CN104023068A - Method of implementing passive mode elastic computing and resource scheduling in load balancing - Google Patents
Method of implementing passive mode elastic computing and resource scheduling in load balancing Download PDFInfo
- Publication number
- CN104023068A CN104023068A CN201410263629.5A CN201410263629A CN104023068A CN 104023068 A CN104023068 A CN 104023068A CN 201410263629 A CN201410263629 A CN 201410263629A CN 104023068 A CN104023068 A CN 104023068A
- Authority
- CN
- China
- Prior art keywords
- host computer
- fictitious host
- application service
- load balancing
- service pond
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Abstract
The invention provides a method of implementing passive mode elastic computing and resource scheduling in load balancing. The method comprises the following steps that: an IAAS (Infrastructure as a Service) cloud computing environment including a load balancing gateway, a virtual machine management center and a plurality of application service pools is constructed; the load balancing gateway receives a resource request from each client, and then the resource request is distributed to each application service pool according to a predetermined load balancing strategy, and meanwhile, a logging module is configured on the load balancing gateway, and the following information is recorded through the logging module: the number of online sessions, a throughout flow and the number of requests per second of a movable virtual machine; a business pressure model is established based on the logging information, if the business pressure model reaches a critical trigger threshold of a state, a scheduling command is actively transmitted to the virtual machine management center, and resources of the virtual machine are elastically scheduled. In the premise of greatly improving the practicability and the accuracy of computing resources under the cloud environment, the load and the expenditure of the whole system are remarkably reduced, and the service quality of a whole cloud computing platform is improved.
Description
Technical field
The invention belongs to communication technical field, be specifically related to realize in a kind of load balancing the method for Passive Mode elastometer operator resource scheduling.
Background technology
Under cloud computing environment, computer hardware resource is virtual after, obtain some virtual machines, many virtual machine recomposition virtual resource ponds, then carry out calculation task by virtual resource pond.How making full use of the virtual resource in virtual resource pond, improve cloud service platform parallel processing capability, is the focus of current research.
In prior art, conventionally the method for taking is: for every active virtual machine setting in each virtual resource pond independently detects demons, by cpu resource or the memory source of this detection demons collection activity virtual machine, statistics gatherer arrives again cpu resource or memory source, obtain the pressure state of this active virtual machine; Then the pressure state of each active virtual machine is fed back to load-balancing device, load-balancing device is analyzed the pressure state of each the active virtual machine receiving, judge whether a certain virtual resource pond transships, and then determine whether to need in virtual resource pond, increase or reduce virtual machine, realize the load balancing in each virtual resource pond.
Although said method can be by constantly monitoring the pressure state of each virtual machine, and then adjust virtual resource allocation situation, on certain procedures, improve the utilance of virtual resource, still, due to the current resource situation of the continuous collection activity virtual machine of needs, increase the weight of the burden of whole system; In addition, for each active virtual machine setting independently detects demons, also strengthened the expense of whole system.
Summary of the invention
The defect existing for prior art, the invention provides a kind of method that realizes the scheduling of Passive Mode elastometer operator resource in load balancing, in order to overcome the problems referred to above.
The technical solution used in the present invention is as follows:
The invention provides a kind of method that realizes the scheduling of Passive Mode elastometer operator resource in load balancing, comprise the following steps:
S1, builds IAAS cloud computing environment, comprising: load balancing gateway, Virtual Machine Manager center and multiple application services pond; Each application service pond is by several member compositions, and described member is fictitious host computer; And the fictitious host computer that can participate in scheduling in each application service pond disposes unified participation scheduling identification ID, by this participation scheduling identification ID, differentiation can participate in the fictitious host computer of scheduling and not participate in the fictitious host computer of scheduling; In each application service pond, fictitious host computer has two states: active state and inactive state, and the fictitious host computer of active state is made a comment or criticism at the fictitious host computer of operation; The fictitious host computer of inactive state is further divided into closed condition and is gradually moved back state, the fictitious host computer of closed condition refers to the fictitious host computer having rolled off the production line, the fictitious host computer that gradually moves back state refers to no longer to receive new client requests, the client requests that is currently connected to self is continued to process, and waits to discharge all resources after being disposed and enter closed condition;
S2, described load balancing gateway is configured to lower initial configuration parameter by configuration module: elasticity is calculated the control option that triggers minimum limit value, startup or the forbidding elasticity computing function of active members in threshold values, members IP that each application service pond comprises, active members IP that each application service pond comprises, each application service pond;
S3, in the time that the control option of described startup or forbidding elasticity computing function is set to starting state, described load balancing gateway receives the resource request from each client, then according to default load balancing, described resource request is distributed to each application service pond, described load balancing gateway configuration has logging modle simultaneously, records following information by logging modle: on-line session number, handle up flow and the number of request per second of movable fictitious host computer in each application service pond;
Meanwhile, described load balancing gateway reads described configuration module, obtains the active members IP that each application service pond comprises, and obtains the active members quantity that each application service pond comprises, and for any one application service pond I, all carries out following operation:
S31, first described load balancing gateway judges whether in described application service pond I, to increase a fictitious host computer, and step is as follows:
S311, establishing fictitious host computer quantity movable under the I current state of application service pond is n;
S312, reads described logging modle, obtains following parameter:
Ci: the on-line session number that represents fictitious host computer i movable in the I of application service pond;
Bi: the flow of handling up that represents fictitious host computer i movable in the I of application service pond;
Ri: the number of request per second that represents fictitious host computer i movable in the I of application service pond;
S313, the elasticity reading in described configuration module is calculated triggering threshold values, specifically obtains following parameter:
Sch: the on-line session that every fictitious host computer of expression standard can be supported is counted the threshold values upper limit;
Sbh: the flow threshold values upper limit that every fictitious host computer of expression standard can be supported;
Srh: the number of request threshold values upper limit per second that every fictitious host computer of expression standard can be supported;
S314, calculates δ value by formula (), and wherein, δ represents whether to increase the logical value of fictitious host computer, if δ value is 1, carries out S315, otherwise, carry out S316;
(1);
S315, described load balancing gateway sends to described Virtual Machine Manager center to the dispatch command that increases a fictitious host computer in the I of application service pond;
Fictitious host computer of described Virtual Machine Manager center Remote Wake Up, and be increased in described application service pond I, the state of putting the fictitious host computer newly increasing is active state, upgrades the following parameter in described configuration module simultaneously: the members IP that application service pond I comprises, the active members IP that application service pond I comprises;
S316, described load balancing gateway further judges whether fictitious host computer quantity n movable under the I current state of application service pond is greater than the minimum limit value of active members in the set application service pond I of configuration module, if be not more than, carries out S33; Otherwise, carry out S32;
S32, secondly described load balancing gateway judges whether to reduce a fictitious host computer in described application service pond I, and step is as follows:
S321, establishing fictitious host computer quantity movable under the I current state of application service pond is n;
S322, reads described logging modle, obtains following parameter:
Ci: the on-line session number that represents fictitious host computer i movable in the I of application service pond;
Bi: the flow of handling up that represents fictitious host computer i movable in the I of application service pond;
Ri: the number of request per second that represents fictitious host computer i movable in the I of application service pond;
S323, the elasticity reading in described configuration module is calculated triggering threshold values, specifically obtains following parameter:
Scl: the on-line session that every fictitious host computer of expression standard can be supported is counted lower threshold;
Sbl: the flow lower threshold that every fictitious host computer of expression standard can be supported;
Srl: the number of request lower threshold per second that every fictitious host computer of expression standard can be supported;
S324, calculates γ value by formula (two), and wherein, γ represents whether to reduce the logical value of fictitious host computer, if δ value is 1, carries out S325, otherwise, carry out S33;
(2)
S325, described load balancing gateway further judges the lightest particular virtual main frame of current pressure in the service pool I that is applied, and then sends to described Virtual Machine Manager center the dispatch command that reduces particular virtual main frame described in the I of application service pond;
Long-range this particular virtual main frame of cutting out in described Virtual Machine Manager center, puts the state of described particular virtual main frame for gradually moving back state; In the time that described particular virtual host process completes the client requests that is currently connected to self, close described particular virtual main frame, simultaneously, the state of putting described particular virtual main frame is closed condition, upgrades the following parameter in described configuration module simultaneously: the members IP that application service pond I comprises, the active members IP that application service pond I comprises; Then carry out S33;
S33, described load balancing gateway finishes epicycle flexible scheduling, and in the time that arrive in next dispatching cycle, S3 step is carried out in circulation.
Preferably, in S325, described load balancing gateway further judges the lightest particular virtual main frame of current pressure in the service pool I that is applied, and specifically comprises the following steps:
S3251, reads described logging modle, obtains following parameter:
Ci: the on-line session number that represents fictitious host computer i movable in the I of application service pond;
Bi: the flow of handling up that represents fictitious host computer i movable in the I of application service pond;
Ri: the number of request per second that represents fictitious host computer i movable in the I of application service pond;
S3252, reads the set initial configuration parameter of described configuration module, obtains following parameter:
Δ α is the weight coefficient of on-line session number;
Δ β is flow weight coefficient;
Δ γ is that current request is processed number weight coefficient;
S3253, calculates Ri by formula (three), and wherein, Ri represents the definite fictitious host computer of closing and discharging of needs, is the particular virtual main frame that current pressure is the lightest;
(3).
Preferably, in S33, described load balancing gateway finishes epicycle flexible scheduling, and in the time that arrive in next dispatching cycle, S3 step is carried out in circulation, is specially:
Described configuration module is also provided with quiet period initial parameter value;
In epicycle scheduling process, when execution wakes up or closes after the operation of a certain virtual machine, one, interval quiet period, then carry out next round scheduling.
In load balancing provided by the invention, realize the method for Passive Mode elastometer operator resource scheduling, have the following advantages: the load balancing gateway with elasticity computing capability is combined with Intel Virtualization Technology, load balancing gateway is in the service request process of distributing from client, directly record the on-line session number of fictitious host computer movable in each application service pond, handle up flow and number of request per second, and then judge whether in application service pond, to increase or to reduce fictitious host computer, greatly promoting under cloud environment under the practicality of computational resource and the prerequisite of accuracy, burden and the expense of whole system are also obviously reduced, improve the service quality of whole cloud computing platform.
Brief description of the drawings
Fig. 1 is the method flow schematic diagram of realizing the scheduling of Passive Mode elastometer operator resource in load balancing provided by the invention.
Embodiment
Below in conjunction with accompanying drawing, the present invention is described in detail:
As shown in Figure 1, the invention provides a kind of method that realizes the scheduling of Passive Mode elastometer operator resource in load balancing, the flexibility of virtual resource scheduling in the perception of load-balancing technique, subsequent use service pool and virtual platform is closely linked, calculates solution for cloud computing environment provides more complete elasticity.
Specifically comprise the following steps:
S1, builds IAAS cloud computing environment, comprising: load balancing gateway, Virtual Machine Manager center and multiple application services pond; Each application service pond is by several member compositions, and described member is fictitious host computer; And the fictitious host computer that can participate in scheduling in each application service pond disposes unified participation scheduling identification ID, by this participation scheduling identification ID, differentiation can participate in the fictitious host computer of scheduling and not participate in the fictitious host computer of scheduling; In each application service pond, fictitious host computer has two states: active state and inactive state, and the fictitious host computer of active state is made a comment or criticism at the fictitious host computer of operation; The fictitious host computer of inactive state is further divided into closed condition and is gradually moved back state, the fictitious host computer of closed condition refers to the fictitious host computer having rolled off the production line, the fictitious host computer that gradually moves back state refers to no longer to receive new client requests, the client requests that is currently connected to self is continued to process, and waits to discharge all resources after being disposed and enter closed condition; In the present invention, one of fictitious host computer existence is set and gradually moves back state, can guarantee that lasting impact minimizes on client business.
The fictitious host computer that can participate in flexible resource scheduling in application service pond arranges special identifier, is in order to facilitate load balancing gateway in the time carrying out flexible scheduling, can to identify the fictitious host computer that can allow flexible scheduling.
S2, described load balancing gateway is configured to lower initial configuration parameter by configuration module: elasticity is calculated the control option that triggers minimum limit value, startup or the forbidding elasticity computing function of active members in threshold values, members IP that each application service pond comprises, active members IP that each application service pond comprises, each application service pond;
The members comprising due to application service pond and active members are dynamic change, therefore, in the time that members and active members change, need to dynamically update in real time this two parameters, facilitate follow-up load balancing gateway to carry out flexible scheduling according to these two parameters.
For guaranteeing business idle period of time, the state normal and service of guaranteeing in application service pond is not interrupted, the minimum limit value that application service pond active members need to be set, when application service pond, active members quantity reaches after minimum value, and the lowest limit threshold parameter of no longer calculating according to elasticity is continued to close fictitious host computer.Certainly, meet the minimum limit value of application service pond active members if do not participate in the active members quantity of scheduling, elasticity computing engines can be closed the fictitious host computer of all participation scheduling.
S3, in the time that the control option of described startup or forbidding elasticity computing function is set to starting state, described load balancing gateway receives the resource request from each client, then according to default load balancing, described resource request is distributed to each application service pond, described load balancing gateway configuration has logging modle simultaneously, records following information by logging modle: on-line session number, handle up flow and the number of request per second of movable fictitious host computer in each application service pond;
The most important innovation of the present invention is: load-balancing device, in the time of dispensing applications flow, is recorded the parameter of these three important its pressure conditions of reflection in detail: on-line session number (Concurrent Connection), the flow of handling up (Through Put) and the number of request per second (Request per Second) of movable fictitious host computer in application service pond; Then parameter and the automatic start-stop of virtual machine of these reflection pressure conditions are linked together, realize the flexible scheduling of virtual resource, completely without increasing load-balancing device overhead; In addition, also need in the fictitious host computer in application service pond, special system resource monitoring demons be installed, with dispose specially the pattern of demons in prior art in fictitious host computer compared with, the present invention has better IaaS system compatibility.
Simultaneously, described load balancing gateway reads described configuration module, obtain the active members IP that each application service pond comprises, obtain the active members quantity that each application service pond comprises, for any one application service pond I, set and increase the fictitious host computer priority in application service pond higher than the default convention of closing and discharge the fictitious host computer in an application service pond.That is to say, first judging whether to increase fictitious host computer, further judges whether if NO to reduce by a fictitious host computer again, operation below concrete execution:
S31, first described load balancing gateway judges whether in described application service pond I, to increase a fictitious host computer, and step is as follows:
S311, establishing fictitious host computer quantity movable under the I current state of application service pond is n;
S312, reads described logging modle, obtains following parameter:
Ci: the on-line session number that represents fictitious host computer i movable in the I of application service pond;
Bi: the flow of handling up that represents fictitious host computer i movable in the I of application service pond;
Ri: the number of request per second that represents fictitious host computer i movable in the I of application service pond;
S313, the elasticity reading in described configuration module is calculated triggering threshold values, specifically obtains following parameter:
Sch: the on-line session that every fictitious host computer of expression standard can be supported is counted the threshold values upper limit;
Sbh: the flow threshold values upper limit that every fictitious host computer of expression standard can be supported;
Srh: the number of request threshold values upper limit per second that every fictitious host computer of expression standard can be supported;
S314, calculates δ value by formula (), and wherein, δ represents whether to increase the logical value of fictitious host computer, if δ value is 1, carries out S315, otherwise, carry out S316;
(1);
S315, described load balancing gateway sends to described Virtual Machine Manager center to the dispatch command that increases a fictitious host computer in the I of application service pond; In practical application, for the balanced gateway of proof load and Virtual Machine Manager center normal communication, can at load balancing gateway, following communications parameter be set in advance: IP address, Virtual Machine Manager center, virtual machine store path, keeper's account at Virtual Machine Manager center, i.e. usemame/password.
Fictitious host computer of described Virtual Machine Manager center Remote Wake Up, and be increased in described application service pond I, the state of putting the fictitious host computer newly increasing is active state, upgrades the following parameter in described configuration module simultaneously: the members IP that application service pond I comprises, the active members IP that application service pond I comprises;
S316, described load balancing gateway further judges whether fictitious host computer quantity n movable under the I current state of application service pond is greater than the minimum limit value of active members in the set application service pond I of configuration module, if be not more than, carries out S33; Otherwise, carry out S32;
S32, secondly described load balancing gateway judges whether to reduce a fictitious host computer in described application service pond I, and step is as follows:
S321, establishing fictitious host computer quantity movable under the I current state of application service pond is n;
S322, reads described logging modle, obtains following parameter:
Ci: the on-line session number that represents fictitious host computer i movable in the I of application service pond;
Bi: the flow of handling up that represents fictitious host computer i movable in the I of application service pond;
Ri: the number of request per second that represents fictitious host computer i movable in the I of application service pond;
S323, the elasticity reading in described configuration module is calculated triggering threshold values, specifically obtains following parameter:
Scl: the on-line session that every fictitious host computer of expression standard can be supported is counted lower threshold;
Sbl: the flow lower threshold that every fictitious host computer of expression standard can be supported;
Srl: the number of request lower threshold per second that every fictitious host computer of expression standard can be supported;
S324, calculates γ value by formula (two), and wherein, γ represents whether to reduce the logical value of fictitious host computer, if δ value is 1, carries out S325, otherwise, carry out S33;
(2)
S325, described load balancing gateway further judges the lightest particular virtual main frame of current pressure in the service pool I that is applied, and then sends to described Virtual Machine Manager center the dispatch command that reduces particular virtual main frame described in the I of application service pond;
Long-range this particular virtual main frame of cutting out in described Virtual Machine Manager center, puts the state of described particular virtual main frame for gradually moving back state; In the time that described particular virtual host process completes the client requests that is currently connected to self, close described particular virtual main frame, simultaneously, the state of putting described particular virtual main frame is closed condition, upgrades the following parameter in described configuration module simultaneously: the members IP that application service pond I comprises, the active members IP that application service pond I comprises; Then carry out S33;
In this step, load balancing gateway can judge the lightest particular virtual main frame of current pressure in the service pool I that is applied by the following method:
S3251, reads described logging modle, obtains following parameter:
Ci: the on-line session number that represents fictitious host computer i movable in the I of application service pond;
Bi: the flow of handling up that represents fictitious host computer i movable in the I of application service pond;
Ri: the number of request per second that represents fictitious host computer i movable in the I of application service pond;
S3252, reads the set initial configuration parameter of described configuration module, obtains following parameter:
Δ α is the weight coefficient of on-line session number;
Δ β is flow weight coefficient;
Δ γ is that current request is processed number weight coefficient;
S3253, calculates Ri by formula (three), and wherein, Ri represents the definite fictitious host computer of closing and discharging of needs, is the particular virtual main frame that current pressure is the lightest;
(3).
S33, described load balancing gateway finishes epicycle flexible scheduling, and in the time that arrive in next dispatching cycle, S3 step is carried out in circulation.
Configuration module is also provided with quiet period initial parameter value; In epicycle scheduling process, when execution wakes up or closes after the operation of a certain virtual machine, one, interval quiet period, then carry out next round scheduling.
Due to the dynamic characteristic of application system, for preventing that threshold parameter from shaking extremely, cause fictitious host computer On/Off too frequent, therefore, in the present invention, configuration module is also provided with quiet period initial parameter value; In epicycle scheduling process, when execution wakes up or closes after the operation of a certain virtual machine, one, interval quiet period, then carry out next round scheduling.
Silent period parameters can be revised, and system default quiet period configuration parameter can illustrate as follows:
(a) quiet period strategy is described: record a secondary data every 3 seconds logging modles, the record of a decision-making period is continuously 3 minutes (60 times), calculates the average of this cycle internal valve value record as the threshold values data of scheduling of resource; After increasing or closing fictitious host computer behavior generation, system is mourned in silence after 2 record periods, continues record and carries out threshold values judgement.
(b) close the strategy of fictitious host computer: for ensureing the integrality of applied business, need to close fictitious host computer time, first fictitious host computer is designated to the state that gradually moves back, now this fictitious host computer is not as movable fictitious host computer.Only have when this fictitious host computer without be flexibly connected time, just can formally close this fictitious host computer.
In load balancing provided by the invention, realize the method for Passive Mode elastometer operator resource scheduling, the situation of change of the each fictitious host computer flow load of load balancing gateway real time record, trigger by artificial default triggering threshold values as application service pond flexible scheduling, interlock Virtual Machine Manager center, the long-range member who arouses or close subsequent use service pool, has following characteristics:
(1) load balancing gateway, by API and vCenter or the interlock of XenCenter Virtual Machine Manager center, is realized the unlatching of fictitious host computer or closes;
(2) by linking with Virtual Machine Manager center, load balancing gateway can be understood fictitious host computer deployment scenario automatically, automatically selects the member of fictitious host computer as application service pond;
(3) load balancing gateway is the configuration that application service pond creates elasticity calculating triggering threshold parameter, builds traffic pressure model, i.e. formula () or formula (two) by parameter combinations; After traffic pressure model is enabled, if reach the critical triggering threshold values of certain state, load balancing gateway initiatively sends dispatch command, flexible scheduling resources of virtual machine to Virtual Machine Manager center;
(4) load balancing gateway can be monitored the member's running status in each application service pond in real time, comprises the health status (application health examination) of number of nodes (Pool member) and node, for flexible scheduling provides foundation;
(5) by configure elasticity computing capability in load balancing gateway, can realize the computational resource automatic governing in IAAS cloud computing environment, thereby optimize the utilization rate of computational resource, reach the effect of the cloud computing IAAS platform of efficient energy-saving.
In load balancing provided by the invention, realize the method for Passive Mode elastometer operator resource scheduling, under the IAAS platform environment of realizing at multiple virtual machines, dispose multiple application services pond, by in load balancing gateway configuration elasticity computing function, can easily computational resource (being virtual machine) superfluous in certain application service pond be discharged, and dynamically allotment is given in the application service pond under the application system that need to increase computational resource, thereby maximum using computational resource, realizes intensive IAAS cloud computing environment.Also obviously reduce burden and the expense of whole system, improved the service quality of whole cloud computing platform.
The above is only the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, under the premise without departing from the principles of the invention; can also make some improvements and modifications, these improvements and modifications also should be looked protection scope of the present invention.
Claims (3)
1. a method that realizes the scheduling of Passive Mode elastometer operator resource in load balancing, is characterized in that, comprises the following steps:
S1, builds IAAS cloud computing environment, comprising: load balancing gateway, Virtual Machine Manager center and multiple application services pond; Each application service pond is by several member compositions, and described member is fictitious host computer; And the fictitious host computer that can participate in scheduling in each application service pond disposes unified participation scheduling identification ID, by this participation scheduling identification ID, differentiation can participate in the fictitious host computer of scheduling and not participate in the fictitious host computer of scheduling; In each application service pond, fictitious host computer has two states: active state and inactive state, and the fictitious host computer of active state is made a comment or criticism at the fictitious host computer of operation; The fictitious host computer of inactive state is further divided into closed condition and is gradually moved back state, the fictitious host computer of closed condition refers to the fictitious host computer having rolled off the production line, the fictitious host computer that gradually moves back state refers to no longer to receive new client requests, the client requests that is currently connected to self is continued to process, and waits to discharge all resources after being disposed and enter closed condition;
S2, described load balancing gateway is configured to lower initial configuration parameter by configuration module: elasticity is calculated the control option that triggers minimum limit value, startup or the forbidding elasticity computing function of active members in threshold values, members IP that each application service pond comprises, active members IP that each application service pond comprises, each application service pond;
S3, in the time that the control option of described startup or forbidding elasticity computing function is set to starting state, described load balancing gateway receives the resource request from each client, then according to default load balancing, described resource request is distributed to each application service pond, described load balancing gateway configuration has logging modle simultaneously, records following information by logging modle: on-line session number, handle up flow and the number of request per second of movable fictitious host computer in each application service pond;
Meanwhile, described load balancing gateway reads described configuration module, obtains the active members IP that each application service pond comprises, and obtains the active members quantity that each application service pond comprises, and for any one application service pond I, all carries out following operation:
S31, first described load balancing gateway judges whether in described application service pond I, to increase a fictitious host computer, and step is as follows:
S311, establishing fictitious host computer quantity movable under the I current state of application service pond is n;
S312, reads described logging modle, obtains following parameter:
Ci: the on-line session number that represents fictitious host computer i movable in the I of application service pond;
Bi: the flow of handling up that represents fictitious host computer i movable in the I of application service pond;
Ri: the number of request per second that represents fictitious host computer i movable in the I of application service pond;
S313, the elasticity reading in described configuration module is calculated triggering threshold values, specifically obtains following parameter:
Sch: the on-line session that every fictitious host computer of expression standard can be supported is counted the threshold values upper limit;
Sbh: the flow threshold values upper limit that every fictitious host computer of expression standard can be supported;
Srh: the number of request threshold values upper limit per second that every fictitious host computer of expression standard can be supported;
S314, calculates δ value by formula (), and wherein, δ represents whether to increase the logical value of fictitious host computer, if δ value is 1, carries out S315, otherwise, carry out S316;
(1);
S315, described load balancing gateway sends to described Virtual Machine Manager center to the dispatch command that increases a fictitious host computer in the I of application service pond;
Fictitious host computer of described Virtual Machine Manager center Remote Wake Up, and be increased in described application service pond I, the state of putting the fictitious host computer newly increasing is active state, upgrades the following parameter in described configuration module simultaneously: the members IP that application service pond I comprises, the active members IP that application service pond I comprises;
S316, described load balancing gateway further judges whether fictitious host computer quantity n movable under the I current state of application service pond is greater than the minimum limit value of active members in the set application service pond I of configuration module, if be not more than, carries out S33; Otherwise, carry out S32;
S32, secondly described load balancing gateway judges whether to reduce a fictitious host computer in described application service pond I, and step is as follows:
S321, establishing fictitious host computer quantity movable under the I current state of application service pond is n;
S322, reads described logging modle, obtains following parameter:
Ci: the on-line session number that represents fictitious host computer i movable in the I of application service pond;
Bi: the flow of handling up that represents fictitious host computer i movable in the I of application service pond;
Ri: the number of request per second that represents fictitious host computer i movable in the I of application service pond;
S323, the elasticity reading in described configuration module is calculated triggering threshold values, specifically obtains following parameter:
Scl: the on-line session that every fictitious host computer of expression standard can be supported is counted lower threshold;
Sbl: the flow lower threshold that every fictitious host computer of expression standard can be supported;
Srl: the number of request lower threshold per second that every fictitious host computer of expression standard can be supported;
S324, calculates γ value by formula (two), and wherein, γ represents whether to reduce the logical value of fictitious host computer, if δ value is 1, carries out S325, otherwise, carry out S33;
(2)
S325, described load balancing gateway further judges the lightest particular virtual main frame of current pressure in the service pool I that is applied, and then sends to described Virtual Machine Manager center the dispatch command that reduces particular virtual main frame described in the I of application service pond;
Long-range this particular virtual main frame of cutting out in described Virtual Machine Manager center, puts the state of described particular virtual main frame for gradually moving back state; In the time that described particular virtual host process completes the client requests that is currently connected to self, close described particular virtual main frame, simultaneously, the state of putting described particular virtual main frame is closed condition, upgrades the following parameter in described configuration module simultaneously: the members IP that application service pond I comprises, the active members IP that application service pond I comprises; Then carry out S33;
S33, described load balancing gateway finishes epicycle flexible scheduling, and in the time that arrive in next dispatching cycle, S3 step is carried out in circulation.
2. in load balancing according to claim 1, realize the method for Passive Mode elastometer operator resource scheduling, it is characterized in that, in S325, described load balancing gateway further judges the lightest particular virtual main frame of current pressure in the service pool I that is applied, and specifically comprises the following steps:
S3251, reads described logging modle, obtains following parameter:
Ci: the on-line session number that represents fictitious host computer i movable in the I of application service pond;
Bi: the flow of handling up that represents fictitious host computer i movable in the I of application service pond;
Ri: the number of request per second that represents fictitious host computer i movable in the I of application service pond;
S3252, reads the set initial configuration parameter of described configuration module, obtains following parameter:
Δ α is the weight coefficient of on-line session number;
Δ β is flow weight coefficient;
Δ γ is that current request is processed number weight coefficient;
S3253, calculates Ri by formula (three), and wherein, Ri represents the definite fictitious host computer of closing and discharging of needs, is the particular virtual main frame that current pressure is the lightest;
(3).
3. the method that realizes the scheduling of Passive Mode elastometer operator resource in load balancing according to claim 1, is characterized in that, in S33, described load balancing gateway finishes epicycle flexible scheduling, in the time that arrive in next dispatching cycle, S3 step is carried out in circulation, is specially:
Described configuration module is also provided with quiet period initial parameter value;
In epicycle scheduling process, when execution wakes up or closes after the operation of a certain virtual machine, one, interval quiet period, then carry out next round scheduling.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410263629.5A CN104023068B (en) | 2014-06-13 | 2014-06-13 | A kind of method that Passive Mode elastic calculation scheduling of resource is realized in load balancing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410263629.5A CN104023068B (en) | 2014-06-13 | 2014-06-13 | A kind of method that Passive Mode elastic calculation scheduling of resource is realized in load balancing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104023068A true CN104023068A (en) | 2014-09-03 |
| CN104023068B CN104023068B (en) | 2017-12-15 |
Family
ID=51439637
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410263629.5A Active CN104023068B (en) | 2014-06-13 | 2014-06-13 | A kind of method that Passive Mode elastic calculation scheduling of resource is realized in load balancing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104023068B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015192556A1 (en) * | 2014-06-16 | 2015-12-23 | 中兴通讯股份有限公司 | Management method, management center and management system for cloud scheduling |
| CN105450728A (en) * | 2015-11-04 | 2016-03-30 | 合一网络技术(北京)有限公司 | Method and apparatus for realizing real device allocation of virtual machine |
| CN106254408A (en) * | 2015-06-12 | 2016-12-21 | 财团法人工业技术研究院 | Control method, network system and service platform for mobile edge calculation |
| CN106845746A (en) * | 2016-06-15 | 2017-06-13 | 曹大海 | A kind of cloud Workflow Management System for supporting extensive example intensive applications |
| CN107645731A (en) * | 2017-09-21 | 2018-01-30 | 北京邮电大学 | Load-balancing method based on self-organizing resource allocation in a kind of non-orthogonal multiple access system |
| CN108170520A (en) * | 2018-01-29 | 2018-06-15 | 北京搜狐新媒体信息技术有限公司 | A kind of cloud computing resources management method and device |
| CN110166435A (en) * | 2019-04-18 | 2019-08-23 | 杭州电子科技大学 | The mimicry Web gateway system and method for dynamic dispatching are carried out using load balancing |
| CN111357257A (en) * | 2017-08-31 | 2020-06-30 | 格林伊登美国控股有限责任公司 | System and method for load balancing media server instances |
| US10972535B2 (en) | 2018-08-20 | 2021-04-06 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and device for load balancing, and storage medium |
| CN114598656A (en) * | 2022-05-09 | 2022-06-07 | 张家港保税科技集团电子商务有限公司 | Load balancing method and system applied to delivery service platform |
| CN117076057A (en) * | 2023-10-18 | 2023-11-17 | 成都阿加犀智能科技有限公司 | AI service request scheduling method, device, equipment and medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050002399A1 (en) * | 2001-04-24 | 2005-01-06 | Atitania Ltd. | Method and apparatus for converting data between two dissimilar systems |
| CN101719081A (en) * | 2009-12-01 | 2010-06-02 | 北京大学 | Method for scheduling virtual machines |
| CN102244685A (en) * | 2011-08-11 | 2011-11-16 | 中国科学院软件研究所 | Distributed type dynamic cache expanding method and system supporting load balancing |
| CN102427475A (en) * | 2011-12-08 | 2012-04-25 | 曙光信息产业(北京)有限公司 | Load balance scheduling system in cloud computing environment |
| CN102932413A (en) * | 2012-09-26 | 2013-02-13 | 华为软件技术有限公司 | Computing resource allocation method, cloud management platform node and computing resource cluster |
| CN103051564A (en) * | 2013-01-07 | 2013-04-17 | 杭州华三通信技术有限公司 | Dynamic resource allocation method and device |
-
2014
- 2014-06-13 CN CN201410263629.5A patent/CN104023068B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050002399A1 (en) * | 2001-04-24 | 2005-01-06 | Atitania Ltd. | Method and apparatus for converting data between two dissimilar systems |
| CN101719081A (en) * | 2009-12-01 | 2010-06-02 | 北京大学 | Method for scheduling virtual machines |
| CN102244685A (en) * | 2011-08-11 | 2011-11-16 | 中国科学院软件研究所 | Distributed type dynamic cache expanding method and system supporting load balancing |
| CN102427475A (en) * | 2011-12-08 | 2012-04-25 | 曙光信息产业(北京)有限公司 | Load balance scheduling system in cloud computing environment |
| CN102932413A (en) * | 2012-09-26 | 2013-02-13 | 华为软件技术有限公司 | Computing resource allocation method, cloud management platform node and computing resource cluster |
| CN103051564A (en) * | 2013-01-07 | 2013-04-17 | 杭州华三通信技术有限公司 | Dynamic resource allocation method and device |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015192556A1 (en) * | 2014-06-16 | 2015-12-23 | 中兴通讯股份有限公司 | Management method, management center and management system for cloud scheduling |
| CN106254408A (en) * | 2015-06-12 | 2016-12-21 | 财团法人工业技术研究院 | Control method, network system and service platform for mobile edge calculation |
| CN105450728A (en) * | 2015-11-04 | 2016-03-30 | 合一网络技术(北京)有限公司 | Method and apparatus for realizing real device allocation of virtual machine |
| CN106845746A (en) * | 2016-06-15 | 2017-06-13 | 曹大海 | A kind of cloud Workflow Management System for supporting extensive example intensive applications |
| CN111357257B (en) * | 2017-08-31 | 2022-06-14 | 格林伊登美国控股有限责任公司 | System and method for load balancing media server instances |
| CN111357257A (en) * | 2017-08-31 | 2020-06-30 | 格林伊登美国控股有限责任公司 | System and method for load balancing media server instances |
| CN107645731A (en) * | 2017-09-21 | 2018-01-30 | 北京邮电大学 | Load-balancing method based on self-organizing resource allocation in a kind of non-orthogonal multiple access system |
| CN107645731B (en) * | 2017-09-21 | 2019-10-08 | 北京邮电大学 | Load-balancing method based on self-organizing resource allocation in a kind of NOMA system |
| CN108170520A (en) * | 2018-01-29 | 2018-06-15 | 北京搜狐新媒体信息技术有限公司 | A kind of cloud computing resources management method and device |
| US10972535B2 (en) | 2018-08-20 | 2021-04-06 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and device for load balancing, and storage medium |
| CN110166435A (en) * | 2019-04-18 | 2019-08-23 | 杭州电子科技大学 | The mimicry Web gateway system and method for dynamic dispatching are carried out using load balancing |
| CN114598656A (en) * | 2022-05-09 | 2022-06-07 | 张家港保税科技集团电子商务有限公司 | Load balancing method and system applied to delivery service platform |
| CN117076057A (en) * | 2023-10-18 | 2023-11-17 | 成都阿加犀智能科技有限公司 | AI service request scheduling method, device, equipment and medium |
| CN117076057B (en) * | 2023-10-18 | 2024-01-26 | 成都阿加犀智能科技有限公司 | AI service request scheduling method, device, equipment and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104023068B (en) | 2017-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104023068A (en) | Method of implementing passive mode elastic computing and resource scheduling in load balancing | |
| US11507430B2 (en) | Accelerated resource allocation techniques | |
| CN105634804B (en) | A kind of service dial testing method, device and system | |
| CN103294521B (en) | A kind of method reducing data center's traffic load and energy consumption | |
| Caglar et al. | iOverbook: intelligent resource-overbooking to support soft real-time applications in the cloud | |
| CN106506670B (en) | A kind of cloud platform virtual resource high speed dispatching method and system | |
| CN104572307B (en) | The method that a kind of pair of virtual resource carries out flexible scheduling | |
| Tsai | SEIRA: An effective algorithm for IoT resource allocation problem | |
| Guazzone et al. | Exploiting VM migration for the automated power and performance management of green cloud computing systems | |
| Nithya et al. | SDCF: A software-defined cyber foraging framework for cloudlet environment | |
| CN104038392A (en) | Method for evaluating service quality of cloud computing resources | |
| CN107018175A (en) | The dispatching method and device of mobile cloud computing platform | |
| CN106209482A (en) | A kind of data center monitoring method and system | |
| CN107659609A (en) | A kind of deep learning support platform and deep learning training method based on cloud computing | |
| Liu et al. | A fast algorithm for energy-saving offloading with reliability and latency requirements in multi-access edge computing | |
| CN106210129B (en) | A kind of current-limiting method and system based on Web server configuration | |
| CN105868004A (en) | Cloud computing based business system scheduling method and apparatus | |
| KR20220030050A (en) | Method and server for adjusting the allocation of computing resources to a plurality of vnf(virtualized network function) | |
| CN105302641A (en) | Node scheduling method and apparatus in virtual cluster | |
| CN109144661A (en) | A kind of deep learning management method based on docker | |
| CN103699433A (en) | Method and system for performing dynamic adjustment on number of tasks in Hadoop platform | |
| Hanafy et al. | A new infrastructure elasticity control algorithm for containerized cloud | |
| Somma et al. | When less is more: Core-restricted container provisioning for serverless computing | |
| Theeten et al. | Towards the optimization of a parallel streaming engine for telco applications | |
| KR20160032881A (en) | Cloud based intelligent power control system |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220111 Address after: 100176 2403, floor 24, building 1, yard 8, KEGU 1st Street, Beijing Economic and Technological Development Zone, Daxing District, Beijing (Yizhuang group, high-end industrial area of Beijing Pilot Free Trade Zone) Patentee after: Tongming Zhiyun (Beijing) Technology Co.,Ltd. Address before: 101111 Room 302, 3 / F, building 4, yard 29, Jinghai Second Road, Beijing Economic and Technological Development Zone, Daxing District, Beijing Patentee before: SINOGRID SOFTWARE SYSTEMS LTD. |