the present invention, the type of client is unlimited, can be image point for example, can be third party's client
Analyse (Image Analysis, IA) equipment etc., can by the difference of IP address (Internet Protocol Address),
Distinguish different clients.
According to different network environments, the combination of picture business can there are many, for example, included under certain network environment
Picture type of service can more than the type enumerated in table 1, in another example, the included picture service class under certain network environment
Type can only include the wherein several types enumerated in table 1.The present embodiment to this with no restriction.Included by different network environments
In the case that picture type of service is different, correspondingly, the corresponding weight specific gravity of different picture types of service can be also adaptively adjusted,
The sum of corresponding weight specific gravity of all picture types of service included by same network environment is not more than 100%.According to reality
Border demand carries out networking to various picture types of service, and the mode for carrying out priority setting to different picture types of service is flexibly more
Sample, in the present embodiment with no restriction to this.
Step S13 reads picture from the memory node according to client described in the IO information and two priority classes
Frequency.
In the case where management node is communicated with multiple memory nodes, management node receives each can be on memory node
The iowait information of report.Management node is provided with the iowait critical value range of the disk on memory node, corresponding division side
There are many formulas, and the present embodiment lists one of division mode by table 2.
Influence performance factor | It is good | It is bad | It is bad |
Disk iowait% | < 20% | 20%~50% | > 50% |
Table 2
Fig. 3 is please referred to, step S13 can be accomplished by the following way.
Step S131, management node parsing picture scheduling request obtain the target storage node where picture to be read.
Wherein, the memory node where the picture that management node can contain in the URL by picture to be read is practical
Coding, obtains the target storage node where picture to be read.
Step S132, the IO information governance client reported according to the corresponding priority of client and target storage node from
The frequency of picture is read in target storage node.
In the case where management node is communicated with multiple storages and multiple client, management node can also record different visitors
The corresponding priority in family end is different figures according to the difference of the priority carried in the picture scheduling request of each client transmission
The client of piece type of service distributes different resources, guarantees the operation of the picture business of high priority.
Fig. 4 is please referred to, the picture business of one of client for adjusting different priorities is present embodiments provided
The scheme of reading frequency controls client reading frequency using token bucket algorithm by management node, and implementation is as follows.
Token bucket is respectively created for each memory node in step S133.
Step S134 is determined in the token bucket of the memory node according to the IO information that each memory node reports
Token number.
Each picture is read, and is required into the token bucket of memory node where picture acquisition token and is read out visit
It asks.
In the present embodiment, the critical value range according to locating for iowait determines the token in the token bucket of memory node
Number, for example, can preset iowait is in corresponding token number when different critical value range.Table 3 is please referred to,
The present embodiment lists token number method of determination corresponding with the critical value range division mode in table 2.
Report iowait value | Distribute token number |
< 20% | A |
20%~50% | B |
> 50% | C |
Table three
Wherein, the numerical value of A, B and C can flexibly be set, and the value of A is greater than the value of B, and the value of B is greater than the value of C.For example, A
Value can be 10000, correspondingly, if iowait information < 20% of memory node, then dividing into the token bucket of memory node
With 10000 tokens, thus, within the single iowait information reporting period, 10000 figures can be read on the memory node
Piece.In another example the value of C can be 0, correspondingly, if iowait information > 50% of memory node, then to memory node
0 token is distributed in token bucket, thus, within the single iowait information reporting period, 0 can be read on the memory node
Picture.For each memory node, when the token number in the token bucket of the memory node is 0, the client is limited
Hold the access to the memory node.
It should be appreciated that the critical value range division mode of iowait information can also have other, the iowait of memory node
Information when being located at different critical value range corresponding token number can also flexibly set, for example, according to memory node server
Parameter, performance is different and iowait information reporting period difference is correspondingly configured, the present embodiment does not limit this
System.
First iowait information reporting period, the token number in the token bucket of each memory node is to believe with iowait
The corresponding token number of the critical value range of breath, in view of within an iowait information reporting period, the token of each memory node
Token in bucket may have residue, thus, in next iowait information reporting period, will believe with the iowait of each memory node
The corresponding token number of the critical value range of breath is added separately in the token bucket of respective stored node, if certain memory node is upper
Token in one iowait information reporting period in token bucket has residue, then, the memory node is in current iowait information
Token number in report cycle in token bucket is remaining token number and critical with the iowait information of the memory node
It is worth the sum of corresponding token number of range.
In order to avoid the token number in the token bucket of memory node excessively normally reads to influence picture, optionally,
For each memory node, after finding out token quantity corresponding with the iowait information of the memory node, judge to increase
Add whether the token number after the token quantity in the token bucket of the memory node is greater than pre-stored maximum value, if greatly
In the maximum value, the token number in the token bucket is remained into the maximum value.If being less than the maximum value, Xiang Suoshu is deposited
Store up the token for increasing the token quantity in the token bucket of node.Wherein, maximum value can flexibly be set, as long as not influencing single
Iowait information reporting cycle memory stores up the reading of the picture on node.
Step S135, for each memory node, according to each client for reading the picture on the memory node
Corresponding priority is held, the token in the token bucket of the memory node is distributed into each client, to control each visitor
The frequency of picture is read from the memory node in family end.
Fig. 5 is please referred to, can be realized by following steps to the token distribution of each client and picture reading frequency
Control.
Step S1351, for each memory node, each client that acquisition needs to access the memory node is corresponding
Priority, obtain the classification for the corresponding priority of all clients for needing to access the memory node.
Be exemplified by Table 1, it is assumed that some memory node, need to access the memory node client have it is multiple, access should
The corresponding priority 1~5 of the client of memory node has, then, all clients for needing to access the memory node are corresponding
Priority have 5 classifications.
Step S1352, inquiry obtain token number distribution weight corresponding with the priority of each classification.
In the present embodiment, the priority of each classification corresponding token number distribution weight can according to actual needs, with
And the difference of network environment flexibly setting.
Step S1353, according to the token number and token number distribution power in the token bucket of the memory node
Weight, is calculated distribution token quantity corresponding with the priority of each classification.
If the token number in the token bucket of some memory node is M, the corresponding token of the priority of one of classification
Number assignment weight is N, then, the corresponding distribution token quantity of the priority of the category is M multiplied by N.
Step S1354, according to the quantity of the corresponding client of the priority of each classification, by being calculated with each class
The corresponding distribution token quantity of other priority divides equally the token quantity that single client distribution is obtained to each client, control
It makes each client and picture reading is carried out according to the token distributed within the setting period.
If the corresponding distribution token quantity of the priority that some classification is calculated is E, priority is the client of the category
End quantity is F, then, the token quantity that each client of the category is assigned to is E divided by F.
As shown in table 4, the embodiment of the invention provides the corresponding priority of client for accessing a certain memory node four
Class, priority are that different classes of client terminal quantity is different, and the token number of the token bucket of the memory node is 10000, each visitor
One of token method of salary distribution at family end.
Table 4
By table 4 it is found that in the token bucket for learning memory node in the case where token number, according to different classes of excellent
The corresponding weight of first grade, the corresponding token quantity of priority that each classification can be calculated multiplied by weight in token number,
Further according to the identical client terminal quantity of priority class, by the corresponding token quantity of the priority of each classification divided by the category
The token number of each client distribution can be calculated in the client terminal quantity of priority.
Step S14 detects whether there is the picture scheduling request being not carried out in the preset duration after distributing token, if it exists
The picture scheduling request being not carried out deletes the picture scheduling request, re-starts token distribution.
If be not carried out in a certain picture scheduling request long-time, management node if, deletes its aging, deletes the figure
After piece scheduling request, weight distribution is re-started for remaining picture scheduling request and distribution token number calculates, avoids wasting
Idling-resource.
Using the scheme in the embodiment of the present invention, when client needs to read the picture on memory node, saved by management
Point obtains the token of memory node, and management node obtains token from the token bucket of memory node and is sent to client, client
End can directly read picture according to token from memory node, if enabled within the iowait information reporting period of memory node
Token in board bucket runs out, then needs update the token number in token bucket equal iowait information reporting period next time, and enable
When board number is not 0, client just has an opportunity to be assigned to token, and then completes picture and read.
On the basis of the above, as shown in fig. 6, the embodiment of the invention provides a kind of picture traffic scheduling device 20, application
In the management node with memory node and client communication, the picture traffic scheduling device 20 include information receiving module 21,
Request receiving module 22 and reading control module 23.
Wherein, information receiving module 21 is for receiving the IO information that memory node reports.
Since information receiving module 21 is similar with the realization principle of step S11 in Fig. 2, thus do not illustrate more herein.
Request receiving module 22 is used to receive the picture scheduling request of client transmission, parses the picture scheduling request and obtains
Priority corresponding to picture type of service to the client.
Since request receiving module 22 is similar with the realization principle of step S12 in Fig. 2, thus do not illustrate more herein.
It reads control module 23 and is used for the client according to the IO information and two priority classes from the memory node
The middle frequency for reading picture.
It is similar with the realization principle of step S13 in Fig. 2 due to reading control module 23, thus do not illustrate more herein.
Picture business scheduling method, device, management node and computer readable storage medium in the embodiment of the present invention lead to
It crosses management node and Real-Time Scheduling is carried out to picture business, so that the upper access frequency of iowait relatively low memory node can increase
Adding, the upper access frequency of iowait relatively high memory node can reduce, thus realize that the reading of each memory node is balanced,
Ensure the reliability of picture traffic scheduling.
In several embodiments provided by the embodiment of the present invention, it should be understood that disclosed device and method, it can also
To realize by another way.Device and method embodiment described above is only schematical, for example, in attached drawing
Flow chart and block diagram show that the devices of multiple embodiments according to the present invention, method and computer program product are able to achieve
Architecture, function and operation.In this regard, each box in flowchart or block diagram can represent module, a program
A part of section or code, a part of the module, section or code include that one or more is patrolled for realizing defined
Collect the executable instruction of function.It should also be noted that in some implementations as replacement, function marked in the box
It can occur in a different order than that indicated in the drawings.For example, two continuous boxes can actually be held substantially in parallel
Row, they can also be executed in the opposite order sometimes, and this depends on the function involved.It is also noted that block diagram and/or
The combination of each box in flow chart and the box in block diagram and or flow chart, can the function as defined in executing or dynamic
The dedicated hardware based system made is realized, or can be realized using a combination of dedicated hardware and computer instructions.
In addition, each functional module in each embodiment of the present invention can integrate one independent portion of formation together
Point, it is also possible to modules individualism, an independent part can also be integrated to form with two or more modules.
It, can be with if the function is realized and when sold or used as an independent product in the form of software function module
It is stored in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially in other words
The part of the part that contributes to existing technology or the technical solution can be embodied in the form of software products, the meter
Calculation machine software product is stored in a storage medium, including some instructions are used so that a computer equipment (can be a
People's computer, management node or network equipment etc.) execute all or part of step of each embodiment the method for the present invention
Suddenly.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), deposits at random
The various media that can store program code such as access to memory (RAM, Random Access Memory), magnetic or disk.
It should be noted that, in this document, the terms "include", "comprise" or its any other variant are intended to the packet of nonexcludability
Contain, so that the process, method, article or equipment for including a series of elements not only includes those elements, but also including
Other elements that are not explicitly listed, or further include for elements inherent to such a process, method, article, or device.
In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including the element
Process, method, article or equipment in there is also other identical elements.
The foregoing is merely alternative embodiments of the invention, are not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of picture business scheduling method, applied to the management node with memory node and client communication, which is characterized in that
The described method includes:
Receive the IO information that memory node reports;
The picture scheduling request that client is sent is received, the picture scheduling request is parsed and obtains the picture business of the client
Priority corresponding to type;
The frequency of picture is read from the memory node according to client described in the IO information and two priority classes.
2. picture business scheduling method according to claim 1, which is characterized in that the management node described is deposited with multiple
Store up node communication;
The step of frequency of picture is read from the memory node according to client described in the IO information and two priority classes
Include:
It parses the picture scheduling request and obtains the target storage node where picture to be read;
The IO information reported according to the priority and the target storage node controls the client and stores from the target
The frequency of picture is read in node.
3. picture business scheduling method according to claim 1, which is characterized in that the management node described is deposited with multiple
Store up node and multiple client communications, the client according to the IO information and two priority classes is from the storage
The step of frequency of picture is read in node, comprising:
Token bucket is respectively created for each memory node;
According to the IO information that each memory node reports, the token number in the token bucket of the memory node is determined;
It is corresponding preferential according to each client for reading the picture on the memory node for each memory node
Grade, distributes to each client for the token in the token bucket of the memory node, is deposited with controlling each client from described
Store up the frequency that picture is read in node.
4. picture business scheduling method according to claim 3, which is characterized in that the method also includes: for each
The memory node limits the client to the memory node when token number in the token bucket of the memory node is 0
Access.
5. picture business scheduling method according to claim 3, which is characterized in that reported according to each memory node
IO information, the step of determining the token number in the token bucket of the memory node, comprising:
For each memory node, token quantity corresponding with the IO information of the memory node is found out;
Judge to increase the token number after the token quantity in the token bucket of the memory node whether be greater than it is pre-stored
Token number in the token bucket is remained the maximum value if more than the maximum value by maximum value;If being less than the maximum
It is worth, increases the token of the token quantity in the token bucket of Xiang Suoshu memory node.
6. picture business scheduling method according to claim 3, which is characterized in that be directed to each memory node, root
According to the corresponding priority of each client of the picture read on the memory node, by the order in the token bucket of the memory node
The step of board distributes to each client, reads the frequency of picture to control each client from the memory node,
Include:
For each memory node, the corresponding priority of each client for needing to access the memory node is obtained, is obtained
Need to access the classification of the corresponding priority of all clients of the memory node;
Inquiry obtains token number distribution weight corresponding with the priority of each classification;
According in the token bucket of the memory node token number and the token number distribute weight, be calculated with
The corresponding distribution token quantity of the priority of each classification;
It is corresponding with the priority of each classification by what is be calculated according to the quantity of the corresponding client of the priority of each classification
Distribution token quantity divide equally the token quantity that single client distribution is obtained to each client, control each client
Picture reading is carried out according to the token distributed within the setting period.
7. according to the described in any item picture business scheduling methods of claim 3~6, which is characterized in that the method also includes:
Detect whether there is the picture scheduling request being not carried out in the preset duration after distributing token, the picture being not carried out if it exists
Scheduling request deletes the picture scheduling request, re-starts token distribution.
8. a kind of picture traffic scheduling device, applied to the management node with memory node and client communication, which is characterized in that
The picture traffic scheduling device includes:
Information receiving module, the IO information reported for receiving memory node;
Request receiving module parses the picture scheduling request and obtains institute for receiving the picture scheduling request of client transmission
State priority corresponding to the picture type of service of client;
Control module is read, is read from the memory node for the client according to the IO information and two priority classes
The frequency of picture.
9. a kind of management node characterized by comprising
Memory;
Processor;And
Picture traffic scheduling device, the picture traffic scheduling device include:
Information receiving module, the IO information reported for receiving memory node;
Request receiving module parses the picture scheduling request and obtains institute for receiving the picture scheduling request of client transmission
State priority corresponding to the picture type of service of client;
Control module is read, is read from the memory node for the client according to the IO information and two priority classes
The frequency of picture.
10. a kind of computer readable storage medium, which is characterized in that the computer readable storage medium includes computer journey
Sequence, management node perform claim requires 1~7 the computer program controls the computer readable storage medium when running where
Described in any item picture business scheduling methods.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810242805.5A CN110300138B (en) | 2018-03-22 | 2018-03-22 | Picture service scheduling method, device, management node and computer readable storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810242805.5A CN110300138B (en) | 2018-03-22 | 2018-03-22 | Picture service scheduling method, device, management node and computer readable storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110300138A true CN110300138A (en) | 2019-10-01 |
CN110300138B CN110300138B (en) | 2022-04-08 |
Family
ID=68025795
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810242805.5A Active CN110300138B (en) | 2018-03-22 | 2018-03-22 | Picture service scheduling method, device, management node and computer readable storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110300138B (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101414269A (en) * | 2007-04-09 | 2009-04-22 | 英特尔公司 | Priority based throttling for power/performance quality of service |
US20110246730A1 (en) * | 2010-04-06 | 2011-10-06 | Fujitsu Limited | Computer-readable medium storing storage control program, storage control method, and storage control device |
US20140112147A1 (en) * | 2012-10-19 | 2014-04-24 | Broadcom Corporation | Refresh mechanism for a token bucket |
CN103761051A (en) * | 2013-12-17 | 2014-04-30 | 北京同有飞骥科技股份有限公司 | Performance optimization method for multi-input/output stream concurrent writing based on continuous data |
CN104765577A (en) * | 2015-04-28 | 2015-07-08 | 杭州中天微系统有限公司 | High-speed storage system achieving self-adaptive frequency |
CN106559683A (en) * | 2016-11-16 | 2017-04-05 | 浙江宇视科技有限公司 | The QoS priority dynamic regulating method of monitoring business and device |
US9703602B1 (en) * | 2015-03-31 | 2017-07-11 | Amazon Technologies, Inc. | Burst throttling for multi-tenant storage services |
-
2018
- 2018-03-22 CN CN201810242805.5A patent/CN110300138B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101414269A (en) * | 2007-04-09 | 2009-04-22 | 英特尔公司 | Priority based throttling for power/performance quality of service |
US20110246730A1 (en) * | 2010-04-06 | 2011-10-06 | Fujitsu Limited | Computer-readable medium storing storage control program, storage control method, and storage control device |
US20140112147A1 (en) * | 2012-10-19 | 2014-04-24 | Broadcom Corporation | Refresh mechanism for a token bucket |
CN103761051A (en) * | 2013-12-17 | 2014-04-30 | 北京同有飞骥科技股份有限公司 | Performance optimization method for multi-input/output stream concurrent writing based on continuous data |
US9703602B1 (en) * | 2015-03-31 | 2017-07-11 | Amazon Technologies, Inc. | Burst throttling for multi-tenant storage services |
CN104765577A (en) * | 2015-04-28 | 2015-07-08 | 杭州中天微系统有限公司 | High-speed storage system achieving self-adaptive frequency |
CN106559683A (en) * | 2016-11-16 | 2017-04-05 | 浙江宇视科技有限公司 | The QoS priority dynamic regulating method of monitoring business and device |
Non-Patent Citations (1)
Title |
---|
BINGGOOGLE: "QOS以及QOS中的令牌桶机制", 《CSDN》 * |
Also Published As
Publication number | Publication date |
---|---|
CN110300138B (en) | 2022-04-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10855545B2 (en) | Centralized resource usage visualization service for large-scale network topologies | |
CN106899680B (en) | The fragment treating method and apparatus of multi-tiling chain | |
US7523454B2 (en) | Apparatus and method for routing a transaction to a partitioned server | |
CN108776934B (en) | Distributed data calculation method and device, computer equipment and readable storage medium | |
Chaczko et al. | Availability and load balancing in cloud computing | |
US9647904B2 (en) | Customer-directed networking limits in distributed systems | |
CN102347876B (en) | Multilink aggregation control device for cloud computing network | |
US9258197B2 (en) | Prioritizing service requests | |
CN109167812B (en) | Method for evaluating service quality and determining adjustment strategy, server and storage medium | |
CN109656685A (en) | Container resource regulating method and system, server and computer readable storage medium | |
CN109039787A (en) | log processing method, device and big data cluster | |
CN112685287B (en) | Product data testing method and device, storage medium and electronic device | |
CN110417886A (en) | A kind of load-balancing method of integrated service, apparatus and system | |
CN105872082B (en) | Fine granularity resource response system based on container cluster load-balancing algorithm | |
CN109286617B (en) | Data processing method and related equipment | |
CN114430383A (en) | Method and device for screening detection nodes, electronic equipment and storage medium | |
CN116760655B (en) | POP point method for providing CPE optimal access in SD-WAN application | |
CN108063814B (en) | Load balancing method and device | |
CN112102063B (en) | Data request method, device, equipment, platform and computer storage medium | |
WO2023088362A1 (en) | Network traffic processing method and apparatus, and medium and electronic device | |
CN110300138A (en) | Picture business scheduling method, device, management node and computer readable storage medium | |
CN112001454B (en) | Clustering method and device for special line for collecting guests | |
Gilesh et al. | Resource availability–aware adaptive provisioning of virtual data center networks | |
US20060179342A1 (en) | Service aggregation in cluster monitoring system with content-based event routing | |
CN109218463B (en) | Acquisition system maintenance method and acquisition 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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |