CN102004673A - Processing method and system of multi-core processor load balancing - Google Patents
Processing method and system of multi-core processor load balancing Download PDFInfo
- Publication number
- CN102004673A CN102004673A CN2010105642054A CN201010564205A CN102004673A CN 102004673 A CN102004673 A CN 102004673A CN 2010105642054 A CN2010105642054 A CN 2010105642054A CN 201010564205 A CN201010564205 A CN 201010564205A CN 102004673 A CN102004673 A CN 102004673A
- Authority
- CN
- China
- Prior art keywords
- hash
- polycaryon processor
- factor
- nuclear
- identification information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a processing method and a system of multi-core processor load balancing. The method comprises the following steps: a multi-core processor obtains the identification information of an input message; the identification information and a selected HASH factor are subject to the HASH operation by the multi-core processor to obtain the HASH value corresponding to the identification information, and transmits the output message to the core corresponding to the HASH value; and the selected HASH factor is adjusted according to the load condition of each core in the multi-core processor. The method and system of the invention can enhance the flexibility and the performance of the multi-core processor, and enhance the user experience.
Description
Technical field
The present invention relates to the communications field, relate in particular to a kind of disposal route and system of polycaryon processor load balancing.
Background technology
Polycaryon processor has obtained widely using in various fields such as network communication equipments, how to allow the processing power of each nuclear of polycaryon processor obtain having fully played problem into a key.If certain nuclear load is too much, then it becomes the bottleneck of total system probably, causes the total system performance decrease.In communication facilities, for message sequence that guarantees each data stream inside and the information of preserving data stream, the same nuclear that usually needs same data flow data to be distributed to polycaryon processor is handled, because the dispersion degree of the quantity of stream, differentiating method, traffic identifier may be very different in different occasions, so the method for load balancing and rule just have multiple choices.
At present, the method for polycaryon processor load balancing is normally by to the intellectuality of distributor and complicated the realization.For example, in correlation technique, provide a kind of and examined the method that loading condition carries out load sharing, need in distributor, write down the numbering of the nuclear of every stream distribution last time, so that the message of same stream mails to same nuclear according to each.And dynamically adjust in the method for corresponding relation of HASH space and nuclear at another kind, when certain nuclear load is light, need enlarge the HASH space that it is responsible for, make it might be assigned to more task, and the HASH space dividing need the distributor memory and dynamically adjust.Because the distribution of polycaryon processor generally need be in the inner realization of processor, and be subjected to the restriction of processor hardware scale, distributor can not be done too complicatedly, so, in above-mentioned correlation technique, distribution method is difficult to utilize the distributor of polycaryon processor itself to realize, often needs to utilize several nuclears of polycaryon processor to use the mode of software programming to realize.Like this, distributor has consumed polycaryon processor when realizing distribution function resource, and distribution nuclear itself becomes bottleneck of performance probably.
And, in correlation technique, network communication apparatus polycaryon processor load-balancing method is under normally the fabricator was just solidified in the development phase, the employed HASH hash factor is that the developer set before dispatching from the factory according to the experience of self, to the user all is sightless, and this mode can't satisfy all application scenarioss.
Summary of the invention
Fundamental purpose of the present invention is to provide a kind of processing scheme of polycaryon processor load balancing, with the method complexity height that solves polycaryon processor load balancing in the above-mentioned correlation technique at least and the problem of bad adaptability.
To achieve these goals, according to an aspect of the present invention, provide a kind of disposal route of polycaryon processor load balancing.
Disposal route according to polycaryon processor load balancing of the present invention may further comprise the steps: polycaryon processor extracts the identification information of incoming message; Polycaryon processor through the HASH computing, obtains the HASH value corresponding with this identification information with the HASH hash factor of this identification information and selection, and incoming message sent in the nuclear corresponding with the HASH value handles; And the HASH hash factor of selecting is adjusted according to the loading condition of each nuclear in the polycaryon processor.
Further, polycaryon processor is with the HASH hash factor process HASH computing of identification information and selection, obtain before the HASH value corresponding with this identification information, this method also comprises: polycaryon processor carries out the HASH computing respectively with this identification information and all HASH hash factors that are used to calculate, and obtains the distribution situation of all possible HASH hash space in polycaryon processor each nuclear corresponding with this identification information; And select the best HASH hash factor of dispersion degree as the HASH hash factor of selecting according to distribution situation.
Further, select the best HASH hash factor of dispersion degree to comprise as the HASH hash factor of selecting according to distribution situation: polycaryon processor calculates the HASH result's who answers with all HASH hash factor pairs that are used to calculate variance respectively according to distribution situation, and from wherein selecting the HASH hash factor corresponding with the minimum variance value as the best HASH hash factor of dispersion degree.
Further, before according to the loading condition of each nuclear in the polycaryon processor HASH hash factor of selecting being adjusted, this method also comprises: collect the loading condition of each nuclear in the polycaryon processor according to scheduled time slot, and present to the user with the form of table and/or figure.
Further, above-mentioned incoming message is the model of network traffics.
Further, above-mentioned identification information is the feature field of message, comprise following one of at least: IP address, protocol type.
Further, in the polycaryon processor loading condition of each nuclear comprise following one of at least: the message total of cpu busy percentage, user's linking number, processing or forwarding.
To achieve these goals, according to a further aspect in the invention, also provide a kind of disposal system of polycaryon processor load balancing.
According to the disposal system of polycaryon processor load balancing of the present invention, comprise polycaryon processor device and HASH hash factor adjusting gear, wherein, the polycaryon processor device comprises: extraction module is used to extract the identification information of incoming message; The HASH computing module is used for the HASH hash factor process HASH computing with this identification information and selection, obtains the HASH value corresponding with this identification information; And distribution module, be used for and incoming message sent to the nuclear corresponding with the HASH value handling; HASH hash factor adjusting gear comprises, adjusting module is used for according to the loading condition of each nuclear of polycaryon processor device the HASH hash factor of selecting being adjusted.
Further, the polycaryon processor device also comprises: the HASH hash factor is selected module, be used for this identification information and all HASH hash factors that are used to calculate are carried out the HASH computing respectively, obtain the distribution situation of this all possible HASH hash space in polycaryon processor each nuclear corresponding with identification information; And select the best HASH hash factor of dispersion degree as the HASH hash factor of selecting according to distribution situation.
Further, HASH hash factor adjusting gear also comprises: the load information acquisition module is used for the loading condition according to each nuclear of scheduled time slot collection polycaryon processor device; And display module, the loading condition of each nuclear in the polycaryon processor device of load information acquisition module collection is presented to the user with the form of table and/or figure.
By the present invention, the mode that employing is adjusted the HASH hash factor according to the loading condition of its each nuclear, the method complexity height of polycaryon processor load balancing and the problem of bad adaptability in the correlation technique have been solved, strengthen the dirigibility and the performance of multi-nuclear processor equipment, improved user experience.
Description of drawings
Accompanying drawing described herein is used to provide further understanding of the present invention, constitutes the application's a part, and illustrative examples of the present invention and explanation thereof are used to explain the present invention, do not constitute improper qualification of the present invention.In the accompanying drawings:
Fig. 1 is the process flow diagram according to the disposal route of the polycaryon processor load balancing of the embodiment of the invention;
Fig. 2 is the structured flowchart according to the disposal system of the polycaryon processor load balancing of the embodiment of the invention;
Fig. 3 is the structured flowchart of the disposal system of polycaryon processor load balancing according to the preferred embodiment of the invention;
Fig. 4 is the schematic flow sheet according to the message distribution of the preferred embodiment of the present invention two;
Fig. 5 is the schematic flow sheet according to the HASH factor preliminary election of the preferred embodiment of the present invention three;
Fig. 6 is the result of calculation synoptic diagram according to the HASH factor preliminary election of the preferred embodiment of the present invention three;
Fig. 7 is the schematic flow sheet of dynamically adjusting according to the HASH factor of the preferred embodiment of the present invention four;
Fig. 8 is according to the cpu load situation of the preferred embodiment of the present invention four synoptic diagram that presents to the user; And
Fig. 9 is according to another cpu load situation of the preferred embodiment of the present invention four synoptic diagram that presents to the user.
Embodiment
Hereinafter will describe the present invention with reference to the accompanying drawings and in conjunction with the embodiments in detail.Need to prove that under the situation of not conflicting, embodiment and the feature among the embodiment among the application can make up mutually.
According to the embodiment of the invention, provide a kind of disposal route of polycaryon processor load balancing.Fig. 1 is the process flow diagram according to the disposal route of the polycaryon processor load balancing of the embodiment of the invention, and as shown in Figure 1, this method may further comprise the steps:
Step S102, polycaryon processor extracts the identification information of incoming message;
Step S104, polycaryon processor through the HASH computing, obtains the HASH value corresponding with this identification information with the HASH hash factor of this identification information and selection, and incoming message sent in the nuclear corresponding with the HASH value handles; And
Step S106 adjusts the HASH hash factor of selecting according to the loading condition of each nuclear in the polycaryon processor.
Pass through above-mentioned steps, the mode that employing is adjusted the HASH hash factor according to the loading condition of its each nuclear, the method complexity height of polycaryon processor load balancing and the problem of bad adaptability in the correlation technique have been solved, strengthen the dirigibility and the performance of multi-nuclear processor equipment, improved user experience.
Need to prove, the HASH algorithm is that message flow is identified, for example, the traffic identifier that the source IP address of message, purpose IP address, protocol type, source port number, destination slogan are formed, through a HASH computing, transform to the HASH space,, the message flow of input is divided as far as possible equably less than in the HASH space through after the conversion.Wherein, the HASH effect of uniform is determined by the HASH factor fully.
Preferably, before step S104, polycaryon processor can carry out the HASH computing respectively with this identification information and all HASH hash factors that are used to calculate, and obtains the distribution situation of all possible HASH hash space in polycaryon processor each nuclear corresponding with this identification information; And select the best HASH hash factor of dispersion degree as the HASH hash factor of selecting according to distribution situation.This method selects dispersion degree best from all available HASH hash factors, can to send in the polycaryon processor load distribution of each nuclear after the HASH computing more even so that use, thus reach the purpose of load balancing, improved networks efficiency.
Preferably, select the best HASH hash factor of dispersion degree to comprise as the HASH hash factor of selecting according to distribution situation: polycaryon processor calculates the HASH result's who answers with all HASH hash factor pairs that are used to calculate variance respectively according to distribution situation, and from wherein selecting the HASH hash factor corresponding with the minimum variance value as the best HASH hash factor of dispersion degree.This method is simple and practical, workable.
For example, the selection of the HASH hash factor also can be the distinctive HASH hash factor that the user is provided with voluntarily.
Preferably, before step S106, can collect the loading condition of each nuclear in the polycaryon processor, and present to the user with the form of table and/or figure according to scheduled time slot.Like this, each examines the loading condition on border in the polycaryon processor so that the user can observe, thereby adjusts, and has improved the validity and the practicality of system.
In specific implementation process, the identification information that polycaryon processor can reconfigure the new HASH hash factor and incoming message according to the loading condition of its each nuclear carries out the HASH computing, and this incoming message sent in the nuclear corresponding with the HASH result of the new HASH hash factor handles.
Preferably, incoming message can be the model of network traffics.This method has improved the dirigibility and the adaptive faculty of system.In specific implementation process, can be the Model of network traffic of the network equipment connection of precognition, for example,, can import inner information such as IP address network segment for the router of enterprise, for gateway, can import information such as user's IP address scope.
Preferably, the above-mentioned identification information of incoming message can be the feature field of message, comprise following one of at least: IP address, protocol type.This method is simple and practical, workable.
Preferably, in the polycaryon processor loading condition of each nuclear can comprise following one of at least: the message total of cpu busy percentage, user's linking number, processing or forwarding.This method is simple and practical, workable.
According to the embodiment of the invention, also provide a kind of disposal system of polycaryon processor load balancing.Fig. 2 is the structured flowchart according to the disposal system of the polycaryon processor load balancing of the embodiment of the invention, as shown in Figure 2, this system 20 comprises polycaryon processor device 22 and HASH hash factor adjusting gear 24, wherein, polycaryon processor device 22 comprises: extraction module 222 is used to extract the identification information of incoming message; HASH computing module 224 is coupled to extraction module 222, is used for the HASH hash factor process HASH computing with this identification information and selection, obtains the HASH value corresponding with this identification information; And distribution module 226, be coupled to HASH computing module 224, be used for and incoming message sent to the nuclear corresponding with the HASH value handling; HASH hash factor adjusting gear 24 comprises that adjusting module 242 is coupled to HASH computing module 224, is used for according to the loading condition of polycaryon processor device 22 each nuclears the HASH hash factor of selecting being adjusted.
Pass through said system, the mode that the loading condition of each nuclear is adjusted the HASH hash factor in the employing HASH hash factor adjusting gear 24 polycaryon processor devices 22, the method complexity height of polycaryon processor load balancing and the problem of bad adaptability in the correlation technique have been solved, strengthen the dirigibility and the performance of multi-nuclear processor equipment, improved user experience.
Fig. 3 is the structured flowchart of the disposal system of polycaryon processor load balancing according to the preferred embodiment of the invention, as shown in Figure 3, polycaryon processor device 22 also comprises: the HASH hash factor is selected module 228, be coupled to HASH computing module 224, be used for this identification information and all HASH hash factors that are used to calculate are carried out the HASH computing respectively, obtain the distribution situation of all possible HASH hash space in polycaryon processor each nuclear corresponding with this identification information; And select the best HASH hash factor of dispersion degree as the HASH hash factor of selecting according to distribution situation.
Preferably, HASH hash factor adjusting gear 24 also comprises: load information acquisition module 244, be coupled to polycaryon processor device 22, and be used for collecting the loading condition of polycaryon processor device 22 each nuclears according to scheduled time slot; And display module 246, be coupled to load information acquisition module 244, the loading condition of each nuclear in the polycaryon processor device 22 of load information acquisition module 244 collections is presented to the user with the form of table and/or figure.
Need to prove that the disposal system of the polycaryon processor load balancing of describing among the device embodiment is corresponding to above-mentioned method embodiment, its concrete implementation procedure had been carried out detailed description in method embodiment, do not repeat them here.
Be elaborated below in conjunction with the implementation procedure of preferred embodiments and drawings to the foregoing description.
Preferred embodiment one
Present embodiment is in the network communication equipment that uses polycaryon processor to realize, increased the configuration of the load balancing HASH hash factor, adopt and use HASH ways of distribution simple and that be easy to realize, make and use the equipment of polycaryon processor exploitation to reach effect of load balance.
Particularly, polycaryon processor meets the dynamic debugging system of sharing and can comprise in the present embodiment: Operation ﹠ Maintenance System, polycaryon processor parts and loading condition collector, wherein, the polycaryon processor parts comprise distributor, and distributor can comprise message characteristic extraction apparatus, HASH unit, the HASH factor and message Dispatching Unit.Below each module is described in detail.
Operation ﹠ Maintenance System is used for software and interface that the user operates equipment.
The polycaryon processor parts are used for using the network communication equipment realization access of polycaryon processor exploitation, the parts of processing capacity.Wherein, distributor is used for being responsible for message distribution to each nuclear in polycaryon processor inside; The message characteristic extraction apparatus is used for extracting from message content the feature field (for example, IP address, protocol type etc., these feature field are as the sign of message flow) of message; The HASH unit is used for according to the feature field of message and the HASH factor as input, and the HASH result who calculates message is as the distribution foundation; The HASH factor is used for the polynomial expression divisor as the HASH computing; The message Dispatching Unit is used for according to message HASH result the communication mechanism of message by processor inside sent to the nuclear of being responsible for handling this message.
The loading condition collector is used to run on the program on the polycaryon processor, the cpu busy percentage of each nuclear of record polycaryon processor, and information such as the message flow process of processing feed back to Operation ﹠ Maintenance System.
In specific implementation process, adopt the message distribution method of polycaryon processor HASH distribution to be: at first, feature extraction, that is, the message characteristic extraction unit extracts the traffic identifier of message.Secondly, the HASH mapping promptly, as input, is calculated the HASH value of message characteristic field with message characteristic field and the HASH factor; Once more, the message distribution promptly, according to HASH value and the responsible HASH spatial dimension of each nuclear, sends to message the nuclear of appointment.
In the preferred embodiment of the present invention, the method that the polycaryon processor load sharing is dynamically adjusted can may further comprise the steps:
Step 1, HASH factor preliminary election function.The user imports known message input range, by the HASH value of all possible message of system-computed, determines the dispersion degree of HASH space distribution, therefrom selects the best HASH factor of dispersion degree.
Step 2, the load information collection.Polycaryon processor is regularly collected the loading condition of its each nuclear, sends to Operation ﹠ Maintenance System.
Step 3, load information presents.In the user interface of Operation ﹠ Maintenance System, use the mode of form and figure to present to the user multinuclear loading condition in a period of time.
Step 4, the HASH factor is reset.The user reselects the HASH hash factor, and the own distinctive HASH hash factor perhaps is set.
As seen, the preferred embodiment of the present invention in multi-nuclear processor equipment, introduced the polycaryon processor loading condition collect and present, configurable two functions of the HASH hash factor.Wherein, the collection of polycaryon processor loading condition is used for the load state that the user observes polycaryon processor, and the configuration of the HASH hash factor makes the user adjust the effect of HASH hash, makes the bigger performance of multi-nuclear processor equipment performance.
Preferred embodiment two
Fig. 4 is that the polycaryon processor load sharing system in this preferred embodiment comprises according to the schematic flow sheet of the message distribution of the preferred embodiment of the present invention two:
Operation ﹠ Maintenance System is used to run on the cover of one on PC or server software, is connected with equipment by network, and equipment is configured, monitors and manages.
Multi-nuclear processor equipment (using the communication network device of polycaryon processor exploitation) is used for the forwarding or the processing of data message.
Distributor is used for being responsible for message distribution to each nuclear in polycaryon processor inside.
The message characteristic extraction apparatus is used for from the feature field of message content extraction message.For example, IP address, protocol type or the like, these feature field are as the sign of message flow.
The HASH unit is used for according to the feature field of message and the HASH factor as input, and the HASH result who calculates message is as the distribution foundation.
The HASH factor is used for the polynomial expression divisor as the HASH computing.
The message Dispatching Unit is used for according to message HASH result, with the communication mechanism of message by processor inside, sends to the nuclear of being responsible for handling this message.
The loading condition collector is used to run on the program on the polycaryon processor, the cpu busy percentage of each nuclear of record polycaryon processor, and information such as the message flow process of processing feed back to Operation ﹠ Maintenance System.
As shown in Figure 4, the message ways of distribution of the polycaryon processor load-balancing method of present embodiment can comprise following step:
Step S402, feature extraction.Message at first by the message characteristic extraction apparatus, can identify the feature field of a stream after being received by polycaryon processor from the network during it extracts from content of message.For example, for IP stream, can extract source IP address, purpose IP address, next layer protocol; For TCP stream, can extract source IP address, purpose IP address, source tcp port number, purpose tcp port number.These feature field are combined, stream of sign that can be unique.
Step S404, the HASH mapping.The message characteristic extraction apparatus is input to the feature field KEY of message in the HASH unit, and the HASH unit uses the KEY and the HASH factor that presets to carry out the HASH computing, and its result is exactly the HASH value of message characteristic field correspondence.
Step S406, the message distribution.According to step S404 result calculated HASH value and the responsible HASH spatial dimension of each nuclear, message is sent to the nuclear of appointment.For example, a nuclear of HASH result's the respectively corresponding polycaryon processor of low 4Bit like this, as long as determined HASH result, just can know that sending it to which nuclear handles.
As seen, above-mentioned message distribution method is fairly simple, is easy to realize with hardware, and the same characteristic features field can be distributed on the same nuclear through distributor, and distributor can not possess memory function, so the efficient of the distribution of distributor is higher, and realizes that cost is low.
Preferred embodiment three
Present embodiment provides a kind of and has reached the method for multinuclear load balancing by dynamically adjusting the HASH factor, comprises the initial setting up of the HASH factor and on-the-fly modifies two processes.Fig. 5 is that as shown in Figure 5, the initial setting up of the HASH factor comprises following step according to the schematic flow sheet of the HASH factor preliminary election of the preferred embodiment of the present invention three:
Step S502, the setting of message input model.Under many circumstances, the Model of network traffic that certain network equipment connects might be known, such as, the router of an enterprise, it is known certainly that the IP address network segment of enterprises distributes.For another example, the GGSN gateway of GSM network, it is responsible for the distribution of cellphone subscriber's IP address, so the user's IP address scope can be predicted.In these cases, the model of user's fan-in network flow will extract the field and the distribution (for example, comprising which network segment) thereof of stream.
Step S504, the assessment of HASH factorial effect.In specific implementation process, the HASH computing is carried out in all HASH factors and these inputs that system can be known, obtains it at HASH spatial distributions state.For example, Fig. 6 is the result of calculation synoptic diagram according to the HASH factor preliminary election of the preferred embodiment of the present invention three, as shown in Figure 6, after calculating each HASH factor of use, whole sample space can be formed form in the distribution situation of each nuclear, the user can determine recently whether this multiple core processing system reach load balancing according to the load flow percentage of each CPU, and from the data of Fig. 6 as seen, the HASH3 factor is that wherein load balancing is best.
Step S506, the selection of the HASH factor.For example, select the wherein best HASH factor of HASH spatial spreading degree; Perhaps the user can carry out the selection of the HASH factor as required.And the user can add the HASH factor voluntarily, participates in the selection of the HASH factor.For example, among Fig. 6, HASH3 result's variance minimum, it is the most even to prove that whole sample space distributes on each nuclear.
This preferred embodiment is in equipment configuration parameter, increase the configuration of the polycaryon processor load balancing HASH hash factor, make the user to select the appropriate H ASH factor according to the situation of incoming message in advance, and under the discovery multinuclear is loaded uneven situation, can also adjust the HASH hash factor, make the load between each nuclear balanced more.
Preferred embodiment four
Fig. 7 is the schematic flow sheet of dynamically adjusting according to the HASH factor of the preferred embodiment of the present invention four, and as shown in Figure 7, the method for dynamically adjusting the HASH factor in operational process can may further comprise the steps:
Step S702, the collection of loading condition.Multi-nuclear processor equipment is regularly collected the load state of each nuclear.In specific implementation process, can comprise message total of cpu busy percentage, processing or forwarding or the like.
Step S704, the presenting of problem of load balancing.Polycaryon processor mails to Operation ﹠ Maintenance System with these information, and Operation ﹠ Maintenance System is data with table and the form of figure dedicates the user to.The user can observe the load balancing situation and not satisfy customer requirements according to the presenting of load balancing situation, thereby confirmed the existence of problem of load balancing.
Fig. 8 is according to the cpu load situation of the preferred embodiment of the present invention four synoptic diagram that presents to the user, Fig. 9 is according to another cpu load situation of the preferred embodiment of the present invention four synoptic diagram that presents to the user, as Fig. 8 and shown in Figure 9, identify the loading condition of each nuclear respectively with histogram, use cpu busy percentage and two data of user's linking number to add up.
Step S706, multi-nuclear processor equipment load balancing strategy is adjusted.The user uses the new HASH hash factor of Operation ﹠ Maintenance System configuration (that is, the HASH factor adjust), and the distributor of polycaryon processor uses the new HASH factor to carry out meeting between each nuclear to share.Like this, can make the HASH distribution possess stronger adaptive faculty by adjusting the HASH factor.
After step S706, can also carry out the load balancing effect observation, that is, after repeating step S702, user can be observed the new load-balancing method or the HASH hash factor, the load balancing situation of multi-nuclear processor equipment.
In sum, the load balancing mode that the embodiment of the invention proposes utilizes the HASH algorithm, adopts simple distribution rules, for the equipment of polycaryon processor exploitation has increased a kind of function of the dynamic adjustment load balancing HASH factor, realized the polycaryon processor load balancing.
Obviously, those skilled in the art should be understood that, above-mentioned each module of the present invention or each step can realize with the general calculation device, they can concentrate on the single calculation element, perhaps be distributed on the network that a plurality of calculation element forms, alternatively, they can be realized with the executable program code of calculation element, carry out by calculation element thereby they can be stored in the memory storage, perhaps they are made into each integrated circuit modules respectively, perhaps a plurality of modules in them or step are made into the single integrated circuit module and realize.Like this, the present invention is not restricted to any specific hardware and software combination.
The above is the preferred embodiments of the present invention only, is not limited to the present invention, and for a person skilled in the art, the present invention can have various changes and variation.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 disposal route of a polycaryon processor load balancing is characterized in that, may further comprise the steps:
Polycaryon processor extracts the identification information of incoming message;
Described polycaryon processor through the HASH computing, obtains the HASH value corresponding with described identification information with the HASH hash factor of described identification information and selection, and described incoming message sent in the nuclear corresponding with described HASH value handles; And
Loading condition according to each nuclear in the described polycaryon processor is adjusted the HASH hash factor of described selection.
2. method according to claim 1 is characterized in that, described polycaryon processor obtains the HASH hash factor process HASH computing of described identification information and described selection also to comprise before the HASH value corresponding with described identification information:
Described polycaryon processor carries out the HASH computing respectively with the described identification information and all HASH hash factors that are used to calculate, and obtains the distribution situation of all possible HASH hash space in described polycaryon processor each nuclear corresponding with described identification information; And
Select the HASH hash factor of the best HASH hash factor of dispersion degree according to described distribution situation as described selection.
3. method according to claim 2 is characterized in that, selects the best HASH hash factor of dispersion degree to comprise as the HASH hash factor of described selection according to described distribution situation:
Described polycaryon processor calculates the HASH result's who answers with the described all HASH hash factor pairs that are used to calculate variance respectively according to described distribution situation, and from wherein selecting the HASH hash factor corresponding with the minimum variance value as the best HASH hash factor of described dispersion degree.
4. method according to claim 1 is characterized in that, before according to the loading condition of each nuclear in the described polycaryon processor HASH hash factor of described selection being adjusted, also comprises:
Collect the loading condition of each nuclear in the described polycaryon processor according to scheduled time slot, and present to the user with the form of table and/or figure.
5. method according to claim 1 is characterized in that, described incoming message is the model of network traffics.
6. according to each described method in the claim 1 to 5, it is characterized in that described identification information is the feature field of described message, comprise following one of at least: IP address, protocol type.
7. method according to claim 1 is characterized in that, in the described polycaryon processor loading condition of each nuclear comprise following one of at least: the message total of cpu busy percentage, user's linking number, processing or forwarding.
8. the disposal system of a polycaryon processor load balancing is characterized in that, comprises polycaryon processor device and HASH hash factor adjusting gear, wherein,
Described polycaryon processor device comprises: extraction module is used to extract the identification information of incoming message; The HASH computing module is used for the HASH hash factor process HASH computing with described identification information and selection, obtains the HASH value corresponding with described identification information; And distribution module, be used for and described incoming message sent to the nuclear corresponding with described HASH value handling;
Described HASH hash factor adjusting gear comprises, adjusting module is used for according to the loading condition of described each nuclear of polycaryon processor device the HASH hash factor of described selection being adjusted.
9. system according to claim 8 is characterized in that, described polycaryon processor device also comprises:
The HASH hash factor is selected module, be used for the described identification information and all HASH hash factors that are used to calculate are carried out the HASH computing respectively, obtain the distribution situation of all possible HASH hash space in described polycaryon processor each nuclear corresponding with described identification information; And select the HASH hash factor of the best HASH hash factor of dispersion degree as described selection according to described distribution situation.
10. according to Claim 8 or 9 described systems, it is characterized in that described HASH hash factor adjusting gear also comprises:
The load information acquisition module is used for collecting according to scheduled time slot the loading condition of described each nuclear of polycaryon processor device; And
Display module is presented to the user with the loading condition of each nuclear in the described polycaryon processor device of described load information acquisition module collection with the form of table and/or figure.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2010105642054A CN102004673A (en) | 2010-11-29 | 2010-11-29 | Processing method and system of multi-core processor load balancing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2010105642054A CN102004673A (en) | 2010-11-29 | 2010-11-29 | Processing method and system of multi-core processor load balancing |
Publications (1)
Publication Number | Publication Date |
---|---|
CN102004673A true CN102004673A (en) | 2011-04-06 |
Family
ID=43812051
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2010105642054A Pending CN102004673A (en) | 2010-11-29 | 2010-11-29 | Processing method and system of multi-core processor load balancing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102004673A (en) |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102387219A (en) * | 2011-12-13 | 2012-03-21 | 曙光信息产业(北京)有限公司 | Multi-network-card load balancing system and method |
CN102521047A (en) * | 2011-11-15 | 2012-06-27 | 重庆邮电大学 | Method for realizing interrupted load balance among multi-core processors |
CN102546310A (en) * | 2011-12-31 | 2012-07-04 | 曙光信息产业股份有限公司 | Network flow monitoring method |
CN102594621A (en) * | 2012-02-21 | 2012-07-18 | 德科仕通信(上海)有限公司 | Method and system for monitoring high-performance IP (Internet Protocol) media stream |
CN102664815A (en) * | 2012-05-21 | 2012-09-12 | 华为技术有限公司 | Load sharing method, device and system for message traffic |
CN102970244A (en) * | 2012-11-23 | 2013-03-13 | 上海寰创通信科技股份有限公司 | Network message processing method of multi-CPU (Central Processing Unit) inter-core load balance |
CN103049336A (en) * | 2013-01-06 | 2013-04-17 | 浪潮电子信息产业股份有限公司 | Hash-based network card soft interrupt and load balancing method |
CN103582857A (en) * | 2011-04-22 | 2014-02-12 | 高通股份有限公司 | Method and system for thermal load management in a portable computing device |
CN103647708A (en) * | 2013-11-29 | 2014-03-19 | 曙光信息产业(北京)有限公司 | ATCA-based data message processing board |
CN104038437A (en) * | 2014-06-30 | 2014-09-10 | 蓝盾信息安全技术有限公司 | Method for implementing Cavium multi-core processor platform data packet load balance |
CN104301241A (en) * | 2014-06-05 | 2015-01-21 | 中国人民解放军信息工程大学 | SOA dynamic load distribution method and system |
CN106407055A (en) * | 2016-09-13 | 2017-02-15 | 郑州云海信息技术有限公司 | RSS Key value discreteness evaluation method and system |
CN106656842A (en) * | 2015-10-29 | 2017-05-10 | 华为数字技术(苏州)有限公司 | Load balancing method and flow forwarding device |
CN106789705A (en) * | 2015-11-25 | 2017-05-31 | 华为技术有限公司 | The method and apparatus of transmission control message |
CN108460002A (en) * | 2017-12-29 | 2018-08-28 | 中国平安财产保险股份有限公司 | Affiliate's automatic butt method, apparatus, equipment and storage medium |
CN108718259A (en) * | 2018-05-30 | 2018-10-30 | 新华三信息安全技术有限公司 | A kind of message processing method and multi-core processor |
CN108989311A (en) * | 2013-05-31 | 2018-12-11 | 华为技术有限公司 | Generate the method and apparatus of input parameter |
CN111262792A (en) * | 2020-01-17 | 2020-06-09 | 新华三信息安全技术有限公司 | Message forwarding method, device, network equipment and storage medium |
CN113806083A (en) * | 2021-09-06 | 2021-12-17 | 杭州迪普科技股份有限公司 | Method and device for processing aggregation stream data |
CN116820787A (en) * | 2023-08-31 | 2023-09-29 | 湖北芯擎科技有限公司 | Load balancing method, load balancing device, electronic equipment and computer readable storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1937574A (en) * | 2005-09-19 | 2007-03-28 | 北京大学 | Network flow classifying, state tracking and message processing device and method |
CN101610209A (en) * | 2008-11-28 | 2009-12-23 | 北京网康科技有限公司 | A kind of Network method for stream processing and equipment of multi-core parallel concurrent |
CN101217467B (en) * | 2007-12-28 | 2010-10-27 | 杭州华三通信技术有限公司 | An inter-core load dispensing device and method |
-
2010
- 2010-11-29 CN CN2010105642054A patent/CN102004673A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1937574A (en) * | 2005-09-19 | 2007-03-28 | 北京大学 | Network flow classifying, state tracking and message processing device and method |
CN101217467B (en) * | 2007-12-28 | 2010-10-27 | 杭州华三通信技术有限公司 | An inter-core load dispensing device and method |
CN101610209A (en) * | 2008-11-28 | 2009-12-23 | 北京网康科技有限公司 | A kind of Network method for stream processing and equipment of multi-core parallel concurrent |
Cited By (31)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103582857B (en) * | 2011-04-22 | 2017-06-23 | 高通股份有限公司 | Heat load management in portable computing device |
CN103582857A (en) * | 2011-04-22 | 2014-02-12 | 高通股份有限公司 | Method and system for thermal load management in a portable computing device |
CN102521047A (en) * | 2011-11-15 | 2012-06-27 | 重庆邮电大学 | Method for realizing interrupted load balance among multi-core processors |
CN102521047B (en) * | 2011-11-15 | 2014-07-09 | 重庆邮电大学 | Method for realizing interrupted load balance among multi-core processors |
CN102387219A (en) * | 2011-12-13 | 2012-03-21 | 曙光信息产业(北京)有限公司 | Multi-network-card load balancing system and method |
CN102387219B (en) * | 2011-12-13 | 2014-05-28 | 曙光信息产业(北京)有限公司 | Multi-network-card load balancing system and method |
CN102546310A (en) * | 2011-12-31 | 2012-07-04 | 曙光信息产业股份有限公司 | Network flow monitoring method |
CN102594621A (en) * | 2012-02-21 | 2012-07-18 | 德科仕通信(上海)有限公司 | Method and system for monitoring high-performance IP (Internet Protocol) media stream |
CN102664815A (en) * | 2012-05-21 | 2012-09-12 | 华为技术有限公司 | Load sharing method, device and system for message traffic |
CN102970244A (en) * | 2012-11-23 | 2013-03-13 | 上海寰创通信科技股份有限公司 | Network message processing method of multi-CPU (Central Processing Unit) inter-core load balance |
CN102970244B (en) * | 2012-11-23 | 2018-04-13 | 上海寰创通信科技股份有限公司 | A kind of network message processing method of multi -CPU inter-core load equilibrium |
CN103049336A (en) * | 2013-01-06 | 2013-04-17 | 浪潮电子信息产业股份有限公司 | Hash-based network card soft interrupt and load balancing method |
CN108989311A (en) * | 2013-05-31 | 2018-12-11 | 华为技术有限公司 | Generate the method and apparatus of input parameter |
CN103647708A (en) * | 2013-11-29 | 2014-03-19 | 曙光信息产业(北京)有限公司 | ATCA-based data message processing board |
CN104301241B (en) * | 2014-06-05 | 2018-01-09 | 中国人民解放军信息工程大学 | A kind of SOA dynamic load distributing methods and system |
CN104301241A (en) * | 2014-06-05 | 2015-01-21 | 中国人民解放军信息工程大学 | SOA dynamic load distribution method and system |
CN104038437A (en) * | 2014-06-30 | 2014-09-10 | 蓝盾信息安全技术有限公司 | Method for implementing Cavium multi-core processor platform data packet load balance |
CN106656842A (en) * | 2015-10-29 | 2017-05-10 | 华为数字技术(苏州)有限公司 | Load balancing method and flow forwarding device |
CN106789705B (en) * | 2015-11-25 | 2020-04-28 | 华为技术有限公司 | Method and device for transmitting control message |
CN106789705A (en) * | 2015-11-25 | 2017-05-31 | 华为技术有限公司 | The method and apparatus of transmission control message |
CN106407055B (en) * | 2016-09-13 | 2019-09-24 | 郑州云海信息技术有限公司 | A kind of RSS Key value discreteness evaluation method and system |
CN106407055A (en) * | 2016-09-13 | 2017-02-15 | 郑州云海信息技术有限公司 | RSS Key value discreteness evaluation method and system |
CN108460002A (en) * | 2017-12-29 | 2018-08-28 | 中国平安财产保险股份有限公司 | Affiliate's automatic butt method, apparatus, equipment and storage medium |
CN108718259A (en) * | 2018-05-30 | 2018-10-30 | 新华三信息安全技术有限公司 | A kind of message processing method and multi-core processor |
CN108718259B (en) * | 2018-05-30 | 2020-07-03 | 新华三信息安全技术有限公司 | Message processing method and multi-core processor |
CN111262792A (en) * | 2020-01-17 | 2020-06-09 | 新华三信息安全技术有限公司 | Message forwarding method, device, network equipment and storage medium |
CN111262792B (en) * | 2020-01-17 | 2022-04-01 | 新华三信息安全技术有限公司 | Message forwarding method, device, network equipment and storage medium |
CN113806083A (en) * | 2021-09-06 | 2021-12-17 | 杭州迪普科技股份有限公司 | Method and device for processing aggregation stream data |
CN113806083B (en) * | 2021-09-06 | 2023-07-25 | 杭州迪普科技股份有限公司 | Method and device for processing aggregate flow data |
CN116820787A (en) * | 2023-08-31 | 2023-09-29 | 湖北芯擎科技有限公司 | Load balancing method, load balancing device, electronic equipment and computer readable storage medium |
CN116820787B (en) * | 2023-08-31 | 2023-12-15 | 湖北芯擎科技有限公司 | Load balancing method, load balancing device, electronic equipment and computer readable storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102004673A (en) | Processing method and system of multi-core processor load balancing | |
CN109547517B (en) | Method and device for scheduling bandwidth resources | |
CN103369601B (en) | For cell-phone customer terminal provides the method for large concurrent processing and flow control | |
CN103581274B (en) | Message forwarding method and device in stacking system | |
CN102918813A (en) | Device and method for data load balancing | |
CN110266822B (en) | Shared load balancing implementation method based on nginx | |
CN104378270A (en) | Data center system and data center method for networks | |
CN103581966A (en) | Real-time monitoring optimizing configuration method for smart phone | |
CN106878193A (en) | A kind of load sharing method and device | |
CN109412878A (en) | Multi-tenant service access implementation method, device and electronic equipment | |
CN106330610A (en) | Load balance processing system and method | |
CN102904824A (en) | Service providing entity selection method and device | |
CN104852859A (en) | Aggregate interface service processing method and aggregate interface service processing equipment | |
CN107846365A (en) | A kind of load balancing based on SDN realizes system and method | |
CN105376347A (en) | IP address allocation method and system | |
CN103312682A (en) | Method and system for accessing gateway safely | |
CN107579925A (en) | Message forwarding method and device | |
CN102546364A (en) | Network data distribution method and device | |
CN109348434A (en) | A kind of sending method of scene information, sending device and terminal device | |
CN100372323C (en) | A method for accessing server group | |
CN101764761B (en) | Method for limiting transmission flow capacity of service provider interfaces and multi-media service gateway | |
CN103297542A (en) | Operating system bus and balancing method supporting online expansion and retraction of components | |
CN103812707B (en) | A kind of method for processing forwarding of line identification message | |
CN116233200B (en) | Electric energy meter communication method and system based on subsequent frame dynamic registration | |
CN104168206B (en) | Adapter gateway load balancing control method, device and 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 | ||
C12 | Rejection of a patent application after its publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20110406 |