CN1744593B - Transmission link selecting method - Google Patents
Transmission link selecting method Download PDFInfo
- Publication number
- CN1744593B CN1744593B CN200410064285.1A CN200410064285A CN1744593B CN 1744593 B CN1744593 B CN 1744593B CN 200410064285 A CN200410064285 A CN 200410064285A CN 1744593 B CN1744593 B CN 1744593B
- Authority
- CN
- China
- Prior art keywords
- processing unit
- transmission link
- transport layer
- application layer
- load balancing
- 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.)
- Expired - Fee Related
Links
- 230000005540 biological transmission Effects 0.000 title claims abstract description 89
- 238000000034 method Methods 0.000 title claims abstract description 62
- 238000012545 processing Methods 0.000 claims abstract description 230
- 230000008569 process Effects 0.000 claims abstract description 23
- 230000006854 communication Effects 0.000 claims description 33
- 238000004891 communication Methods 0.000 claims description 22
- 238000001514 detection method Methods 0.000 claims description 5
- 230000005856 abnormality Effects 0.000 claims description 3
- 238000013523 data management Methods 0.000 abstract description 6
- 238000004883 computer application Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 4
- 238000012546 transfer Methods 0.000 description 4
- 238000007726 management method Methods 0.000 description 3
- 238000004148 unit process Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005538 encapsulation Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
Images
Abstract
The method separates load-sharing algorithm from load data management, and sets them at application layer and transport layer of processing unit respectively. The application layer cooperates with the transport layer to accomplish procedure of distributing process tasks of distributed application. The method includes steps: application layer detects whether there is a transmission link to destination process unit capable of processing the task; if yes, using the link to accomplish distribution tasks; otherwise, selecting proper load-sharing algorithm, application layer through transmission port requests transport layer for usable transmission link; using load-sharing algorithm, the transport layer selects a proper process unit from candidate set of maintained process units, and builds transmission link to the destination process unit; application layer saves and uses the transmission link to accomplish distribution. Features are: quick distribution and flexibility of system.
Description
Technical field
The present invention relates to the communication technology between the processing unit of computer distribution type application system, specially refer to the method that the transmission link when communicating by letter with a plurality of purpose processing units is selected.
Background technology
Along with the continuous development of information technology, the scale of computer application system is just at acceleration expansion, and in order to satisfy the needs that the computer application system complexity constantly rises, the modularization of application program and modularization have been trend of the times.The more and more module division of labor of specialization has appearred in a lot of computer application systems, the application program division of complexity is called a plurality of modules, handle respectively by a plurality of processing units, wherein each module only paid close attention to very little aspect of application system, for example, use special processing unit to insert the user, use special processing unit controlled state data or the like.Intercom mutually by various means of communication between each processing unit, result is combined, externally form an integral body, powerful service function is provided.Processing unit described herein also can be understood as usually said process.
The operating type of above-mentioned this kind streamline has improved the retractility of computer application system greatly, promptly bottleneck has appearred in the performance of certain processing unit in computer application system, as long as the disposal ability that increases this processing unit just can satisfy the requirement of system, therefore, can carry out real-time adjustment to the disposal ability of using system easily.
But, the cooperation of computer application system each processing unit is to rely on communications platform to support, therefore, though the operating type of above-mentioned class streamline has reduced the complexity of system, improved the retractility of system, but also increased the communication pressure between the each processing unit simultaneously, and then may influence the performance of whole system.So, improve the efficient of cooperating between each processing unit and will become important means that improve the computer application system overall performance.
, how to improve the problem that cooperation efficiency comprises following two aspects here: at first, the notice of how will cooperating sends to the purpose processing unit with the fastest speed; Secondly, how in the processing unit that can handle current task, to obtain the fastest purpose processing unit of processing speed.
At first problem of above-mentioned raising cooperation efficiency, can adopt the most commonly usually, also be that most basic socket (Socket) communication pattern is realized cooperate and notified the purpose that sends to the purpose processing unit with the fastest speed.Socket is high-speed traffic between the implementation process effectively, but directly between application Socket implementation process communication restriction the retractility of system, flexibility is relatively poor, and the software reuse ability is bad.For this reason, the network programming personnel are encapsulated in Socket in the communication unit as the basic module of communication usually, and in the interprocess communication process of reality, direct calling communication unit like this can be flexibly, simply communicate by letter between implementation process.
Second problem at above-mentioned raising cooperation efficiency; promptly solve problem how in the processing unit that can handle current task, to select the fastest processing unit of processing speed; generally can introduce load balancing mechanism; promptly before process communication, select the lightest processing unit of present load to handle by load balancing mechanism earlier.
Current generally is to adopt the form of layering to realize in conjunction with the solution aspect above-mentioned two effectively in Distributed Application, promptly takes out a transport layer and is responsible for communication between the management processing unit specially.The application module of whole system will be developed based on the communication support of this transport layer, and transport layer is upwards for application program provides the communication service interface, and application module is by this interface, and the transfer of data between the processing unit is finished in the service of using transport layer to provide.
Fig. 1 has shown that current a kind of typical computer application system realizes the abstract model of communicating by letter between the processing unit.As can be seen from Figure 1, each processing unit of this model can be abstracted into following which floor module: bottom module, transport layer and application layer.Wherein, the bottom module comprises operating system, middleware and various general module; The transmission link between the processing unit is in charge of and is safeguarded to transport layer, and use the load balancing algorithm of customization to select processing unit from Candidate Set; Application layer realizes service logic, and data communication is finished in the service of application transport layer.
In the running of system, application layer is submitted the transmission request to transport layer, the task of transport layer analysis request, application load are shared one of algorithm selection can handle this task handling unit, and finishes transmission course by the transmission link of setting up and use this processing unit.
As can be seen, in the process that processing unit shown in Figure 1 communicates, each time communication process all at first use the load balancing algorithm determined from can handle this task handling unit processing unit of selection, and then the transmission path that passes through to this processing unit is finished data communication.
So, the time that communication process spent just all will comprise each time: retrieval can be handled the purpose processing unit of this subtask, and is determined to the required time of transmission path of respective handling unit; And in this used time of transmission path transmitting data bag.Here retrieving and locate transmission link is the auxiliary processing unit cooperation, is not the time that the cooperation of distributed system processing unit must spend.
In addition, the characteristics of a mainframe computer applications system are various, therefore, use predetermined load balancing algorithm in transport layer and always can both effectively not play a role to all signal intelligences.For example, in the prior art, load balancing is normally selected in can handling the purpose processing unit of this subtask at random by transport layer, this stochastic selection algorithm shows as the effect of all processing unit load averages on the whole, i.e. the transport layer task that the mean allocation request is handled in all processing unit in Candidate Set.Though this load balancing algorithm is simple, be easy to realize, and can on average bring into play the disposal ability of each processing unit basically,, in the application of reality be have necessarily circumscribed.For example, when system dynamics adds certain new processing unit or upgrades certain processing unit, the disposal ability of this processing unit, comprise central processing unit and internal memory etc., all may bigger deviation be arranged with original processing unit, in this case, stochastic selection algorithm obviously can not be given full play to the disposal ability of the new adding or the processing unit of newly upgrading.Cause the processing unit of new adding or new upgrading can't be fast and the problem of whole system fusion.
Summary of the invention
In order to solve the problems of the technologies described above, the invention provides a kind of system of selection of transmission link, can farthest reduce the time of retrieval processing unit, realize Processing tasks distribution fast.
Further aim of the present invention is the more effective load that distributes between processing unit, improves the cooperation performance of each processing unit.
The system of selection of transmission link of the present invention, with load balancing algorithm process management separates and be arranged at respectively processing unit with load data application layer and transport layer, and between application layer and transport layer, transmitting the load balancing algorithm information by coffret, the distribution procedure of Processing tasks is finished in cooperation;
This method may further comprise the steps:
A, application layer detect whether there is the transmission link that can handle this task handling unit, if exist, then use this transmission link to finish the distribution procedure of this subtask, finish then; Otherwise, execution in step b;
B, application layer are selected suitable load balancing algorithm according to the characteristics of application program; And provide suitable transmission link by coffret request transport layer, execution in step c then;
C, transport layer judge whether to exist can handle this task handling unit, if exist, then the load balancing algorithm of selecting according to application layer is selected suitable processing unit from can handle this task handling unit, then, the transmission link that uses selected processing unit is finished the communication between the processing unit; Otherwise, carry out abnormality processing, finish then.
Described method further comprises: after transport layer was set up transmission link and returned transport layer link mark, application layer was created the application layer routing label, and was established to the association of this transmission link by intelligent pointer.
And when the transmission link state changed, transport layer was revised the transport layer link flag state of its preservation; The application layer routing label is known the variation of this transmission link state automatically by intelligent pointer.
Whether the described detection of step a exists the method for the transmission link that can handle this task handling unit to be specially: whether the intelligent pointer of application layer detection application layer routing label is effective.
The described load balancing algorithm of selecting to be fit to according to the characteristics of using of step b is specially: application program is according to using required processor ability, or required memory source, or required network bandwidth resources, or three's combination in any is selected suitable load balancing algorithm.
Step c further comprises: after described transport layer was selected the purpose processing unit, transport layer at first judged whether to exist the transmission link of selected processing unit, if exist, then used this transmission link to finish communication process between the processing unit; Otherwise transport layer is established to the transmission link of selected processing unit, and the transmission link of use foundation is finished the communication process between the processing unit then.
Described method further comprises: processing unit receives the load data of other processing units broadcasting, and classifies according to the load data of other processing units and to gather, and generates the processing unit candidate collection; Step c is described to select the purpose processing unit to be specially from handling this task handling unit: select the purpose processing unit during can handling this task handling unit candidate collection.
Described load data comprises the disposal ability of the manageable task type of processing unit, processing unit and the present load amount of processing unit.
The described load balancing algorithm of step c is: when the transport layer of pretreatment unit can be handled the lightest processing unit of search load in this task handling unit candidate collection at all.
Processing unit is carrying out before the described load balancing of step c the step-size in search value being set; And a search benchmark pointer is set in can handling this task handling unit candidate collection, point to the processing unit of search original position; The described load balancing algorithm of described step b may further comprise the steps:
B1, this processing unit in can handling this task handling unit candidate collection, from the processing unit of described search benchmark pointer indication, the loading condition of the processing unit of comparison search step value number, and therefrom select the lightest processing unit of load;
B2, described search benchmark pointer is moved forward half of described step-size in search value, point to the processing unit of search next time original position.
This shows, use transmission link system of selection of the present invention and can obtain following beneficial effect:
1, the present invention is by using the transmission link mark, when Processing tasks is distributed, the transmission link of selecting use application layer routing label to be preserved, and without the repeated retrieval processing unit, thereby reduced the time that transmission link is selected, realized that transmission link is selected fast, also avoided the secondary of transmission link to set up process simultaneously;
2, the present invention is by allowing application program realize the load balancing algorithm function, and realize the function that load data is managed in transport layer, can fully introduce various load balancing algorithms flexibly according to the characteristics of process, thereby the disposal ability of more effective performance each processing unit, improve the cooperation performance between the processing unit, and system's retractility of Distributed Application.
Description of drawings
Fig. 1 has shown the abstract model of communicating by letter between the at present typical computer application system processing unit;
Fig. 2 has shown that the described application system of a preferred embodiment of the invention realizes the abstract model of communicating by letter between the processing unit;
Fig. 3 is the flow chart of the described application layer communication method of a preferred embodiment of the invention;
Fig. 4 is the schematic diagram that concerns of a preferred embodiment of the invention described processing unit application layer routing label and transport layer link mark;
Fig. 5 is the described load balancing algorithm flow chart of a preferred embodiment of the invention.
Embodiment
The present invention is described in further detail below in conjunction with accompanying drawing and specific embodiment.
In order to realize that the distribution application system each processing unit better cooperates, a preferred embodiment of the present invention is improved existing load balancing processing method.Fig. 2 has shown that the described computer application system of this embodiment realizes the abstract model of communicating by letter between the processing unit.As shown in Figure 2, described embodiment is with load balancing algorithm process management separates and be arranged at respectively processing unit with load data application layer and transport layer, and between application layer and transport layer by coffret (TI, Transport Interface) transmit the load balancing algorithm information, the distribution procedure of Processing tasks is finished in cooperation.
Wherein, the load balancing algorithm processing module that is positioned at application layer can be selected suitable load balancing algorithm according to the characteristics of using self.For example, certain application is very high to the requirement of disposal ability and memory source, and the designed load balancing algorithm of application program just can be selected the processing unit that current disposal ability is the strongest, memory source is the most idle; And if certain process requires than higher bandwidth resources, the designed load balancing algorithm of application program just can be chosen the wideest processing unit of this processing unit transmission link bandwidth.The load balancing algorithm processing module of application layer even can consider along with the load balancing algorithm is upgraded in the variation of processing unit load state in the processing unit candidate collection, such as, if the load of processing unit candidate collection processing unit is generally lighter, just there is no need to select too accurate load share algorithm.
And the load data administration module that is positioned at transport layer is used to collect and manage the load data of other processing units, for example, the manageable task type of other processing units, the task number that is moving and maximum processing capability or the like, then the load data of other processing units is classified and gather, formation is at the processing unit candidate collection of all kinds of processing targets, and generates the loading index of other processing units.The reflection that each processing unit candidate collection is real-time current number of processing units and the state that can be the service of certain processing target.
Like this, by the described abstract model of Fig. 2, changed in the prior art owing to the load balancing function realizes all restrictions of bringing in transport layer, disposal ability that can more effective each processing unit of performance in distribution application system improves the cooperation performance between the processing unit.
In addition, in order further to reduce the required time of transmission link of setting up and choosing the purpose processing unit, realize link selection and communication process fast, the preferred embodiments of the present invention have been introduced the application layer routing label in the load balancing algorithm processing module of processing unit application layer, this application layer routing label is used to write down the transmission route of this processing unit to corresponding purpose processing unit, with as the subsequent transmission process foundation of gravity treatment transmission link whether.
Describe below in conjunction with Fig. 2, if processing unit 1 is set up and the transmission link of processing unit 2, then the load balancing algorithm processing module of processing unit 1 application layer will be preserved one to the application layer routing label of handling unit 2.And if processing unit 1 occurs unusually to the transmission link of handling unit 2, then the load balancing algorithm processing module of processing unit 1 application layer should discharge this to the application layer routing label of handling unit 2.Like this, in the communication process of processing unit next time 1 and processing unit 2, whether the application layer of processing unit 1 can at first detect application corresponding layer routing label effective, if this application layer routing label is effective, then processing unit 1 transmission link of having set up that just can directly be write down by this application layer routing label carries out data communication, share process and need not once more starting load, thereby saved search, select time of processing unit, realized that transmission link is selected fast.Thus, application program just can make full use of the transmission link of having set up and carry out transfer of data, and need all not retrieve and select processing unit in communication process each time, and then has realized the high-speed traffic between process.
Fig. 3 has shown the selection course of the described transmission link of a preferred embodiment of the invention.As shown in Figure 3, the selection course of transmission link of the present invention comprises the steps:
Step 301: application program sends the load balancing algorithm processing module that transmission asks to handle the unit application layer, comprises the task for the treatment of other processing unit processes in this transmission request;
Step 302: the load balancing algorithm processing module detect whether exist with can the corresponding application layer routing label in treatment step 301 described task handling unit, if having, then execution in step 306; Otherwise, then carry out the step 303;
Step 303: the load balancing algorithm that the load balancing algorithm processing module select to be fit to, the characteristic information that uses this subtask by coffret to transport layer request transmission link; It is to have corresponding processing unit candidate collection that transport layer is judged, if having, then carries out the step 304; Otherwise, carry out the step 307;
Step 304: transport layer is selected suitable processing unit according to the load balancing algorithm of selecting, and judges whether to exist the transmission link of selected processing unit, if exist, then carries out the step 305; Otherwise transport layer is established to the transmission link of selected processing unit, then, carries out the step 305;
Step 305: the load balancing algorithm processing module is preserved this application layer routing label, then, carries out the step 306;
Step 306: the transmission link that the load balancing algorithm processing module uses this application layer routing label to be write down is finished interprocess communication;
Step 307: application program is carried out abnormality processing, indicates this Processing tasks distribution procedure failure.
In order to realize method shown in Figure 3, must guarantee the state of transmission link between current processing unit of reflection that application layer routing label that the load balancing algorithm processing module of processing unit application layer preserved is real-time and other processing units.But, because the hierarchy of processing unit and the communication unit of system design encapsulation principle, though be in charge of and safeguard the state of the acquisition transmission link that the transport layer of transmission link between the processing unit can be real-time, transport layer does not have direct interface can notify the state of application layer transmission link to change.Like this, the application layer routing label of being preserved with regard to the load balancing algorithm processing module that produces application layer possibly and the actual inconsistent situation of Link State.Clearly, this will cause the failure of communication process.
This asks that a preferred embodiment of the present invention is when setting up transmission link in order to determine, and the transport layer of processing unit is also set up a transport layer link mark, writes down the transmission link of this processing unit to the respective handling unit.Fig. 4 is the schematic diagram of the preferred embodiment described processing unit application layer routing label and transport layer link mark relation.As shown in Figure 4, the transport layer link mark of processing unit is specially the connecting object of transport layer, and the application layer routing label is specially the routing object of application layer, these two same transmission links of the equal mark of object.Wherein, the routing object of application layer further comprises an intelligent pointer that points to the transport layer connector object, by this intelligent pointer, guarantee transport layer link mark and application layer routing label state consistency, realize the purpose of transport layer transmission link state to application layer implicit expression transparent transmission, if promptly this intelligent pointer is not empty, then this application layer routing label is an effective status; And if this intelligent pointer is empty, then this application layer routing label is a disarmed state.
When transport layer was set up transmission link, processing unit was set up a connector object, a routing object respectively in transport layer and application layer, and the intelligent pointer of the routing object of application layer points to corresponding transport layer connector object.These two objects all mark to the transmission link of alignment processing unit.Like this, under the normal situation of transmission link, application program can find the transmission link of institute's mark by the intelligent pointer of its application layer routing object, and finishes data communication by this transmission link.And when transmission link is unavailable, the operation that transport layer discharges the corresponding connectors object will cause the routing object of application layer to become disarmed state.The application layer routing label just can be by detecting the state of the real-time acquisition respective transmissions link of its intelligent pointer like this.
Thus, the operation that whether effectively can performing step 302 described detection application layer routing labels whether exists of the intelligent pointer of the routing object of load balancing algorithm processing module by detecting application layer.
In addition, in the step 304 of the method for the invention, the load balancing algorithm processing module can be selected the algorithm of sharing at random of the prior art, also can select multiple other load balancing algorithm, and the present invention does not limit the selection of load balancing algorithm.
A preferred embodiment of the present invention provides the method for a kind of load balancing algorithm and load data management, and the load balancing algorithm of this embodiment is suitable for using load balancing required precision condition with higher.In addition, the disposal ability of the load data management and utilization purpose processing unit of this embodiment and the load data of present load situation etc., the processing unit of actual loading and disposal ability ratio minimum, the i.e. the lightest processing unit of present load in the processing unit candidate collection.Wherein, the processing unit candidate collection is to be set up and safeguarded by the load data administration module of transport layer.
In order to obtain the loading condition of other processing units in real time, disposal ability and manageable task type, these are worth the load data administration module of other processing units the broadcasting in each processing unit meeting cycle, the cycle of broadcasting is short more, the data of transport layer link label record are just accurate more, and then the load balancing algorithm is just effective more.But on the other hand, the state information amount that the shortening of broadcast cycle will cause transmitting between the processing unit strengthens, and causes the load of network to become heavy, reduces the cooperative system performance.This broadcast cycle is generally set according to using desired load information accuracy.
From above-mentioned load balancing algorithm and load data management method as can be seen, the described load data management of present embodiment can't limit the size of processing unit candidate collection, therefore, can constantly add new processing unit in system.In addition, because the processing unit present load of the processing unit of new adding system or new upgrading is zero, can in a period of time, become the main target that Processing tasks distributes, till the load of the load of this processing unit and other processing units maintains an equal level gradually, thus present embodiment can guarantee new processing unit quick, seamless incorporate whole distributed system.
But the time complexity of searching for the lightest purpose processing unit of load in entire process unit candidate collection is linear, will occur the problem that algorithm is degenerated like this along with the expansion of distribution application system scale.Therefore, in another embodiment of the present invention, its load balancing algorithm has adopted the algorithm of suboptimum search, be that searching algorithm is no longer searched in the scope of entire process unit Candidate Set, but according to the experience step-size in search value Step of prior agreement, the processing unit of step-length Step number is specified in each only search, and finds out the purpose processing unit of a processing unit that load is the lightest as the distribution of this time Processing tasks in this step-size in search.This load balancing algorithm also is fit to current task distributing processing unit load condition with higher, so just can as early as possible the task branch be sent out, and reduces disposal ability and takies.
Fig. 5 is the described load balancing algorithm flow chart of the preferred embodiment.As shown in Figure 5, load balancing algorithm of the present invention is in can handling this task handling unit candidate collection, a search benchmark pointer is set, indicate the original position of this search, and by the selected chance of each processing unit equalization in the mobile realization Candidate Set of this pointer of control; Algorithm has been set an experience step-size in search value Step in addition, and this value is changed to 5 in the present embodiment; Load balancing algorithm of the present invention may further comprise the steps:
Step 501: the processing unit from search benchmark pointer indication, compare the loading condition of Step processing unit of step value, and therefrom select the processing unit of a present load and disposal ability ratio minimum, the i.e. the lightest processing unit of load;
Step 502: will search for half that the benchmark pointer moves forward step value, promptly mobile Step/2 processing unit is as the initial search point of searching for next time;
Step 503: be established to the transmission link of this processing unit, realize transfer of data.
In this embodiment, by load balancing search procedure repeatedly, search benchmark pointer can circulate in the candidate collection of processing unit, make that all processing units can both participate in load balancing in the processing unit candidate collection, thereby guaranteed the fairness of load balancing algorithm, the basic realization given full play to all processing unit processes abilities, realizes the purpose of effectively cooperating between each processing unit.
In addition, described load balancing algorithm is to select the lightest processing unit of load between a predefined Step processing unit, rather than in entire process unit candidate collection, retrieve, like this, just make the time of retrieval significantly reduce, and then improved the efficient of application program search purpose processing unit.
Need to prove, above-mentioned processing unit processes ability is virtual disposal ability, this disposal ability may be the nominal index of system, it also may be a configuration parameter of system, even might be to realize on the hardware system under the situation of a plurality of processing units, when the actual treatment ability of this hardware system is assigned to above-mentioned a plurality of processing unit in some way, the disposal ability that each processing unit disposed.
As can be seen from the above-described embodiment, method of the present invention is by moving on to load balancing algorithm process function the application layer of processing unit, and increase the function of load data management in transport layer, can fully introduce various load balancing algorithms flexibly according to the characteristics of using, thereby the more effective disposal ability of bringing into play processing unit has improved the cooperation performance between the processing unit.
In addition, the method of the invention is by setting up the method for mark for transmission link, can in transfer of data, use the transmission link that is kept at the application layer routing label, and without quadratic search purpose processing unit, thereby reduced the time of selecting transmission link, realized cooperating between processing unit fast and effectively.
More than lift preferred embodiment; the purpose, technical solutions and advantages of the present invention have been carried out further detailed description; institute is understood that; the above is the preferred embodiments of the present invention; not in order to show the present invention; within the spirit and principles in the present invention all, any modification of being done, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (10)
1. the method selected of a transmission link, it is characterized in that, this method is with load balancing algorithm process management separates and be arranged at respectively processing unit with load data application layer and transport layer, and between application layer and transport layer, transmitting the load balancing algorithm information by coffret, the distribution procedure of Processing tasks is finished in cooperation;
This method may further comprise the steps:
A, application layer detect whether there is the transmission link that can handle this task handling unit, if exist, then use this transmission link to finish the distribution procedure of this subtask, finish then; Otherwise, execution in step b;
B, application layer are selected suitable load balancing algorithm according to the characteristics of using; And provide suitable transmission link by coffret request transport layer, execution in step c then;
C, transport layer judge whether to exist can handle this task handling unit, if exist, then the load balancing algorithm of selecting according to application layer is selected suitable processing unit from can handle this task handling unit, then, the transmission link that uses selected processing unit is finished the communication between the processing unit; Otherwise, carry out abnormality processing, finish then.
2. the method for claim 1, it is characterized in that, described method further comprises: after transport layer was set up transmission link and returned transport layer link mark, application layer was created the application layer routing label, and was established to the association of this transmission link by intelligent pointer.
3. method as claimed in claim 2 is characterized in that, when the transmission link state changed, transport layer on-the-fly modified its transport layer link flag state of preserving; The application layer routing label is known the variation of this transmission link state automatically by intelligent pointer.
4. as claim 1,2 or 3 described methods, it is characterized in that whether the described detection of step a exists the method for the transmission link that can handle this task handling unit to be specially: whether the intelligent pointer of application layer detection application layer routing label is effective.
5. the method for claim 1, it is characterized in that, the described load balancing algorithm of selecting to be fit to according to the characteristics of using of step b is specially: application program is according to the processor ability of this required by task, or required memory source, or required network bandwidth resources, or three's combination in any is selected suitable load balancing algorithm.
6. the method for claim 1, it is characterized in that, step c further comprises: after described transport layer is selected the purpose processing unit, transport layer at first judges whether to exist the transmission link of selected processing unit, if exist, then use this transmission link to finish communication process between the processing unit; Otherwise transport layer is established to the transmission link of selected processing unit, and the transmission link of use foundation is finished the communication process between the processing unit then.
7. the method for claim 1 is characterized in that, described method further comprises: processing unit receives the load data of other processing units broadcasting, and classifies according to the load data of other processing units and to gather, and generates the processing unit candidate collection;
Step c is described to select the purpose processing unit to be specially from handling this task handling unit: select the purpose processing unit during can handling this task handling unit candidate collection.
8. method as claimed in claim 7 is characterized in that, described load data comprises the disposal ability of the manageable task type of processing unit, processing unit and the present load amount of processing unit.
9. method as claimed in claim 7 is characterized in that, the described load balancing algorithm of step c is: when the transport layer of pretreatment unit can be handled the lightest processing unit of search load in this task handling unit candidate collection at all.
10. method as claimed in claim 7 is characterized in that, processing unit was provided with the step value of search before carrying out the described load balancing of step c; And a search benchmark pointer is set in can handling this task handling unit candidate collection, point to the processing unit that is positioned at the search original position;
The described load balancing algorithm of described step b may further comprise the steps:
B1, this processing unit in can handling this task handling unit candidate collection, from the processing unit of described search benchmark pointer indication, the loading condition of the processing unit of comparison search step value number, and therefrom select the lightest processing unit of load;
B2, described search benchmark pointer is moved forward half of described step-size in search value, as the initial search point of search next time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200410064285.1A CN1744593B (en) | 2004-08-30 | 2004-08-30 | Transmission link selecting method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200410064285.1A CN1744593B (en) | 2004-08-30 | 2004-08-30 | Transmission link selecting method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1744593A CN1744593A (en) | 2006-03-08 |
| CN1744593B true CN1744593B (en) | 2010-04-28 |
Family
ID=36139787
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200410064285.1A Expired - Fee Related CN1744593B (en) | 2004-08-30 | 2004-08-30 | Transmission link selecting method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1744593B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011141580A (en) * | 2010-01-05 | 2011-07-21 | Sony Corp | Access control apparatus, data processor, access control method and program |
| US8898505B2 (en) | 2011-12-01 | 2014-11-25 | International Business Machines Corporation | Dynamically configureable placement engine |
| US10554782B2 (en) | 2011-12-01 | 2020-02-04 | International Business Machines Corporation | Agile hostpool allocator |
| US8874751B2 (en) | 2011-12-01 | 2014-10-28 | International Business Machines Corporation | Candidate set solver with user advice |
| CN103023815B (en) * | 2012-12-26 | 2015-05-13 | 杭州华三通信技术有限公司 | Aggregation link load sharing method and device |
| CN115037612B (en) * | 2022-03-10 | 2024-03-26 | 广东统信软件有限公司 | Server configuration method, computing device and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1030484A2 (en) * | 1999-01-29 | 2000-08-23 | Nortel Networks Corporation | Data link layer quality of service for UMTS |
| CN1504036A (en) * | 2001-03-12 | 2004-06-09 | �����ɷ� | Method and apparatus for providing multiple quality of service levels in wireless packet data services connection |
-
2004
- 2004-08-30 CN CN200410064285.1A patent/CN1744593B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1030484A2 (en) * | 1999-01-29 | 2000-08-23 | Nortel Networks Corporation | Data link layer quality of service for UMTS |
| CN1504036A (en) * | 2001-03-12 | 2004-06-09 | �����ɷ� | Method and apparatus for providing multiple quality of service levels in wireless packet data services connection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1744593A (en) | 2006-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7370029B2 (en) | Method of changing system configuration in shared-nothing database management system | |
| EP1032175B1 (en) | System and method for transferring partitioned data sets over multiple threads | |
| US5819083A (en) | Minimal sufficient buffer space for data redistribution in a parallel database system | |
| CN109933631A (en) | Distributed parallel database system and data processing method based on Infiniband network | |
| CN104679594B (en) | A kind of middleware distributed computing method | |
| CN101142552A (en) | Resource allocation in computing systems | |
| CN102314506B (en) | Based on the distributed buffering district management method of dynamic index | |
| CN112801565B (en) | Goods allocation dispatching method, system and storage medium for intelligent storage | |
| CN103080903A (en) | Scheduler, multi-core processor system, and scheduling method | |
| CN110162388A (en) | A kind of method for scheduling task, system and terminal device | |
| KR20170097132A (en) | System for efficient processing of transaction requests related to an account in a database | |
| CN113722127A (en) | Efficient lightweight easy-to-use distributed network message middleware | |
| CN107977763A (en) | Resource allocation method and related device | |
| CN116777182B (en) | Task dispatch method for semiconductor wafer manufacturing | |
| CN108900626A (en) | Date storage method, apparatus and system under a kind of cloud environment | |
| CN108073696A (en) | GIS application processes based on distributed memory database | |
| CN111427678A (en) | Virtualized resource scheduling system and method in automobile diagnosis cloud platform | |
| US20050188191A1 (en) | Resource discovery method and cluster manager apparatus | |
| CN106502918A (en) | A kind of scheduling memory method and device | |
| CN106960011A (en) | Metadata of distributed type file system management system and method | |
| CN1744593B (en) | Transmission link selecting method | |
| CN101236570A (en) | Method and system for coordinating access to locally and remotely exported file systems | |
| CN111556123A (en) | Self-adaptive network rapid configuration and load balancing system based on edge calculation | |
| CN105282045B (en) | A kind of distributed computing and storage method based on consistency hash algorithm | |
| CN102510403A (en) | Cluster distributed system for vehicle data receiving and real-time analysis and method therefor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| ASS | Succession or assignment of patent right |
Owner name: CHINA POTEVIO CO., LTD. Free format text: FORMER OWNER: PUTIAN IT TECH INST CO., LTD. Effective date: 20130924 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| C56 | Change in the name or address of the patentee |
Owner name: PUTIAN IT TECH INST CO., LTD. Free format text: FORMER NAME: PUTIAN INST. OF INFORMATION TECHNOLOGY |
|
| CP03 | Change of name, title or address |
Address after: 100080 Beijing, Haidian, North Street, No. two, No. 6, No. Patentee after: PETEVIO INSTITUTE OF TECHNOLOGY Co.,Ltd. Address before: 100085, No. two, 2 street, base of information industry, Beijing Patentee before: POTEVIO Institute of Information Technology |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20130924 Address after: 100080, No. two, 2 street, Zhongguancun science and Technology Park, Beijing, Haidian District Patentee after: CHINA POTEVIO CO.,LTD. Address before: 100080 Beijing, Haidian, North Street, No. two, No. 6, No. Patentee before: PETEVIO INSTITUTE OF TECHNOLOGY Co.,Ltd. |
|
| ASS | Succession or assignment of patent right |
Owner name: PUTIAN IT TECH INST CO., LTD. Free format text: FORMER OWNER: CHINA POTEVIO CO., LTD. Effective date: 20131205 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20131205 Address after: 100080 Beijing, Haidian, North Street, No. two, No. 6, No. Patentee after: PETEVIO INSTITUTE OF TECHNOLOGY Co.,Ltd. Address before: 100080, No. two, 2 street, Zhongguancun science and Technology Park, Beijing, Haidian District Patentee before: CHINA POTEVIO CO.,LTD. |
|
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100428 Termination date: 20210830 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |