CN105262623B - Resource allocation methods and device for the software systems based on service - Google Patents

Resource allocation methods and device for the software systems based on service Download PDF

Info

Publication number
CN105262623B
CN105262623B CN201510714321.2A CN201510714321A CN105262623B CN 105262623 B CN105262623 B CN 105262623B CN 201510714321 A CN201510714321 A CN 201510714321A CN 105262623 B CN105262623 B CN 105262623B
Authority
CN
China
Prior art keywords
resource allocation
resource
parameter
parameter set
allocation program
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201510714321.2A
Other languages
Chinese (zh)
Other versions
CN105262623A (en
Inventor
张岳松
启哲
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
NEUSOFT XIKANG HEALTH TECHNOLOGY Co Ltd
Original Assignee
NEUSOFT XIKANG HEALTH TECHNOLOGY Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by NEUSOFT XIKANG HEALTH TECHNOLOGY Co Ltd filed Critical NEUSOFT XIKANG HEALTH TECHNOLOGY Co Ltd
Priority to CN201510714321.2A priority Critical patent/CN105262623B/en
Publication of CN105262623A publication Critical patent/CN105262623A/en
Application granted granted Critical
Publication of CN105262623B publication Critical patent/CN105262623B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08—Configuration management of networks or network elements
    • H04L41/0893—Assignment of logical groups to network elements
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5003—Managing SLA; Interaction between SLA and QoS
    • H04L41/5019—Ensuring fulfilment of SLA
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5041—Network service management, e.g. ensuring proper service fulfilment according to agreements characterised by the time relationship between creation and deployment of a service

Abstract

The invention discloses a kind of resource allocation methods and device for the software systems based on service.This method includes：Multiple and different resource allocation programs of the Component service for the system are obtained, form initial resource configuration plan set, wherein each resource allocation program includes the resource allocation strategy of Component service；Each resource allocation program in initial resource configuration plan set is divided according to dominance hierarchy, forms multiple resource allocation program subsets；According to the resource allocation program subset with highest dominance hierarchy, target resource configuration plan set is obtained；The target resource configuration strategy obtained in plan set for Component service is configured from target resource；And resource is distributed for Component service according to target resource configuration strategy.Target resource configuration strategy can be quickly determined as a result, keep the resource that ISP consumes minimum.Further, it is also possible to realize the resource allocation based on multiple target, the flexibility of the resource allocation of Component service is improved.

Description

Resource allocation methods and device for the software systems based on service
Technical field
The present invention relates to computer realms, and in particular, to a kind of resource allocation for the software systems based on service Method and device.
Background technology
In order to flexibly adapt to the change of business, software systems is further made to have flexibly configurable, dynamic restructuring The features such as, more and more distributed software systems use Enterprise SOA (Service-Oriented Architecture, abbreviation SOA).This kind of system based on SOA structures is also referred to as the software systems (service- based on service Based software system, abbreviation SBS).This kind of system by several Component services and specified services execution sequence control Flow composition processed.For example, Fig. 1 shows the configuration diagram of an example SBS, as shown in Figure 1, the SBS systems may include five A Component service is Component service S1, Component service S2, Component service S3, Component service S4 and Component service S5 respectively.And And the control flow of specified services execution sequence can be as illustrated by the arrows in fig. 1.
Service-level agreement (service level agreement, abbreviation SLA) defines the Service Quality for the system of being applied to (QoS) demand of measuring, for example, handling capacity, delay time, required cost etc..In order to meet given SLA, need to correspond to for SBS Workflow choose Component service appropriate, and distribute the clothes that computing resource appropriate reaches SLA requirements for Component service Business quality.Therefore, how effectively to distribute computing resource for SBS is the key that operation SBS under cloud environment.
The resource allocation methods for being presently used for SBS are that the corresponding computing resource of each Component service is divided into different grains Degree, and find the quality of the corresponding Component service of varigrained resource.Later, by searching for different resource allocation and group Part services corresponding quality, finds the Component service that disclosure satisfy that SLA and its corresponding resource allocation to build SBS examples. Such methods have two：
First, as the granularity of resource allocation and QoS partition is thinner, the resource allocation for meeting SLA is searched out The required time is bigger.Assuming that a SBS is made of m Component service, the cpu resource of each Component service is divided into k Grade, then it is O (m to search for the required timek), complexity is exponential.Obviously, it is difficult in many cases to the problem It is solved.
Secondly, qos parameter is normalized to a utility function by current method, and then is selected and can be made utility function Maximum Component service and the corresponding resource allocation of service.This processing mode have ignored qos parameter be mutually it is conflicting, For example, handling capacity is higher, resource allocation is bigger, and corresponding price is also higher.Therefore, simultaneously for mutual conflicting qos parameter There is no unique solutions.
Invention content
The object of the present invention is to provide a kind of resource allocation methods and device for the software systems based on service, with reality The now resource allocation based on multiple target.
To achieve the goals above, the present invention provides a kind of resource allocation methods for the software systems based on service, This method includes：Multiple and different resource allocation programs of the Component service for the system are obtained, initial resource configuration is formed Plan set, wherein each resource allocation program includes the resource allocation strategy of the Component service；The initial resource is matched The each resource allocation program set in plan set is divided according to dominance hierarchy, forms multiple resource allocation program subsets, Wherein, the resource allocation program dominance hierarchy having the same in same resource allocation program subset；It is dominated according to highest The resource allocation program subset of grade obtains target resource configuration plan set；It is configured in plan set from the target resource Obtain the target resource configuration strategy for the Component service；And according to the target resource configuration strategy be the component Service distribution resource.
Optionally, each resource allocation program includes multiple resource allocation strategies, wherein each resource distribution Strategy is directed to a kind of user type.
Optionally, the method further includes：There is the resource allocation program subset of highest dominance hierarchy in the basis, obtain After taking target resource configuration plan set, and obtained for described group in described configured during plan is gathered from the target resource Before the target resource configuration strategy of part service, judge whether actual cycle number is equal to preset cycle-index, wherein described The initial value of actual cycle number is 0, and the preset cycle-index is greater than or equal to 1；It is differed in the actual cycle number In the case of the preset cycle-index, the actual cycle number is added 1, and the target resource is configured into plan collection After cooperation is new initial resource configuration plan set, re-execute it is described will be each in the initial resource configuration plan set A resource allocation program is divided according to dominance hierarchy, multiple resource allocation program subsets is formed, until the actual cycle Until number is equal to the preset cycle-index.
Optionally, described that the initial resource is configured into each resource allocation program in plan set according to dominance hierarchy It is divided, forms multiple resource allocation program subsets, including：Determine that the domination between each resource allocation program is closed System, wherein in the dominance relation, the dominance hierarchy of Zhi Peifang is higher than the dominance hierarchy of subject side；According to the domination Resource allocation program with identical dominance hierarchy is divided into a resource allocation program subset, wherein same resource by relationship It is not dominated mutually between resource allocation program in configuration plan subset, also, the resource allocation program with higher dominance hierarchy Any resource allocation program in subset can dominate whole moneys in the resource allocation program subset with relatively low dominance hierarchy Source configuration plan.
Optionally, the dominance relation between the determination each resource allocation program, including：From the initial resource First resource configuration plan and Secondary resource configuration plan are chosen in configuration plan set；Determine the first parameter set and the second parameter Collection, wherein first parameter set includes the resource distribution plan that the Component service includes in the first resource configuration plan QoS parameter under slightly and required total resources, second parameter set include the Component service in the Secondary resource QoS parameter under the resource allocation strategy that configuration plan includes and required total resources；It is true according to first parameter set Determine third parameter set, and the 4th parameter set is determined according to second parameter set, wherein the third parameter set includes described QoS parameter and required total resources of the Component service under first resource configuration plan, the 4th parameter set packet Include QoS parameter and required total resources of the Component service under the Secondary resource configuration plan；Institute is judged respectively It states the first parameter set and whether second parameter set meets preset parameter threshold condition；When first parameter set and described One of second parameter set meets the parameter threshold condition, and when another one is unsatisfactory for the parameter threshold condition, it determines Meet the domination of the resource allocation program corresponding to the parameter set of the parameter threshold condition and is unsatisfactory for the parameter threshold condition Resource allocation program corresponding to parameter set；When first parameter set and second parameter set are satisfied by the parameter threshold When condition, judge that the third parameter set and the 4th parameter set are to meet the first preset condition or the second preset condition, Wherein, first preset condition is not inferior to of the same race in the 4th parameter set for each parameter in the third parameter set Parameter, and at least one of described third parameter set parameter is better than the parameter of the same race in the 4th parameter set；Described Two preset conditions are that each parameter in the 4th parameter set is not inferior to the parameter of the same race in the third parameter set, and institute At least one of the 4th parameter set parameter is stated better than the parameter of the same race in the third parameter set；When the third parameter set and When 4th parameter set meets first preset condition, determine that the first resource configuration plan dominates the Secondary resource Configuration plan；When the third parameter set and the 4th parameter set meet second preset condition, described second is determined Resource allocation program dominates the first resource configuration plan；When the third parameter set and the 4th parameter set are unsatisfactory for institute When stating the first preset condition and being unsatisfactory for second preset condition, or work as first parameter set and second parameter When collection is unsatisfactory for the parameter threshold condition, determine that the first resource configuration plan and the Secondary resource configuration plan it Between do not dominate mutually；Repeat the selection first resource configuration plan and second from the initial resource configuration plan set Resource allocation program, until dominance relation between determining each resource allocation program, wherein choose every time First resource configuration plan and Secondary resource configuration plan are not all the same.
Optionally, each resource allocation program includes multiple resource allocation strategies, wherein each resource distribution Strategy is directed to a kind of user type；And first parameter set includes the Component service in the first resource configuration plan Including each resource allocation strategy under handling capacity, delay time, required cost and required total resources；Second parameter Collection includes handling capacity of the Component service under each resource allocation strategy that the Secondary resource configuration plan includes, postpones Time, required cost and required total resources；The third parameter set includes that the Component service is configured in the first resource Handling capacity, delay time, required cost and required total resources under plan；4th parameter set includes the Component service Handling capacity, delay time, required cost and required total resources under Secondary resource configuration plan.
Optionally, the parameter threshold condition includes：
Wherein, XlIndicate first of resource allocation strategy that resource allocation program includes, first of resource allocation strategy needle To l kind user types；L indicates user type sum；Indicate that the Component service includes in resource allocation program First of resource allocation strategy XlUnder required total resources；Indicate the Component service in resource distribution meter Draw first of resource allocation strategy X for includinglUnder handling capacity；Indicate the Component service in resource allocation program Including first of resource allocation strategy XlUnder delay time；Indicate the Component service in resource allocation program packet First of the resource allocation strategy X includedlUnder required cost；CsourceIndicate the preset required total resources for the system Threshold value；Indicate the preset throughput threshold for being directed to l kind user types；It indicates preset and is directed to l kinds use The delay time threshold value of family type；Indicate the preset required cost threshold value for being directed to l kind user types；CcostIndicate pre- If the required cost threshold value for the system.
Optionally, the basis has the resource allocation program subset of highest dominance hierarchy, obtains target resource configuration meter Set is drawn, including：Chosen from the resource allocation program subset with highest dominance hierarchy multiple resource allocation programs into Row cross and variation generates multiple derivative resource allocation programs；The multiple derivative resource allocation programs generated are formed as derivative Resource allocation program set；By the resource allocation program subset with highest dominance hierarchy and the derivative resource distribution meter It draws union of sets collection and is formed as the target resource configuration plan set.
Optionally, described that the target resource obtained in plan set for the Component service is configured from the target resource Configuration strategy, including：According to the target resource configure plan set in each resource allocation program priority from height to Low sequence obtains the resource allocation program of predetermined quantity as candidate resource configuration plan；Obtain user type information；From every A candidate resource configuration in the works, filters out the resource allocation strategy to match with the user type information；Obtain user couple The preference setting information of resource distribution；From the resource allocation strategy to match with the user type information, filter out The resource allocation strategy to match with the preference setting information, as the target resource configuration strategy.
The present invention also provides a kind of resource allocation device for the software systems based on service, which includes：First Acquisition module, multiple and different resource allocation programs for obtaining the Component service for the system form initial resource and match Set plan set, wherein each resource allocation program includes the resource allocation strategy of the Component service；Subset division module, For dividing each resource allocation program in the initial resource configuration plan set according to dominance hierarchy, formed more A resource allocation program subset, wherein resource allocation program domination having the same in same resource allocation program subset etc. Grade；Second acquisition module, for according to the resource allocation program subset with highest dominance hierarchy, obtaining target resource configuration meter Draw set；Third acquisition module, for configuring the mesh obtained in plan set for the Component service from the target resource Mark resource allocation strategy；And resource distribution module, for being the Component service point according to the target resource configuration strategy With resource.
Optionally, each resource allocation program includes multiple resource allocation strategies, wherein each resource distribution Strategy is directed to a kind of user type.
Optionally, described device further includes：Judgment module, in the basis there is the resource of highest dominance hierarchy to match Plan subset is set, after obtaining target resource configuration plan set, and described from the target resource configures plan set Before obtaining the target resource configuration strategy for the Component service, judge whether actual cycle number is equal to preset cycle Number, wherein the initial value of the actual cycle number is 0, and the preset cycle-index is greater than or equal to 1；Circular flow Module is used in the case where the actual cycle number is not equal to the preset cycle-index, by the actual cycle time Number plus 1, and the target resource is configured into plan and is gathered as after new initial resource configuration plan set, it reruns described Subset division module, until the actual cycle number is equal to the preset cycle-index.
Optionally, the subset division module includes：Dominance relation determination sub-module, for determining that each resource is matched Dominance relation between the plan of setting, wherein in the dominance relation, the dominance hierarchy of Zhi Peifang is higher than the domination of subject side Grade；Subset division submodule, for according to the dominance relation, will be divided with the resource allocation program of identical dominance hierarchy At a resource allocation program subset, wherein it is not dominated mutually between the resource allocation program in same resource allocation program subset, Also, any resource allocation program in the resource allocation program subset with higher dominance hierarchy can be dominated with relatively low branch Whole resource allocation programs in resource allocation program subset with grade.
Optionally, the dominance relation determination sub-module includes：Resource allocation program selection unit is used for from described initial First resource configuration plan and Secondary resource configuration plan are chosen in resource allocation program set；First parameter determination unit is used In determining the first parameter set and the second parameter set, wherein first parameter set includes the Component service in first money QoS parameter under the resource allocation strategy that source configuration plan includes and required total resources, second parameter set include QoS parameter and required money of the Component service under the resource allocation strategy that the Secondary resource configuration plan includes Source total amount；Second parameter determination unit, for determining third parameter set according to first parameter set, and according to described second Parameter set determines the 4th parameter set, wherein the third parameter set includes that the Component service configures meter in the first resource QoS parameter and required total resources, the 4th parameter set under drawing include the Component service in the Secondary resource QoS parameter under configuration plan and required total resources；First judging unit, for judging first parameter respectively Whether collection and second parameter set meet preset parameter threshold condition；First dominance relation determination unit, for when described One of first parameter set and second parameter set meet the parameter threshold condition, and another one is unsatisfactory for the parameter When threshold condition, the determining resource allocation program domination met corresponding to the parameter set of the parameter threshold condition is unsatisfactory for described Resource allocation program corresponding to the parameter set of parameter threshold condition；Second judgment unit, for when first parameter set and When second parameter set is satisfied by the parameter threshold condition, judge that the third parameter set and the 4th parameter set are full The first preset condition of foot or the second preset condition, wherein first preset condition is each in the third parameter set Parameter is not inferior to the parameter of the same race in the 4th parameter set, and at least one of described third parameter set parameter is better than institute State the parameter of the same race in the 4th parameter set；Second preset condition is not inferior to institute for each parameter in the 4th parameter set The parameter of the same race in third parameter set is stated, and at least one of described 4th parameter set parameter is better than the third parameter set In parameter of the same race；Second dominance relation determination unit meets institute for working as the third parameter set and the 4th parameter set When stating the first preset condition, determine that the first resource configuration plan dominates the Secondary resource configuration plan；Third, which dominates, closes It is determination unit, for when the third parameter set and the 4th parameter set meet second preset condition, determining institute It states Secondary resource configuration plan and dominates the first resource configuration plan；4th dominance relation determination unit, for when described the When three parameter sets and the 4th parameter set are unsatisfactory for first preset condition and are unsatisfactory for second preset condition, or Person determines the first resource when first parameter set and second parameter set are unsatisfactory for the parameter threshold condition It is not dominated mutually between configuration plan and the Secondary resource configuration plan；Circular flow unit, for the resource that reruns Configuration plan selection unit, until dominance relation between determining each resource allocation program, wherein choosing every time The first resource configuration plan and Secondary resource configuration plan taken is not all the same.
Optionally, each resource allocation program includes multiple resource allocation strategies, wherein each resource distribution Strategy is directed to a kind of user type；And first parameter set includes the Component service in the first resource configuration plan Including each resource allocation strategy under handling capacity, delay time, required cost and required total resources；Second parameter Collection includes handling capacity of the Component service under each resource allocation strategy that the Secondary resource configuration plan includes, postpones Time, required cost and required total resources；The third parameter set includes that the Component service is configured in the first resource Handling capacity, delay time, required cost and required total resources under plan；4th parameter set includes the Component service Handling capacity, delay time, required cost and required total resources under Secondary resource configuration plan.
Optionally, the parameter threshold condition includes：
Wherein, XlIndicate first of resource allocation strategy that resource allocation program includes, first of resource allocation strategy needle To l kind user types；L indicates user type sum；Indicate that the Component service includes in resource allocation program First of resource allocation strategy XlUnder required total resources；Indicate the Component service in resource distribution meter Draw first of resource allocation strategy X for includinglUnder handling capacity；Indicate the Component service in resource allocation program Including first of resource allocation strategy XlUnder delay time；Indicate the Component service in resource allocation program packet First of the resource allocation strategy X includedlUnder required cost；CsourceIndicate the preset required total resources for the system Threshold value；Indicate the preset throughput threshold for being directed to l kind user types；It indicates preset and is directed to l kinds use The delay time threshold value of family type；Indicate the preset required cost threshold value for being directed to l kind user types；CcostIndicate pre- If the required cost threshold value for the system.
Optionally, second acquisition module includes：Cross and variation submodule, for there is highest dominance hierarchy from described Resource allocation program subset in choose multiple resource allocation programs and carry out cross and variations, generate multiple derivative resource distribution meters It draws；Derivative resource allocation program set forms submodule, for being formed as spreading out by the multiple derivative resource allocation programs generated Raw resource allocation program set；Target resource configuration plan set forms submodule, for that described will have highest dominance hierarchy Resource allocation program subset and the derivative resource allocation program union of sets collection be formed as the target resource configuration plan Set.
Optionally, the third acquisition module includes：Candidate resource configuration plan acquisition submodule, for according to the mesh The sequence of the priority of each resource allocation program in resource allocation program set from high to low is marked, the money of predetermined quantity is obtained Source configuration is intended to be candidate resource configuration plan；User type information acquisition submodule, for obtaining user type information；The One screening submodule, in the works, filtering out the money to match with the user type information from the configuration of each candidate resource Source configuration strategy；Preference setting information acquisition submodule, for obtaining preference setting information of the user to resource distribution；Second sieve Submodule is selected, for from the resource allocation strategy to match with the user type information, filtering out and the preference The resource allocation strategy that setting information matches, as the target resource configuration strategy.
Through the above technical solutions, the group for SBS can quickly be determined according to multiple and different resource allocation programs The target resource configuration strategy of part service, and the resource for making ISP consume is minimum.In addition, the money provided through the invention Source distribution method can consider from a variety of different QoS, seek the multiple resource allocation programs not dominated mutually for meeting SLA, Realize the resource allocation based on multiple target.For example, some plan has medium throughput, the medium response time, is assigned with Medium stock number, and another plan has high-throughput, the faster response time, is assigned with more stock number.In this way, The target of the Component service for SBS can be determined in the resource allocation program not dominated mutually from these according to actual demand Resource allocation strategy, and be that Component service distributes resource according to the target resource configuration strategy.Therefore, it is possible to improve Component service Resource allocation flexibility.
Other features and advantages of the present invention will be described in detail in subsequent specific embodiment part.
Description of the drawings
Attached drawing is to be used to provide further understanding of the present invention, an and part for constitution instruction, with following tool Body embodiment is used to explain the present invention together, but is not construed as limiting the invention.In the accompanying drawings：
Fig. 1 shows the configuration diagram of an example SBS；
Fig. 2 is the flow chart of the resource allocation methods for SBS according to the embodiment of the present invention；
Fig. 3 shows the configuration diagram of the Component service for the SBS that an exemplary embodiment provides；
Fig. 4 A and Fig. 4 B show the coded sequence schematic diagram of resource allocation program according to the embodiment of the present invention；
Fig. 5 shows two moneys how determined in initial resource configuration plan set according to the embodiment of the present invention The flow chart of the process of dominance relation between the configuration plan of source；
Fig. 6 shows the stream of the process for how obtaining target resource configuration plan set according to the embodiment of the present invention Cheng Tu；
Fig. 7 shows the target resource configuration strategy how obtained for Component service according to the embodiment of the present invention Process flow chart；
Fig. 8 shows the flow chart of the resource allocation methods for SBS of another embodiment according to the present invention；
Fig. 9 is the block diagram of the resource allocation device for SBS according to the embodiment of the present invention；
Figure 10 is the block diagram of the resource allocation device for SBS of another embodiment according to the present invention；
Figure 11 is the block diagram of the resource allocation device for SBS of another embodiment according to the present invention；
Figure 12 is the block diagram of the resource allocation device for SBS of another embodiment according to the present invention；
Figure 13 is the block diagram of the resource allocation device for SBS of another embodiment according to the present invention.
Specific implementation mode
The specific implementation mode of the present invention is described in detail below in conjunction with attached drawing.It should be understood that this place is retouched The specific implementation mode stated is merely to illustrate and explain the present invention, and is not intended to restrict the invention.
Fig. 2 is the flow chart of the resource allocation methods for SBS according to the embodiment of the present invention.As shown in Fig. 2, should Method may include：
In step s 201, the multiple and different resource allocation programs for obtaining the Component service for SBS form initial money Source configuration plan set, wherein each resource allocation program may include the resource allocation strategy of Component service.
Fig. 3 shows the configuration diagram of the Component service for the SBS that an exemplary embodiment provides.As shown in Figure 3, it is assumed that SBS systems may include five Component services, be Component service S1, Component service S2, Component service S3, Component service S4 respectively With Component service S5.Each Component service has corresponding several computing resources, wherein computing resource can for example include but It is not limited to：CPU, memory etc..By distributing different resources to Component service, each Component service obtains corresponding Service Quality Amount, for example, handling capacity, required cost, delay time etc..In the present invention, will claim corresponding to the resource distribution of service quality For resource configuration unit.
Each Component service can have several resource configuration units, wherein each resource configuration unit is in component In the limitation for servicing the assignable max calculation resource of corresponding physical machine itself.Also, each resource configuration unit has phase Corresponding service quality.For example, distributing the memory of 1 CPU and 1GB for Component service S1, then the Component service can reach The handling capacity of 1000req/s, and required cost is 10.For each Component service, each resource configuration unit and service Correspondence between quality can rule of thumb data be preset.
In the present invention, resource allocation strategy can be used to indicate that the resource distribution for each of SBS part services The selection result of unit.For example, it is assumed that a Component service corresponds to several resource configuration units, selected if it is the Component service It selects one of resource configuration unit to be configured, then the resource configuration unit can be encoded as 1, and be directed to the Component service Other resource configuration units can be encoded as 0.Resource can be carried out to each Component service all in accordance with such mode to match Selection and the coding for setting unit, so as to form a resource allocation strategy for being directed to all components service, and the resource Configuration strategy can show as a coded sequence (for example, binary code sequence).It should be noted that being selected to Component service Resource configuration unit is selected come when formulating resource allocation strategy, each Component service has and only there are one corresponding resource configuration units It is selected.
In an embodiment of the invention, a resource allocation program may include a resource allocation strategy. In this case, the coded sequence for being formed by resource allocation strategy is the coded sequence of resource allocation program.For example, Fig. 4 A Show the coded sequence schematic diagram of the example resources configuration plan according to this embodiment.
Alternatively, in actual use, user is often divided into different user class to SBS serve it Type, for example, the user of gold card grade and the user of silver card grade.These different types of users have the service quality of Component service Different demands.Therefore, in yet another embodiment of the present invention, when formulating resource allocation program, can first be directed to Each user type services various components selection and the coding for carrying out resource configuration unit, is formed and be directed to each user type Component service resource allocation strategy coded sequence.Later, will match for the resource of the Component service of each user type The coded sequence for setting strategy is spliced into a coded sequence, to form the coded sequence of resource allocation program.That is, one Resource allocation program may include multiple resource allocation strategies, wherein each resource allocation strategy is directed to a kind of user type.Example Such as, Fig. 4 B show that one includes the resource for the resource allocation strategy of the Component service of two kinds of user types (gold card, silver card) Configure the coded sequence schematic diagram of plan.As shown in Figure 4 B, it in the resource allocation program may include two resource distribution plans Slightly, wherein first resource allocation strategy is the user for gold card grade, and second resource allocation strategy is for silver card grade User.
By this embodiment, the computing resource for meeting the respective demand of all types of users can be distributed simultaneously, is made The system that must be disposed can be used in conjunction with by catergories of user.
It can be with by the above-mentioned means, set out multiple (for example, 20) different resource allocation program in advance (can be set by calculating It is standby to be randomly provided, or by being manually arranged), these resource allocation programs can form initial resource configuration plan set.
Referring back to Fig. 2, next, in step S202, initial resource is configured into each resource in plan set Configuration plan is divided according to dominance hierarchy, forms multiple resource allocation program subsets, wherein same resource allocation program The resource allocation program dominance hierarchy having the same of concentration.
Specifically, it when dividing resource allocation program subset, can determine first in initial resource configuration plan set Dominance relation between each resource allocation program.Later, according to dominance relation, by the resource distribution with identical dominance hierarchy It is divided into a resource allocation program subset, wherein between the resource allocation program in same resource allocation program subset It does not dominate mutually, also, any resource allocation program in the resource allocation program subset with higher dominance hierarchy can dominate Whole resource allocation programs in resource allocation program subset with relatively low dominance hierarchy.
In the present invention, the dominance relation between resource allocation program may include that a resource allocation program domination is another A resource allocation program.This dominance relation may indicate that the resource allocation program in Zhi Peifang is better than in terms of service quality Resource allocation program in subject side, thus, in Zhi Peifang resource allocation program dominance hierarchy be higher than in by The dominance hierarchy of the resource allocation program of Zhi Peifang.In addition, the dominance relation between resource allocation program can also include two It is not dominated mutually between resource allocation program.This dominance relation may indicate that two resource allocation programs do not have in terms of service quality It is divided into quality, thus, the two resource allocation programs dominance hierarchy having the same.
Assuming that initial resource configuration plan set expression is P, then can be according to the domination etc. of each resource allocation program Grade, divides it, forms multiple resource allocation program subset P1、P2、…、Pa, wherein a indicates the resource distribution meter to be formed The sum of small rowboat collection.Wherein, the resource allocation program in same resource allocation program subset can with dominance hierarchy having the same, That is, not dominated mutually between resource allocation program in same resource allocation program subset.In addition, with higher domination etc. Each resource allocation program in the resource allocation program subset of grade, can dominate the resource with relatively low dominance hierarchy and match Set all resource allocation programs in plan subset.
Fig. 5 shows two moneys how determined in initial resource configuration plan set according to the embodiment of the present invention The flow chart of the process of dominance relation between the configuration plan of source.As shown in figure 5, determine dominance relation process may include with Lower step：
In step S501, chooses first resource configuration plan from initial resource configuration plan set and Secondary resource is matched The plan of setting.Wherein, first resource configuration plan and Secondary resource configuration plan are arbitrary in initial resource configuration plan set Different two resource allocation programs.
In step S502, the first parameter set and the second parameter set are determined, wherein first parameter set includes Component service QoS parameter under the resource allocation strategy that first resource configuration plan includes and required total resources, second parameter Collection includes QoS parameter and required resource of the Component service under the resource allocation strategy that Secondary resource configuration plan includes Total amount.
In step S503, third parameter set is determined according to the first parameter set, and the 4th is determined according to the second parameter set Parameter set, wherein the third parameter set includes QoS parameter of the Component service under first resource configuration plan and required Total resources, the 4th parameter set include QoS parameter and required resource of the Component service under Secondary resource configuration plan Total amount.
As previously mentioned, each resource allocation program in initial resource configuration plan set may include multiple resource distributions Strategy, wherein each resource allocation strategy is directed to a kind of user type.In this case, the first parameter set may include group Handling capacity of the part service under each resource allocation strategy that first resource configuration plan includes, delay time, required cost and Required total resources；Second parameter set may include each resource distribution plan that Component service includes in Secondary resource configuration plan Handling capacity, delay time, required cost under slightly and required total resources；Third parameter set may include Component service first Handling capacity, delay time, required cost under resource allocation program and required total resources；And the 4th parameter set may include Handling capacity, delay time, required cost and required total resources of the Component service under Secondary resource configuration plan.
In step S504, judge whether the first parameter set and the second parameter set meet preset parameter threshold item respectively Part.
In an example embodiment, shown parameter threshold condition may include：
Wherein, XlFor a vector, first of resource allocation strategy that resource allocation program includes, first of resource are indicated Configuration strategy is directed to l kind user types；L indicates user type sum；Indicate Component service in resource distribution meter Draw first of resource allocation strategy X for includinglUnder required total resources；Indicate Component service in resource distribution First of resource allocation strategy X that plan includeslUnder handling capacity；Indicate Component service in resource allocation program packet First of the resource allocation strategy X includedlUnder delay time；Indicate the l that Component service includes in resource allocation program A resource allocation strategy XlUnder required cost；CsourceIndicate the preset required total resources threshold value for the system；Indicate the preset throughput threshold for being directed to l kind user types；It indicates preset and is directed to l kind user class The delay time threshold value of type；Indicate the preset required cost threshold value for being directed to l kind user types；CcostIndicate preset For the required cost threshold value of the system.
Wherein it is possible to be determined by following equation (2)~(5)With
Wherein, csjk、costjk、latjkAnd thrjkK-th of resource configuration unit of j-th of Component service is indicated respectively Total resources, required cost, delay time and handling capacity, this four parameters can be taken by inquiry for j-th component The preset resource configuration unit of business and the correspondence of service quality obtain.N and M indicates Component service in SBS respectively The quantity of the resource configuration unit of quantity and Component service；N ' indicates the critical path (delay time longest path) of SBS Including Component service quantity；Vectorial XlIn element representation beWherein, j indicates j-th of Component service in SBS, k tables Show corresponding k-th of the resource configuration unit of j-th of Component service, and wherein, if having selected k-th of resource distribution list Member, thenValue be 1, be otherwise 0.
Parameter threshold term restriction, which is can be seen that, by equation (1) is directed to each user type, the service of Component service Quality (that is, handling capacity, delay time, required cost) will meet the corresponding quality of service requirement of the user type, for example, handling up AmountIt cannot be below the preset throughput threshold for user type lIt should be understood that needle To user type not of the same race, WithIt can be different.
In addition, the parameter threshold condition also defines in a resource allocation program, the required total resources of system (that is, required resource of the Component service under the resource allocation strategy for various user types that the resource allocation program includes The sum of total amount) no more than the preset required total resources threshold value C for the systemsourceAnd the required cost of system (that is, required cost of the Component service under the resource allocation strategy for various user types that the resource allocation program includes The sum of) no more than the preset required cost threshold value C for the systemcost, to make the resource allocation program disclosure satisfy that SLA requirements.
When a certain parameter set meets parameter threshold condition, the resource allocation program corresponding to the parameter set can be referred to as " feasible solution ", and when a certain parameter set is unsatisfactory for parameter threshold condition, resource allocation program corresponding to the parameter set can be with It is referred to as " infeasible solution ".
In step S505, when one of the first parameter set and the second parameter set meet parameter threshold condition, and it is another When person is unsatisfactory for parameter threshold condition, determine that the resource allocation program met corresponding to the parameter set of parameter threshold condition dominates not Meet the resource allocation program corresponding to the parameter set of parameter threshold condition.That is, in the present invention, feasible solution dominates not Feasible solution.
For example, it is assumed that the first parameter set meets parameter threshold condition, and the second parameter set is unsatisfactory for parameter threshold condition, then Determine that first resource configuration plan dominates Secondary resource configuration plan.
In step S506, when the first parameter set and the second parameter set are unsatisfactory for parameter threshold condition, first is determined It is not dominated mutually between resource allocation program and Secondary resource configuration plan.
In step s 507, when the first parameter set and the second parameter set are satisfied by parameter threshold condition, judge that third is joined Manifold and the 4th parameter set are to meet the first preset condition or the second preset condition, wherein first preset condition is third Each parameter in parameter set is not inferior to the parameter of the same race in the 4th parameter set, and at least one of third parameter set parameter Better than the parameter of the same race in the 4th parameter set；Second preset condition is that each parameter in the 4th parameter set is not inferior to third ginseng Parameter of the same race in manifold, and at least one of the 4th parameter set parameter is better than the parameter of the same race in third parameter set.
For example, the first preset condition can be expressed as to formula (6), and the second preset condition is expressed as following Equation (7)：
Wherein, piIndicate first resource configuration plan；pjIndicate Secondary resource configuration plan；Thr(pi) indicate Component service Handling capacity under first resource configuration plan, the handling capacity can be that Component service configures plan p in first resourceiIncluding For the sum of the handling capacity under the resource allocation strategy of various user types；Thr(pj) indicate that Component service is matched in Secondary resource Handling capacity under the plan of setting, the handling capacity can be that Component service configures plan p in Secondary resourcejIncluding be directed to various users The sum of handling capacity under the resource allocation strategy of type；Cost(pi) indicate institute of the Component service under first resource configuration plan Cost is needed, which can be that Component service configures plan p in first resourceiIncluding the money for various user types The sum of required cost under the configuration strategy of source；Cost(pj) indicate Component service Secondary resource configuration plan under needed at This, which can be that Component service configures plan p in Secondary resourcejIncluding the resource for various user types match Set the sum of the required cost under strategy；Lat(pi) indicating delay time of the Component service under first resource configuration plan, this prolongs The slow time can be that Component service configures plan p in first resourceiIncluding the resource allocation strategy for various user types Under the sum of delay time；Lat(pj) indicate delay time of the Component service under Secondary resource configuration plan, the delay time Can be that Component service configures plan p in Secondary resourcejIncluding the resource allocation strategy for various user types under prolong The sum of slow time；Source(pi) indicate required total resources of the Component service under first resource configuration plan, the required money Source total amount can be that Component service configures plan p in first resourceiIncluding the resource allocation strategy for various user types Under the sum of required total resources；Source(pj) indicate that required resource of the Component service under Secondary resource configuration plan is total Amount, the required total resources can be that Component service configures plan p in Secondary resourcejIncluding the money for various user types The sum of required total resources under the configuration strategy of source.
In step S508, when third parameter set and the 4th parameter set the first preset condition of satisfaction, first resource is determined Configuration plan dominates Secondary resource configuration plan.
In step S509, when third parameter set and the 4th parameter set the second preset condition of satisfaction, Secondary resource is determined Configuration plan dominates first resource configuration plan.
That is, in the present invention, in the case where two resource allocation programs are feasible solution, in which parameter set Including parameter it is more excellent, corresponding resource allocation program dominates another resource allocation program.
In step S510, when third parameter set and the 4th parameter set are unsatisfactory for the first preset condition and are unsatisfactory for second When preset condition, determines and do not dominated mutually between first resource configuration plan and Secondary resource configuration plan.
The above process (that is, step S501 to step S510) can repeat, until determining each resource distribution meter Until dominance relation between drawing, wherein the first resource configuration plan chosen every time and the infull phase of Secondary resource configuration plan Together.
Referring back to Fig. 2, after marking off resource allocation program subset, in step S203, according to highest The resource allocation program subset of dominance hierarchy obtains target resource configuration plan set.
Fig. 6 shows the mistake for how obtaining target resource configuration plan set of an example embodiment according to the present invention The flow chart of journey.As shown in fig. 6, step S203 may include：
In step s 601, multiple resource distribution meters are chosen from the resource allocation program subset with highest dominance hierarchy It draws and carries out cross and variation, generate multiple derivative resource allocation programs.
For example, it is assumed that the resource allocation program subset with highest dominance hierarchy is expressed as P1, then can be first from P1Middle choosing It takes (for example, randomly selecting) two resource allocation programs to carry out cross and variation, generates two derivative resource allocation programs.Later, It may determine that whether the quantity of generated derivative resource allocation program reaches the quantity of preset derivative resource allocation program.Such as Fruit does not reach, then repeats from P1The middle operation chosen two resource allocation programs and carry out cross and variation, until spreading out for generation Until the quantity of raw resource allocation program reaches the quantity of preset derivative resource allocation program, wherein every time from P1Middle selection Two resource allocation programs it is not all the same.
In step S602, the multiple derivative resource allocation programs generated are added to derivative resource allocation program set In, wherein the derivative resource allocation program set is initially empty set.
In step S603, by resource allocation program subset and derivative resource allocation program collection with highest dominance hierarchy The union of conjunction is formed as target resource configuration plan set.
The process is based on the selection in genetic algorithm, intersection, mutation operation, from the resource distribution with highest dominance hierarchy New individual is generated in several resource allocation programs (can be described as " individual ") that plan subset includes (that is, derivative resource distribution meter Draw), so as to expand target resource configuration plan set, the chance to more excellent solution is sought in promotion.
Referring back to Fig. 2, after getting target resource configuration plan set, in step S204, provided from target The target resource configuration strategy for Component service is obtained in source configuration plan set；And in step S205, according to target Resource allocation strategy is that Component service distributes resource.
As previously mentioned, each resource allocation program in initial resource configuration plan set may include multiple resource distributions Strategy, wherein each resource allocation strategy is directed to a kind of user type.In such a case, it is possible to for example based on user type, And the preference of resource distribution is arranged in user, to obtain the target resource configuration strategy for the Component service of the user.
Specifically, Fig. 7 shows the target resource configuration how obtained for Component service according to this embodiment The flow chart of the process of strategy.As shown in fig. 7, step S204 may include：
In step s 701, the priority of each resource allocation program in plan set is configured according to target resource from height To low sequence, the resource allocation program of predetermined quantity is obtained as candidate resource configuration plan.
In the present invention, the priority of resource allocation program can determine in the following manner：
Assuming that two resource allocation programs are respectively expressed as pxAnd py, then：
(1) if pxFor feasible solution, pyFor infeasible solution, then px＞rpy；
(2) if pxAnd pyIt is feasible solution, and pxDominance hierarchy be higher than pyDominance hierarchy, then px＞rpy；
(3) if pxAnd pyIt is feasible solution, pxAnd pyDominance hierarchy it is identical, and crowding_distance [px]> crowding_distance[py], then px＞rpy。
Wherein, px＞rpyIndicate pxPriority be more than pyPriority；crowding_distance[px] indicate px's Crowding distance；crowding_distance[py] indicate pyCrowding distance.
In step S702, user type information is obtained.For example, user is when logging on to system, system can according to The User ID at family inquires the corresponding user class of the User ID from the mapping relations of preset User ID and user type Type information.
In step S703, in the works from the configuration of each candidate resource, the money to match with user type information is filtered out Source configuration strategy.Assuming that candidate resource configuration plan includes p1、p2、p3、p4, wherein every candidate resource configuration plan can wrap Include the resource allocation strategy for the resource allocation strategy of gold card user and for silver card user.Assuming that getting user class Type information is gold card user, then can traverse candidate resource configuration plan p1、p2、p3、p4, filter out each candidate resource configuration The resource allocation strategy for gold card user for including in the works.
In step S704, preference setting information of the user to resource distribution is obtained.
In one embodiment, the preference setting interface of resource distribution can be provided a user, user passes through on the boundary The operations such as the enterprising edlin in face, click can input its preference setting information to resource distribution.Later, system can obtain To preference setting information input by user.
It, can be by the resource distribution that will directly match with user type information alternatively, in another embodiment Strategy is shown to user, in a manner of selected by user, to obtain preference setting information of the user to resource distribution.
In step S705, from the resource allocation strategy to match with user type information, filters out and be arranged with preference The resource allocation strategy of information match is as target resource configuration strategy.
For example, it is assumed that the preference setting information instruction of user requires handling capacity maximum, then, in step S705, so that it may To choose the maximum resource allocation strategy of handling capacity from the resource allocation strategy to match with user type information as target Resource allocation strategy.For another example assume that the preference setting information instruction of user requires required cost minimum, then, in step In S705, so that it may to choose the minimum resource distribution of required cost from the resource allocation strategy to match with user type information Strategy is as target resource configuration strategy, etc..
By this embodiment, can be to different user types, user with different hobbies provides and customizes clothes Business, so as to meet the use demand of different user, greatly improves user experience.
The resource allocation methods provided through the invention as a result, can be according to multiple and different resource allocation programs, quickly Determine the target resource configuration strategy of the Component service for SBS, and the resource for making ISP consume is minimum.In addition, The resource allocation methods provided through the invention can consider from a variety of different QoS, seek meet SLA it is multiple mutually not The resource allocation program of domination realizes the resource allocation based on multiple target.For example, some plan with medium throughput, in Deng response time, be assigned with medium stock number, and another plan is with high-throughput, faster response time, distribution More stock number.In this way, can determine to be directed in the resource allocation program not dominated mutually from these according to actual demand The target resource configuration strategy of the Component service of SBS, and be that Component service distributes resource according to the target resource configuration strategy.Cause And the flexibility of the resource allocation of Component service can be improved.
Fig. 8 shows the flow chart of the resource allocation methods for SBS of another embodiment according to the present invention.Such as figure Shown in 8, this method can also include：
After step S203 and in the step S206 before step S204, judge whether actual cycle number is equal to Preset cycle-index, wherein the initial value of actual cycle number is 0, and preset cycle-index is greater than or equal to 1.
In step S207, in the case where actual cycle number is not equal to preset cycle-index, by actual cycle time Number plus 1, and using the target resource configuration plan set got after step S203 as new initial resource configuration plan collection After conjunction, return to step S202, with re-execute each resource allocation program in the configuration plan set by initial resource by It is divided according to dominance hierarchy, multiple resource allocation program subsets and follow-up step is formed, until actual cycle number etc. Until preset cycle-index.
The loop iteration process of the embodiment is based on genetic algorithm, wherein each resource allocation program can be considered as An individual.In the genetic algorithm, the population (namely target resource configuration plan set) that can obtain every generation makees For follow-on initial population (namely initial resource configuration plan set), and dominance hierarchy division is carried out again, with from new Initial population in determine the resource allocation program with highest dominance hierarchy, and generated in terms of new target resource configuration by this Draw set.
In addition, being saved in last iterative process with most in the resource allocation program subset with highest dominance hierarchy The solution of high dominance hierarchy compared with the solution (that is, derivative resource allocation program set) that next iteration generates there is highest to dominate The resource allocation program subset of grade includes the priority of solution might not be lower than the solution of derivative resource allocation program set.Cause This in current iteration, remains the resource allocation program with highest dominance hierarchy in a preferred embodiment of the invention The solution of subset, and the resource allocation program subset with highest dominance hierarchy is made with derivative resource allocation program union of sets collection For the input of next iteration.
Thus, it is possible to according to the principle of the survival of the fittest and the survival of the fittest, develops by generation and produces the approximate solution become better and better, So as to realize the optimal resource allocation of Component service.In addition, it is O (NP × η) to search the optimal solution required time, Middle NP is the number (that is, preset cycle-index) of algorithm totality iteration, and η is the number of algorithm performs cross and variation.Therefore, phase For existing method, the resource searching time can be greatly lowered, promote the efficiency of resource searching.
Fig. 9 is the frame of the resource allocation device for the software systems based on service according to the embodiment of the present invention Figure.As shown in figure 9, the device may include：First acquisition module 901 is used to obtain the more of the Component service for being directed to the system A different resource allocation program forms initial resource configuration plan set, wherein each resource allocation program includes component clothes The resource allocation strategy of business；Subset division module 902, by configuring initial resource based on each resource distribution in plan set It draws and is divided according to dominance hierarchy, form multiple resource allocation program subsets, wherein in same resource allocation program subset Resource allocation program dominance hierarchy having the same；Second acquisition module 903, for according to the resource with highest dominance hierarchy Configuration plan subset obtains target resource configuration plan set；Third acquisition module 904 is planned for being configured from target resource The target resource configuration strategy for Component service is obtained in set；And resource distribution module 905, for according to target resource Configuration strategy is that Component service distributes resource.
Optionally, each resource allocation program includes multiple resource allocation strategies, wherein each resource allocation strategy is directed to A kind of user type.
Figure 10 is the resource allocation device for the software systems based on service of another embodiment according to the present invention Block diagram.As shown in Figure 10, which can also include：Judgment module 1001, for there is highest dominance hierarchy in basis Resource allocation program subset after obtaining target resource configuration plan set, and is obtained from target resource configuration plan set Before taking the target resource configuration strategy for Component service, judge whether actual cycle number is equal to preset cycle-index, Wherein, the initial value of the actual cycle number is 0, which is greater than or equal to 1；Circular flow module 1002, For in the case where actual cycle number is not equal to the preset cycle-index, actual cycle number to be added 1, and by target After resource allocation program set is as new initial resource configuration plan set, subset division module 902, Zhi Daoshi is reruned Until border cycle-index is equal to preset cycle-index.
Figure 11 is the resource allocation device for the software systems based on service of another embodiment according to the present invention Block diagram.As shown in figure 11, subset division module 902 may include：Dominance relation determination sub-module 1101, described in determining Dominance relation between each resource allocation program, wherein in dominance relation, the dominance hierarchy of Zhi Peifang is higher than subject side Dominance hierarchy；Subset division submodule 1102, for according to the dominance relation, the resource with identical dominance hierarchy to be matched The plan of setting is divided into a resource allocation program subset, wherein resource allocation program in same resource allocation program subset it Between do not dominate mutually, also, with higher dominance hierarchy resource allocation program subset in any resource allocation program can prop up With whole resource allocation programs in the resource allocation program subset with relatively low dominance hierarchy.
Optionally, which may include：
Resource allocation program selection unit chooses first resource configuration plan for being configured from initial resource in plan set It configures and plans with Secondary resource；
First parameter determination unit, for determining the first parameter set and the second parameter set, wherein first parameter set includes QoS parameter and required total resources of the Component service under the resource allocation strategy that first resource configuration plan includes, should Second parameter set include QoS parameter of the Component service under the resource allocation strategy that Secondary resource configuration plan includes and Required total resources；
Second parameter determination unit, for determining third parameter set according to the first parameter set, and according to the second parameter set Determine the 4th parameter set, wherein the third parameter set includes service quality ginseng of the Component service under first resource configuration plan Number and required total resources, the 4th parameter set include QoS parameter of the Component service under Secondary resource configuration plan and Required total resources；
First judging unit, for judging whether the first parameter set and the second parameter set meet preset parameter threshold respectively Condition；
First dominance relation determination unit, for meeting parameter threshold when one of the first parameter set and the second parameter set Condition, and when another one is unsatisfactory for parameter threshold condition, determine that the resource met corresponding to the parameter set of parameter threshold condition is matched The plan of setting dominates the resource allocation program being unsatisfactory for corresponding to the parameter set of parameter threshold condition；
Second judgment unit, for when the first parameter set and the second parameter set are satisfied by parameter threshold condition, judging Three parameter sets and the 4th parameter set are to meet the first preset condition or the second preset condition, wherein first preset condition is Each parameter in third parameter set is not inferior to the parameter of the same race in the 4th parameter set, and at least one of third parameter set Parameter is better than the parameter of the same race in the 4th parameter set；Second preset condition is that each parameter in the 4th parameter set is not inferior to the Parameter of the same race in three parameter sets, and at least one of the 4th parameter set parameter is better than the same seed ginseng in third parameter set Number；
Second dominance relation determination unit is used for when third parameter set and the 4th parameter set the first preset condition of satisfaction, Determine that first resource configuration plan dominates Secondary resource configuration plan；
Third dominance relation determination unit is used for when third parameter set and the 4th parameter set the second preset condition of satisfaction, Determine that Secondary resource configuration plan dominates first resource configuration plan；
4th dominance relation determination unit, for work as third parameter set and the 4th parameter set be unsatisfactory for the first preset condition, And when being unsatisfactory for the second preset condition, or when the first parameter set and the second parameter set are unsatisfactory for parameter threshold condition, really Determine not dominate mutually between first resource configuration plan and Secondary resource configuration plan；
Circular flow unit, for the resource allocation program selection unit that reruns, until determining each resource Until dominance relation between configuration plan, wherein the first resource configuration plan and Secondary resource configuration plan chosen every time It is not all the same.
Figure 12 is the resource allocation device for the software systems based on service of another embodiment according to the present invention Block diagram.As shown in figure 12, the second acquisition module 903 may include：Cross and variation submodule 1201, for from highest branch Multiple resource allocation programs are chosen in resource allocation program subset with grade and carry out cross and variation, are generated multiple derivative resources and are matched The plan of setting；Derivative resource allocation program adds submodule 1202, for adding the multiple derivative resource allocation programs generated Into derivative resource allocation program set, wherein the derivative resource allocation program set is initially empty set；Target resource configuration meter It draws set and forms submodule 1203, by that will have the resource allocation program subset of highest dominance hierarchy and derive based on resource distribution It draws union of sets collection and is formed as target resource configuration plan set.
Figure 13 is the resource allocation device for the software systems based on service of another embodiment according to the present invention Block diagram.As shown in figure 13, third acquisition module 904 may include：Candidate resource configuration plan acquisition submodule 1301, is used for The sequence of the priority of each resource allocation program in plan set from high to low is configured according to target resource, obtains predetermined number The resource allocation program of amount is configured as candidate resource to be planned；User type information acquisition submodule 1302, for obtaining user Type information；First screening submodule 1303, in the works, being filtered out and user type information from the configuration of each candidate resource The resource allocation strategy to match；Preference setting information acquisition submodule 1304, for obtaining preference of the user to resource distribution Setting information；Second screening submodule 1305, for from the resource allocation strategy to match with user type information, filtering out The resource allocation strategy to match with preference setting information, as target resource configuration strategy.
About the device in the various embodiments described above, wherein modules execute the concrete mode of operation in the related party It is described in detail in the embodiment of method, explanation will be not set forth in detail herein.
The preferred embodiment of the present invention is described in detail above in association with attached drawing, still, the present invention is not limited to above-mentioned realities The detail in mode is applied, within the scope of the technical concept of the present invention, a variety of letters can be carried out to technical scheme of the present invention Monotropic type, these simple variants all belong to the scope of protection of the present invention.
In addition, although the operation of the method for the present invention is described with particular order in the accompanying drawings, this do not require that or Hint must execute these operations according to the particular order, or have to carry out shown in whole operation could realize it is desired As a result.Additionally or alternatively, it is convenient to omit multiple steps are merged into a step and executed by certain steps, and/or by one Step is decomposed into execution of multiple steps.
One skilled in the art will appreciate that embodiments of the present invention can be implemented as a kind of system, device, equipment, method Or computer program product.Therefore, embodiments of the present invention can be with specific implementation is as follows, i.e.,：It is complete hardware, complete The form that full software (including firmware, resident software, microcode etc.) or hardware and software combine.
It is further to note that specific technical features described in the above specific embodiments, in not lance In the case of shield, it can be combined by any suitable means.In order to avoid unnecessary repetition, the present invention to it is various can The combination of energy no longer separately illustrates.
In addition, various embodiments of the present invention can be combined randomly, as long as it is without prejudice to originally The thought of invention, it should also be regarded as the disclosure of the present invention.

Claims (18)

1. a kind of resource allocation methods for the software systems based on service, which is characterized in that this method includes：
Multiple and different resource allocation programs of the Component service for the system are obtained, initial resource configuration plan collection is formed It closes, wherein each resource allocation program includes the resource allocation strategy of the Component service；
Each resource allocation program in the initial resource configuration plan set is divided according to dominance hierarchy, is formed more A resource allocation program subset, wherein resource allocation program domination having the same in same resource allocation program subset etc. Grade；
According to the resource allocation program subset with highest dominance hierarchy, target resource configuration plan set is obtained, wherein dominate Higher grade, and corresponding resource allocation program has more excellent service quality；
The target resource configuration strategy obtained in plan set for the Component service is configured from the target resource；And
It is that the Component service distributes resource according to the target resource configuration strategy.
2. according to the method described in claim 1, it is characterized in that, each resource allocation program includes multiple resources Configuration strategy, wherein each resource allocation strategy is directed to a kind of user type.
3. according to the method described in claim 1, it is characterized in that, the method further includes：
There is the resource allocation program subset of highest dominance hierarchy in the basis, obtain target resource configuration plan set Afterwards, and in the target resource configuration strategy for configuring and obtaining during plan is gathered for the Component service from the target resource Before, judging whether actual cycle number is equal to preset cycle-index, wherein the initial value of the actual cycle number is 0, The preset cycle-index is greater than or equal to 1；
In the case where the actual cycle number is not equal to the preset cycle-index, the actual cycle number is added 1, And after the target resource configured plan set as new initial resource configuration plan set, re-execute it is described will described in Each resource allocation program in initial resource configuration plan set is divided according to dominance hierarchy, forms multiple resource distributions Plan subset, until the actual cycle number is equal to the preset cycle-index.
4. method according to any one of claim 1-3, which is characterized in that described configure the initial resource is planned Each resource allocation program in set is divided according to dominance hierarchy, forms multiple resource allocation program subsets, including：
Determine the dominance relation between each resource allocation program, wherein in the dominance relation, the domination of Zhi Peifang Grade is higher than the dominance hierarchy of subject side；
According to the dominance relation, the resource allocation program with identical dominance hierarchy is divided into resource allocation program Collection, wherein do not dominated mutually between the resource allocation program in same resource allocation program subset, also, with higher domination etc. Any resource allocation program in the resource allocation program subset of grade can dominate the resource distribution meter with relatively low dominance hierarchy Whole resource allocation programs that small rowboat is concentrated.
5. according to the method described in claim 4, it is characterized in that, branch between the determination each resource allocation program With relationship, including：
It is configured from the initial resource and chooses first resource configuration plan and Secondary resource configuration plan in plan set；
Determine the first parameter set and the second parameter set, wherein first parameter set includes the Component service described first QoS parameter under the resource allocation strategy that resource allocation program includes and required total resources, the second parameter set packet Include QoS parameter of the Component service under the resource allocation strategy that the Secondary resource configuration plan includes and required Total resources；
Third parameter set is determined according to first parameter set, and the 4th parameter set is determined according to second parameter set, In, the third parameter set includes QoS parameter of the Component service under the first resource configuration plan and required Total resources, the 4th parameter set include QoS parameter of the Component service under the Secondary resource configuration plan With required total resources；
Judge whether first parameter set and second parameter set meet preset parameter threshold condition respectively；
When one of first parameter set and second parameter set meet the parameter threshold condition, and another one is discontented When the foot parameter threshold condition, determine that the resource allocation program met corresponding to the parameter set of the parameter threshold condition dominates It is unsatisfactory for the resource allocation program corresponding to the parameter set of the parameter threshold condition；
When first parameter set and second parameter set are satisfied by the parameter threshold condition, the third parameter is judged Collection and the 4th parameter set are to meet the first preset condition or the second preset condition, wherein first preset condition is Each parameter in the third parameter set is not inferior to the parameter of the same race in the 4th parameter set, and the third parameter set At least one of parameter better than the parameter of the same race in the 4th parameter set；Second preset condition is the 4th parameter Each parameter concentrated is not inferior to the parameter of the same race in the third parameter set, and at least one of described 4th parameter set Parameter is better than the parameter of the same race in the third parameter set；
When the third parameter set and the 4th parameter set meet first preset condition, determine that the first resource is matched The plan of setting dominates the Secondary resource configuration plan；
When the third parameter set and the 4th parameter set meet second preset condition, determine that the Secondary resource is matched The plan of setting dominates the first resource configuration plan；
When the third parameter set and the 4th parameter set are unsatisfactory for first preset condition and are unsatisfactory for described second in advance If when condition, or when first parameter set and second parameter set are unsatisfactory for the parameter threshold condition, determine It is not dominated mutually between the first resource configuration plan and the Secondary resource configuration plan；
Selection first resource configuration plan and the Secondary resource from the initial resource configuration plan set is repeated to match The plan of setting, until dominance relation between determining each resource allocation program, wherein the first money chosen every time Source configuration plan and Secondary resource configuration plan are not all the same.
6. according to the method described in claim 5, it is characterized in that, each resource allocation program includes multiple resources Configuration strategy, wherein each resource allocation strategy is directed to a kind of user type；And
First parameter set includes each resource distribution plan that the Component service includes in the first resource configuration plan Handling capacity, delay time, required cost under slightly and required total resources；
Second parameter set includes each resource distribution plan that the Component service includes in the Secondary resource configuration plan Handling capacity, delay time, required cost under slightly and required total resources；
The third parameter set include handling capacity of the Component service under the first resource configuration plan, delay time, Required cost and required total resources；
4th parameter set include handling capacity of the Component service under the Secondary resource configuration plan, delay time, Required cost and required total resources.
7. according to the method described in claim 6, it is characterized in that, the parameter threshold condition includes：
Wherein, XlIndicate that first of resource allocation strategy that resource allocation program includes, first of resource allocation strategy are directed to l Kind user type；L indicates user type sum；Indicate the l that the Component service includes in resource allocation program A resource allocation strategy XlUnder required total resources；Indicate that the Component service includes in resource allocation program First of resource allocation strategy XlUnder handling capacity；Indicate that the Component service includes in resource allocation program First of resource allocation strategy XlUnder delay time；Indicate the Component service includes in resource allocation program L resource allocation strategy XlUnder required cost；CsourceIndicate the preset required total resources threshold value for the system；Indicate the preset throughput threshold for being directed to l kind user types；It indicates preset and is directed to l kind user class The delay time threshold value of type；Indicate the preset required cost threshold value for being directed to l kind user types；CcostIndicate preset For the required cost threshold value of the system.
8. method according to any one of claim 1-3, which is characterized in that the basis has highest dominance hierarchy Resource allocation program subset obtains target resource configuration plan set, including：
Multiple resource allocation programs, which are chosen, from the resource allocation program subset with highest dominance hierarchy carries out intersection change It is different, generate multiple derivative resource allocation programs；
The multiple derivative resource allocation programs generated are added in derivative resource allocation program set, wherein the derivative provides Source configuration plan set is initially empty set；
By the resource allocation program subset with highest dominance hierarchy and the derivative resource allocation program union of sets collection Be formed as the target resource configuration plan set.
9. according to the method described in claim 2, it is characterized in that, described obtain from the target resource configuration plan set For the target resource configuration strategy of the Component service, including：
The sequence of the priority of each resource allocation program in plan set from high to low is configured according to the target resource, is obtained Take the resource allocation program of predetermined quantity as candidate resource configuration plan；
Obtain user type information；
In the works from the configuration of each candidate resource, the resource allocation strategy to match with the user type information is filtered out；
Obtain preference setting information of the user to resource distribution；
From the resource allocation strategy to match with the user type information, filter out and the preference setting information phase Matched resource allocation strategy, as the target resource configuration strategy.
10. a kind of resource allocation device for the software systems based on service, which is characterized in that the device includes：
First acquisition module, multiple and different resource allocation programs for obtaining the Component service for the system are formed just Beginning resource allocation program set, wherein each resource allocation program includes the resource allocation strategy of the Component service；
Subset division module, for the initial resource to be configured to each resource allocation program in plan set according to domination etc. Grade is divided, and forms multiple resource allocation program subsets, wherein the resource allocation program in same resource allocation program subset Dominance hierarchy having the same；
Second acquisition module, for according to the resource allocation program subset with highest dominance hierarchy, obtaining target resource configuration Plan set, wherein dominance hierarchy is higher, and corresponding resource allocation program has more excellent service quality；
Third acquisition module is provided for configuring acquisition in plan set from the target resource for the target of the Component service Source configuration strategy；And
Resource distribution module, for being that the Component service distributes resource according to the target resource configuration strategy.
11. device according to claim 10, which is characterized in that each resource allocation program includes multiple moneys Source configuration strategy, wherein each resource allocation strategy is directed to a kind of user type.
12. device according to claim 10, which is characterized in that described device further includes：
Judgment module, the resource allocation program subset for having highest dominance hierarchy in the basis obtain target resource and match After setting plan set, and in the target for configuring and obtaining during plan is gathered for the Component service from the target resource Before resource allocation strategy, judge whether actual cycle number is equal to preset cycle-index, wherein the actual cycle number Initial value be 0, the preset cycle-index be greater than or equal to 1；
Circular flow module is used in the case where the actual cycle number is not equal to the preset cycle-index, by institute It states actual cycle number and adds 1, and target resource configuration plan set is planned as new initial resource configuration after gathering, The subset division module is reruned, until the actual cycle number is equal to the preset cycle-index.
13. according to the device described in any one of claim 10-12, which is characterized in that the subset division module includes：
Dominance relation determination sub-module, for determining the dominance relation between each resource allocation program, wherein described In dominance relation, the dominance hierarchy of Zhi Peifang is higher than the dominance hierarchy of subject side；
Subset division submodule, for according to the dominance relation, will be divided with the resource allocation program of identical dominance hierarchy At a resource allocation program subset, wherein it is not dominated mutually between the resource allocation program in same resource allocation program subset, Also, any resource allocation program in the resource allocation program subset with higher dominance hierarchy can be dominated with relatively low branch Whole resource allocation programs in resource allocation program subset with grade.
14. device according to claim 13, which is characterized in that the dominance relation determination sub-module includes：
Resource allocation program selection unit chooses first resource configuration plan for being configured from the initial resource in plan set It configures and plans with Secondary resource；
First parameter determination unit, for determining the first parameter set and the second parameter set, wherein first parameter set includes institute State QoS parameter and required resource of the Component service under the resource allocation strategy that the first resource configuration plan includes Total amount, second parameter set include the Component service under the resource allocation strategy that the Secondary resource configuration plan includes QoS parameter and required total resources；
Second parameter determination unit for determining third parameter set according to first parameter set, and is joined according to described second Manifold determines the 4th parameter set, wherein the third parameter set includes the Component service in the first resource configuration plan Under QoS parameter and required total resources, the 4th parameter set include that the Component service is matched in the Secondary resource QoS parameter under the plan of setting and required total resources；
First judging unit, for judging whether first parameter set and second parameter set meet preset parameter respectively Threshold condition；
First dominance relation determination unit, described in meeting when one of first parameter set and second parameter set Parameter threshold condition, and when another one is unsatisfactory for the parameter threshold condition, determine the parameter for meeting the parameter threshold condition The corresponding resource allocation program of collection dominates the resource allocation program being unsatisfactory for corresponding to the parameter set of the parameter threshold condition；
Second judgment unit is satisfied by the parameter threshold condition for working as first parameter set and second parameter set When, judge that the third parameter set and the 4th parameter set are to meet the first preset condition or the second preset condition, wherein First preset condition is that each parameter in the third parameter set is not inferior to the parameter of the same race in the 4th parameter set, And at least one of third parameter set parameter is better than the parameter of the same race in the 4th parameter set；Described second is default Condition is that each parameter in the 4th parameter set is not inferior to the parameter of the same race in the third parameter set, and the described 4th At least one of parameter set parameter is better than the parameter of the same race in the third parameter set；
Second dominance relation determination unit meets described first and presets for working as the third parameter set and the 4th parameter set When condition, determine that the first resource configuration plan dominates the Secondary resource configuration plan；
Third dominance relation determination unit meets described second and presets for working as the third parameter set and the 4th parameter set When condition, determine that the Secondary resource configuration plan dominates the first resource configuration plan；
4th dominance relation determination unit is unsatisfactory for described first in advance for working as the third parameter set and the 4th parameter set If condition and when being unsatisfactory for second preset condition, or when first parameter set and second parameter set are discontented When the foot parameter threshold condition, determines and do not propped up mutually between the first resource configuration plan and the Secondary resource configuration plan Match；
Circular flow unit, for the resource allocation program selection unit that reruns, until determining each resource Until dominance relation between configuration plan, wherein the first resource configuration plan and Secondary resource configuration plan chosen every time It is not all the same.
15. device according to claim 14, which is characterized in that each resource allocation program includes multiple moneys Source configuration strategy, wherein each resource allocation strategy is directed to a kind of user type；And
First parameter set includes each resource distribution plan that the Component service includes in the first resource configuration plan Handling capacity, delay time, required cost under slightly and required total resources；
Second parameter set includes each resource distribution plan that the Component service includes in the Secondary resource configuration plan Handling capacity, delay time, required cost under slightly and required total resources；
The third parameter set include handling capacity of the Component service under the first resource configuration plan, delay time, Required cost and required total resources；
4th parameter set include handling capacity of the Component service under the Secondary resource configuration plan, delay time, Required cost and required total resources.
16. device according to claim 15, which is characterized in that the parameter threshold condition includes：
Wherein, XlIndicate that first of resource allocation strategy that resource allocation program includes, first of resource allocation strategy are directed to l Kind user type；L indicates user type sum；Indicate the l that the Component service includes in resource allocation program A resource allocation strategy XlUnder required total resources；Indicate that the Component service includes in resource allocation program First of resource allocation strategy XlUnder handling capacity；Indicate that the Component service includes in resource allocation program First of resource allocation strategy XlUnder delay time；Indicate the l that the Component service includes in resource allocation program A resource allocation strategy XlUnder required cost；CsourceIndicate the preset required total resources threshold value for the system；Indicate the preset throughput threshold for being directed to l kind user types；It indicates preset and is directed to l kind user class The delay time threshold value of type；Indicate the preset required cost threshold value for being directed to l kind user types；CcostIndicate preset For the required cost threshold value of the system.
17. according to the device described in any one of claim 10-12, which is characterized in that second acquisition module includes：
Cross and variation submodule, for choosing multiple resources from the resource allocation program subset with highest dominance hierarchy Configuration plan carries out cross and variation, generates multiple derivative resource allocation programs；
Derivative resource allocation program adds submodule, for the multiple derivative resource allocation programs generated to be added to derivative money In source configuration plan set, wherein the derivative resource allocation program set is initially empty set；
Target resource configuration plan set forms submodule, for the resource allocation program with highest dominance hierarchy is sub Collection is formed as the target resource configuration plan set with the derivative resource allocation program union of sets collection.
18. according to the devices described in claim 11, which is characterized in that the third acquisition module includes：
Candidate resource configuration plan acquisition submodule, for matching according to each resource in the target resource configuration plan set The sequence of the priority of plan from high to low is set, obtains the resource allocation program of predetermined quantity as candidate resource configuration plan；
User type information acquisition submodule, for obtaining user type information；
First screening submodule, in the works, being filtered out and the user type information phase from the configuration of each candidate resource The resource allocation strategy matched；
Preference setting information acquisition submodule, for obtaining preference setting information of the user to resource distribution；
Second screening submodule, for from the resource allocation strategy to match with the user type information, filtering out The resource allocation strategy to match with the preference setting information, as the target resource configuration strategy.
CN201510714321.2A 2015-10-28 2015-10-28 Resource allocation methods and device for the software systems based on service Active CN105262623B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510714321.2A CN105262623B (en) 2015-10-28 2015-10-28 Resource allocation methods and device for the software systems based on service

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510714321.2A CN105262623B (en) 2015-10-28 2015-10-28 Resource allocation methods and device for the software systems based on service

Publications (2)

Publication Number Publication Date
CN105262623A CN105262623A (en) 2016-01-20
CN105262623B true CN105262623B (en) 2018-09-14

Family

ID=55102140

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510714321.2A Active CN105262623B (en) 2015-10-28 2015-10-28 Resource allocation methods and device for the software systems based on service

Country Status (1)

Country Link
CN (1) CN105262623B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11388054B2 (en) * 2019-04-30 2022-07-12 Intel Corporation Modular I/O configurations for edge computing using disaggregated chiplets

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102387173A (en) * 2010-09-01 2012-03-21 中国移动通信集团公司 MapReduce system and method and device for scheduling tasks thereof
CN102780610A (en) * 2012-05-21 2012-11-14 上海斐讯数据通信技术有限公司 Gateway QoS (Quality of Service) guaranteeing method and packet switching network system
CN104283946A (en) * 2014-09-26 2015-01-14 东北大学 Resource self-adaptive adjusting system and method of multiple virtual machines under single physical machine
US9379995B2 (en) * 2012-09-11 2016-06-28 Vmware, Inc. Resource allocation diagnosis on distributed computer systems based on resource hierarchy

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102387173A (en) * 2010-09-01 2012-03-21 中国移动通信集团公司 MapReduce system and method and device for scheduling tasks thereof
CN102780610A (en) * 2012-05-21 2012-11-14 上海斐讯数据通信技术有限公司 Gateway QoS (Quality of Service) guaranteeing method and packet switching network system
US9379995B2 (en) * 2012-09-11 2016-06-28 Vmware, Inc. Resource allocation diagnosis on distributed computer systems based on resource hierarchy
CN104283946A (en) * 2014-09-26 2015-01-14 东北大学 Resource self-adaptive adjusting system and method of multiple virtual machines under single physical machine

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
一种基于服务选取的SBS云资源优化分配方法;赵秀涛,张斌,张长胜;《软件学报》;20150415;第26卷(第4期);867-883页 *

Also Published As

Publication number Publication date
CN105262623A (en) 2016-01-20

Similar Documents

Publication Publication Date Title
Jatoth et al. Optimal fitness aware cloud service composition using an adaptive genotypes evolution based genetic algorithm
Zheng et al. Periodic hierarchical load balancing for large supercomputers
CN108090208A (en) Fused data processing method and processing device
Fan et al. Topology-aware deployment of scientific applications in cloud computing
CN112153700A (en) Network slice resource management method and equipment
Yin et al. A hybrid multiobjective discrete particle swarm optimization algorithm for a SLA-aware service composition problem
CN103346912A (en) Method, device and system for conducting warning correlation analysis
Tao et al. Concept, principle and application of dynamic configuration for intelligent algorithms
Pham et al. Applying Ant Colony System algorithm in multi-objective resource allocation for virtual services
Soualah et al. Online and batch algorithms for VNFs placement and chaining
KR101858715B1 (en) Management System for Service Resource and Method thereof
Armant et al. Semi-online task assignment policies for workload consolidation in cloud computing systems
US20180096052A1 (en) Systems and methods for management of multi-perspective customer segments
CN105488134A (en) Big data processing method and big data processing device
Wang et al. Enhanced ant colony algorithm for cost-aware data-intensive service provision
CN110717647A (en) Decision flow construction method and device, computer equipment and storage medium
Glantz et al. Algorithms for mapping parallel processes onto grid and torus architectures
Thai et al. Executing bag of distributed tasks on the cloud: Investigating the trade-offs between performance and cost
Miriam et al. A double min min algorithm for task metascheduler on hypercubic p2p grid systems
CN106325999A (en) Method and device for distributing resources of host machine
Liu et al. Service organization and recommendation using multi-granularity approach
CN105262623B (en) Resource allocation methods and device for the software systems based on service
Keerthika et al. A multiconstrained grid scheduling algorithm with load balancing and fault tolerance
Mistry et al. Qualitative economic model for long-term IaaS composition
Jung et al. Ostro: Scalable placement optimization of complex application topologies in large-scale data centers

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant