CN106998340B - Load balancing method and device for board resources - Google Patents
Load balancing method and device for board resources Download PDFInfo
- Publication number
- CN106998340B CN106998340B CN201610046025.4A CN201610046025A CN106998340B CN 106998340 B CN106998340 B CN 106998340B CN 201610046025 A CN201610046025 A CN 201610046025A CN 106998340 B CN106998340 B CN 106998340B
- Authority
- CN
- China
- Prior art keywords
- board card
- weight
- board
- current weight
- load state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000012545 processing Methods 0.000 claims abstract description 56
- 238000004891 communication Methods 0.000 claims abstract description 6
- 238000004364 calculation method Methods 0.000 abstract description 5
- 230000004044 response Effects 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 10
- 238000004590 computer program Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000009826 distribution Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 241001522296 Erithacus rubecula Species 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Multi Processors (AREA)
Abstract
The present invention relates to the field of communications, and in particular, to a method and an apparatus for load balancing of board resources. The method is used for fully utilizing the board card resources in a load balancing mode aiming at the problem of single board resource shortage. The method comprises the following steps: when the network element equipment receives the service request, the real-time load state of each board card is determined according to the current weight of each board card, and the service request is distributed to the board card with the lightest current load for processing, wherein the current weight of one board card is obtained by calculation according to the current weight of at least one load state parameter corresponding to the board card. Therefore, the real-time influence degree of each load performance parameter on the board performance is considered, so that the board with the optimal current performance can be quickly selected, and the processing speed and the response time of the service are improved.
Description
Technical Field
The present invention relates to the field of communications, and in particular, to a method and an apparatus for load balancing of board resources.
Background
With the increase of network traffic, the access volume and data traffic of each core part of the network also increase rapidly, and correspondingly, the computing intensity of each board card in the device also increases correspondingly, so that a single board card gradually cannot bear increasing traffic.
In order to solve the above problems, if the existing equipment is discarded to perform a large amount of board card hardware upgrading, a large amount of waste of the existing resources is caused, and if the service volume is increased again and again, the hardware upgrading mode is adopted to solve the problem, unnecessary high investment is caused, and the equipment with superior performance cannot always meet the increase demand of the service volume, so that the load balancing mechanism is very important.
Load balancing has two implications. Firstly, a large amount of concurrent access or data traffic is shared to a plurality of board cards for processing respectively, so that the waiting response time of a user is reduced; secondly, the operation of the heavy load of a single board card is balanced to a plurality of board cards for parallel processing and returned to the user, so that the processing capacity of the system is greatly improved.
The quality of the load balancing strategy and the difficulty of the implementation of the load balancing strategy have two key factors, namely a first load balancing algorithm, a second load balancing algorithm and a detection mode and capability of the network system condition. The current solution generally selects the following two ways.
The first mode is as follows: and (6) polling equalization.
The method specifically comprises the following steps: and assuming that the processing performance of all the boards is the same, distributing the service requests to all the boards in turn in sequence.
The first method is simple to implement, but is not suitable for the situation that the performances of the board cards are inconsistent, and when the difference between the arrival times of the external service requests is large, the load imbalance among the board cards can be caused.
The second way is: weighted round robin equalization.
The method specifically comprises the following steps: according to the actual configuration conditions of different board cards, a weight is respectively assigned to each board card, the weight is expressed by an integer, and the processing performance of the board cards is represented. When a user sends a service request, the service request is preferentially distributed to the board with the largest weight, the board with the largest weight is endowed with more service requests, and after a period of time, the number of the service requests processed by each board tends to the proportion of the respective weight.
The second method considers the difference of the performances of the boards, but does not consider the real-time load condition of the boards, so when the arrival time difference of the service requests is large, the load imbalance among the boards still occurs.
Disclosure of Invention
The embodiment of the invention provides a load balancing method and device for board card resources, which are used for realizing the full utilization of the board card resources in a load balancing mode aiming at the problem of single board resource shortage.
The embodiment of the invention provides the following specific technical scheme:
a load balancing method for board resources comprises the following steps:
the network element equipment receives a service request;
the network element equipment respectively determines the current weight of each local board card, wherein the current weight of one board card represents the current load state of the one board card, and the current weight is obtained according to the current weight of at least one load state parameter corresponding to the one board card;
and the network element equipment selects one board card from the N board cards with the lowest current weight and distributes the service request to the selected board card for processing.
Preferably, the determining, by the network element device, the current weight of each board card configured locally includes:
before receiving a service request, the network element equipment respectively calculates the current weight of each local board card and updates the current weight periodically, and after receiving the service request, the latest current weight calculated for each board card is read; or,
after receiving the service request, the network element device calculates the current weight of each local board card in real time.
Preferably, the calculating, by the network element device, the current weight of any local board card includes:
the network element equipment respectively calculates the difference value between the current value and the appointed historical value of each load state parameter corresponding to any one board card;
the network element equipment obtains the weight adjustment step length of each load state parameter according to the corresponding difference value of each load state parameter;
the network element equipment calculates and obtains the current weight of each load state parameter based on the weight initial value and the weight scheduling step length corresponding to each load state parameter;
and the network element equipment adds the current weight of each load state parameter to obtain the current weight of any one board card.
Preferably, the historical value of one load state parameter is an initial default value set for the one load state parameter, or a value recorded for the one load state parameter when the weight is calculated last time.
Preferably, the load status parameter corresponding to one board card includes one or any combination of the following:
the CPU occupancy rate is higher, and the corresponding weight is higher;
the physical memory occupancy rate is higher, and the corresponding weight is higher;
the number of the bearing cells is greater, and the corresponding weight is higher;
and if the scheduling board card is not the scheduling board card, the corresponding weight is higher.
Preferably, the selecting, by the network element device, one board card from the N board cards with the lowest current weight, and allocating the service request to the selected board card for processing includes:
if a scheduling board card exists in the N board cards with the lowest current weight, selecting the scheduling board card, and distributing the service request to the scheduling board card for processing;
and if the scheduling board card does not exist in the N board cards with the lowest current weight, randomly selecting one board card from the N board cards with the lowest current weight, or selecting one board card with the lowest current weight, and distributing the service request to the selected board card for processing.
Preferably, after the service request is allocated, the method further includes:
and updating the current weight value corresponding to each local board card.
A board resource load balancing device comprises:
a communication unit for receiving a service request;
the processing unit is used for respectively determining the current weight of each local board card, wherein the current weight of one board card represents the current load state of the one board card, and the current weight is obtained according to the current weight of at least one load state parameter corresponding to the one board card;
and the distribution unit is used for selecting one board card from the N board cards with the lowest current weight value and distributing the service request to the selected board card for processing.
Preferably, when the current weight of each board card configured locally is determined, the processing unit is configured to:
before a service request is received, the current weight of each local board card is calculated respectively and is updated periodically, and after the service request is received, the latest current weight calculated aiming at each board card is read; or,
and after receiving the service request, respectively calculating the current weight of each local board card in real time.
Preferably, when calculating the current weight of any local board card, the processing unit is configured to:
respectively calculating the difference value between the current value and the appointed historical value of each load state parameter corresponding to any one board card;
respectively obtaining the weight adjustment step length of each load state parameter according to the corresponding difference value of each load state parameter;
calculating to obtain the current weight of each load state parameter based on the weight initial value and the weight scheduling step length corresponding to each load state parameter;
and adding the current weight of each load state parameter to obtain the current weight of any one board card.
Preferably, the historical value of one load state parameter is an initial default value set for the one load state parameter, or a value recorded for the one load state parameter when the weight is calculated last time.
Preferably, the load status parameter corresponding to one board card includes one or any combination of the following:
the CPU occupancy rate is higher, and the corresponding weight is higher;
the physical memory occupancy rate is higher, and the corresponding weight is higher;
the number of the bearing cells is greater, and the corresponding weight is higher;
and if the scheduling board card is not the scheduling board card, the corresponding weight is higher.
Preferably, when one board card is selected from the N board cards with the lowest current weight and the service request is allocated to the selected board card for processing, the allocating unit is configured to:
if a scheduling board card exists in the N board cards with the lowest current weight, selecting the scheduling board card, and distributing the service request to the scheduling board card for processing;
and if the scheduling board card does not exist in the N board cards with the lowest current weight, randomly selecting one board card from the N board cards with the lowest current weight, or selecting one board card with the lowest current weight, and distributing the service request to the selected board card for processing.
Preferably, after the service request is allocated, the processing unit is further configured to:
and updating the current weight value corresponding to each local board card.
The invention has the following beneficial effects:
in the embodiment of the invention, when the network element equipment receives the service request, the real-time load state of each board card is determined according to the current weight of each board card, and the service request is distributed to the board card with the lightest current load for processing, wherein the current weight of one board card is obtained by calculation according to the current weight of at least one load state parameter corresponding to the board card. Therefore, the network element equipment can update the current weight of each board card in real time by adopting a dynamic weighting mode according to the quantized load state parameters, so as to master the real-time load performance of each board card, so as to select the board card resource with the optimal performance to process the service request, and because the real-time influence degree of each load performance parameter on the board card performance is considered, the board card with the optimal current performance can be quickly selected, and the processing speed and the response time of the service are improved; and under the condition of single board resource shortage, multi-board resource sharing is realized, and board resources are optimized, so that the existing platform evolution thought is reserved, a new board is not required to be developed, and the operation and maintenance cost is effectively saved.
Drawings
Fig. 1 is a schematic structural diagram of a board card in a network element device according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating an embodiment of load balancing among boards according to the present invention;
fig. 3 is a schematic structural diagram of a network element device in an embodiment of the present invention.
Detailed Description
Aiming at the problem of single board resource shortage, in order to realize the full utilization of the board card resources in a load balancing mode, the invention provides a solution for the allocation and the scheduling of the board card resources. The scheme is as follows: that is, when a service request is received, the load state information of each board is detected in real time, the current weight of each board is calculated according to the difference between each load state parameter included in the load state information and the initial state, and the larger the current weight is, the larger the current load of the board is, so that the board with the lowest current weight is selected to process the service request, and thus, load balance among the boards can be dynamically realized.
Preferred embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.
As shown in fig. 1, a current common board card is generally a single baseband board, that is, only three Digital Signal Processing (DSP) modules are provided, and theoretically, signals of six cells can be processed at most, but in practical applications, for a multi-cell scene or a cell combination scene, the number of cells that can be processed is limited due to the limitation of the capability of the single board, and thus the number of users to be carried is limited, and the coverage area is narrow. Therefore, the invention provides a realization method considering dynamic weighting equalization, and the basic idea is as follows: the current weight of a single board card is subjected to dynamic weighting by adopting multiple load state parameters, the current weight represents the load performance, the larger the current weight is, the larger the load is, the board card can be defined as a heavy load board card, and the board card with the relatively small current weight can be defined as a light load board card. The dynamic state refers to a state (such as the number of cells) in which real-time load is considered, and if a new service request exists, the service request should be allocated to a certain current light-load board card for processing.
In the embodiment of the present invention, when calculating the current weight of a board, various load state parameters of the board may be referred to, including but not limited to:
1. CPU occupancy.
Under normal conditions, a cell and a service currently carried by a board card occupy resources of a CPU, and when the CPU occupancy rate at a certain time point is high, a load of the board card is increased, so that the service processing capability of the board card is reduced, and even the board card is halted. Therefore, for a new service request, the board card resource with low CPU occupancy rate is considered firstly, and the normal operation of the service is ensured.
2. Physical memory occupancy.
The physical memory is relative to the logical memory, and the physical memory affects the performance and other resources, but when the physical memory occupancy rate of the board card is too high, the processing performance of the board card on the service data is affected, so that the physical memory occupancy rate is also an important load state parameter to be considered for a new service request.
3. Number of bearer cells.
At present, a single board has the capability of bearing 6 cells at most, so the number of cells currently borne by each board can be used as a load state parameter for measuring the board load degree, and if the number of cells borne by different boards is different, boards bearing a smaller number of cells are considered for a new service request.
4. Whether the board card is a dispatching board card
When the problem of resource scheduling between the board cards is involved, the weights given by the scheduling board card and the scheduled board card are different, the scheduling board card receiving the service request is considered for a new service request, and the scheduled board card is selected if the current weight of the scheduling board card is greater than the current weight of the scheduled board card.
Referring to fig. 2, in the embodiment of the present invention, a specific process of implementing load balancing of board card resources is as follows:
step 200: the network element equipment receives the service request.
Referring to fig. 1, in the embodiment of the present invention, a network element device (e.g., a base station) is generally provided with a plurality of boards, for example, a 1-slot board, a 2-slot board, and an … … 7-slot board, where each board may serve as a scheduling board to receive a service request, calculate a current weight of each board, and determine whether to allocate the service request to a scheduled board for processing or leave the scheduled board for processing.
Step 210: the network element equipment determines the current weight of each local board card respectively, wherein the current weight of one board card represents the current load state of the one board card, and the current weight is obtained according to the current weight of at least one load state parameter corresponding to the one board card.
In the embodiment of the present invention, when step 210 is executed, the following ways may be adopted, but not limited to,
the first mode is as follows: before receiving a service request, the network element device calculates the current weight of each local board card respectively, updates the current weight periodically, and reads the latest current weight calculated for each board card respectively after receiving the service request.
When the first method is adopted, the network element device calculates the current weight of each local board card periodically or according to a trigger event in advance, so that after receiving the service request, the newly calculated current weight of each local board card can be directly read, thereby saving calculation time and improving the adjustment time of load balance among the board cards.
The second way is: after receiving the service request, the network element device calculates the current weight of each local board card in real time.
When the second mode is adopted, the network element equipment calculates the current weight of each local board card in real time after receiving the service request, so that the latest current weight can be obtained in time, and the load balance among the board cards is more favorably realized.
Specifically, taking any board as an example (hereinafter referred to as a board a), when the network element device calculates the current weight of the board a, the following operations may be performed:
firstly, the network element equipment respectively calculates the difference value between the current value and the appointed historical value of each load state parameter corresponding to the board card A.
Specifically, the historical value of one load state parameter may be an initial default value set for the one load state parameter, or may also be a value recorded for the one load state parameter when the weight is calculated last time. For example, taking the load state parameter a as an example, assuming that the current value of the load state parameter a is a2, the initial default value of the load state parameter a is a0, and the current value of the load state parameter a is changed to a1 in the following process, when the difference value from the historical value is calculated, the difference value between a2 and a0 may be calculated, and the difference value between a2 and a1 may also be calculated.
And secondly, the network element equipment obtains the weight adjustment step length of each load state parameter according to the corresponding difference value of each load state parameter.
Specifically, the difference of the values of one load state parameter at different times represents the change state of the one load state parameter, and the change of the value of one load state parameter affects the current weight of the corresponding board card, so that the corresponding weight adjustment step length is set for the change (i.e., the difference) of the value of one load state parameter.
And thirdly, the network element equipment calculates and obtains the current weight of each load state parameter based on the weight initial value and the weight scheduling step length corresponding to each load state parameter.
For example, the load state parameter a is still used as an example.
If the difference between the history value and the current value is calculated by using the difference between a2 and a0, the current weight of the load state parameter a can be calculated by adding the weight adjustment step corresponding to the difference between a2 and a0 to the initial weight value corresponding to a 0.
If the difference between the historical value and the load state parameter a is calculated by using the difference between a2 and a1, the current weight of the load state parameter a can be calculated by adding the weight adjustment step corresponding to the difference between a2 and a0 to the weight value corresponding to a 1.
Taking the load status parameter a as an example, the network element device may calculate the corresponding current weight for each load status parameter of each board card by using the above method.
And finally, the network element equipment adds the current weight of each load state parameter to obtain the current weight of the board card A.
In the above embodiment, the load status parameters of the board card a may be selected from a variety of parameters, including but not limited to one or any combination of the following parameters:
the CPU occupancy rate is higher, and the corresponding weight is higher;
the physical memory occupancy rate is higher, and the corresponding weight is higher;
the number of the bearing cells is greater, and the corresponding weight is higher;
and if the scheduling board card is not the scheduling board card, the corresponding weight is higher.
Step 220: and the network element equipment selects one board card from the N board cards with the lowest current weight and distributes the service request to the selected board card for processing.
Specifically, when step 220 is executed, the following two ways may be adopted, but not limited to:
the first mode is as follows: and if the scheduling board card exists in the N board cards with the lowest current weight, selecting the scheduling board card, and distributing the service request to the scheduling board card for processing.
In practical application, if the current weight of the scheduling board card is lower, the current load of the scheduling board card is lower, and then the scheduling board card is preferentially selected for service request processing, so that load balancing among the board cards can be realized, and resource loss caused by cross-board scheduling can be avoided.
The second way is: and if the scheduling board card does not exist in the N board cards with the lowest current weight, randomly selecting one board card from the N board cards with the lowest current weight, or selecting one board card with the lowest current weight, and distributing the service request to the selected board card for processing.
Because the current load of the scheduling board card is heavier, cross-board scheduling needs to be performed, and one board card is selected from the N board cards with the lowest current weight to perform service request processing, so that the overweight operation load brought to the scheduling board card can be avoided, and meanwhile, the load balance among the board cards can be realized in real time.
Preferably, after the service request is distributed to the corresponding board card and processed, the network element device updates the current weight corresponding to each local board card again, and because the load state parameter of each board card changes with the distribution of the service request, the current weight of each board card needs to be updated in real time.
The above embodiments are further described in detail by a specific application scenario.
Assuming that the set of load state parameters that the network element device needs to calculate is R ═ { R1, R2, …, Rn }, where n represents the total number of load state parameters, in the embodiment of the present invention, n ═ 4, where R1 represents CPU occupancy, R2 represents physical memory occupancy, R3 represents the number of load cells, and R4 represents whether the current board is a dispatch board. In addition, in this embodiment, the service request specifically takes a cell establishment request as an example.
Based on the four load state parameters, according to the importance degree of each load state parameter relative to the board card performance, each load state parameter is quantized, and the current weight Y (R) of different board cards is updated in real time in a weighting mode.
For example, when a new cell establishment request arrives, a board card resource with a light load, that is, y (r), needs to be selected, and assuming that there are currently 2 board cards, which are a 4-slot board card, a 5-slot board card, and a 4-slot board card, which is currently used as a scheduling board card to receive the cell establishment request, then when the board card resource is selected, the current weight y (r) of each board card needs to be calculated.
Each load state parameter is quantized, and according to the influence degree, the temporary assumption is that: r3> R1> R4> R2;
and assuming that the unit of the weight is r, and the initial value of the weight corresponding to each load state parameter is: r3: 5 r; r1: 3 r; r2: 2 r; r1: and r.
Assuming that the initialized value of the CPU occupancy rate is 0, and every time the CPU occupancy rate is increased by 5%, the weight corresponding to the CPU occupancy rate is increased by 3r, namely the weight adjustment step length of the CPU occupancy rate is 3 r;
assuming that the initialization value of the physical memory occupancy rate is 0, and the weight of the physical memory occupancy rate is increased by r every time the physical memory occupancy rate is increased by 5%, namely the weight adjustment step length of the physical memory occupancy rate is r;
assume that the initial value of the number of bearer cells is 0, and every time a cell is added, the weight of the number of bearer cells is increased by 5r, i.e. the weight scheduling step of the number of bearer cells is 5 r.
Assuming that whether the initialization value of the scheduling board card is 0 or not, the weight bit of the scheduled board card is increased by 2r every time of cross-board scheduling, that is, the weight adjustment step of whether the scheduled board card is the scheduling board card is 2r or not.
Assuming that when a received cell establishment request is received, the current CPU occupancy rate of the 4 slot card is 40%, the physical memory occupancy rate is 30%, and the number of load cells is 2; and the CPU occupancy rate of the 5 slot card is 30%, the physical memory occupancy rate is 40%, and the number of the bearing cells is 1. Then
When a new cell establishment request is sent to the 4 slot board cards, the 4 slot board cards need to select appropriate board card resources according to the load of the current 4 slot board cards and 5 slot board cards, and respond to the corresponding request information of the cell establishment request, specifically, the calculation mode of the current weights of the 4 slot board cards and the 5 slot board cards is as follows:
Y(4)=(40%/5%)*3r+(30%/5%)*r+2*5r=40r;
Y(5)=(30%/5%)*3r+(40%/5%)*r+1*5r+2r=33r。
according to the above algorithm, at this time, the current weight Y (5) < Y (4) indicates that the current load of the 5 slot card is light, and therefore, the 4 slot card needs to allocate the received cell establishment request to the 5 slot card for processing, thereby realizing reasonable allocation of card resources.
And after the 5 slot position board cards respond to the distributed cell establishment requests, Y (5) can also dynamically change correspondingly, so that when a next new cell establishment request is received, the scheduling board cards can adopt the updated current weight values for comparison, and then proper board card resources are selected, thereby facilitating the reasonable allocation of the whole resources and the great improvement of the performance.
Based on the above embodiments, referring to fig. 3, in an embodiment of the present invention, a network element device (i.e. a scheduler board card that can be understood to receive a service request) includes at least a communication unit 30, a processing unit 31 and an allocating unit 32, where,
a communication unit 30 for receiving a service request;
the processing unit 31 is configured to determine a current weight of each local board card, where the current weight of one board card represents a current load state of the one board card, and the current weight is obtained according to a current weight of at least one load state parameter corresponding to the one board card;
the allocating unit 32 is configured to select one board card from the N board cards with the lowest current weight, and allocate the service request to the selected board card for processing.
Preferably, when determining the current weight of each board card configured locally, the processing unit 31 is configured to:
before a service request is received, the current weight of each local board card is calculated respectively and is updated periodically, and after the service request is received, the latest current weight calculated aiming at each board card is read; or,
and after receiving the service request, respectively calculating the current weight of each local board card in real time.
Preferably, when calculating the current weight of any local board, the processing unit 31 is configured to:
respectively calculating the difference value between the current value and the appointed historical value of each load state parameter corresponding to any one board card;
respectively obtaining the weight adjustment step length of each load state parameter according to the corresponding difference value of each load state parameter;
calculating to obtain the current weight of each load state parameter based on the weight initial value and the weight scheduling step length corresponding to each load state parameter;
and adding the current weight of each load state parameter to obtain the current weight of any one board card.
Preferably, the historical value of one load state parameter is an initial default value set for the one load state parameter, or a value recorded for the one load state parameter when the weight is calculated last time.
Preferably, the load status parameter corresponding to one board card includes one or any combination of the following:
the CPU occupancy rate is higher, and the corresponding weight is higher;
the physical memory occupancy rate is higher, and the corresponding weight is higher;
the number of the bearing cells is greater, and the corresponding weight is higher;
and if the scheduling board card is not the scheduling board card, the corresponding weight is higher.
Preferably, when one board card is selected from the N board cards with the lowest current weight and the service request is allocated to the selected board card for processing, the allocating unit 32 is configured to:
if the scheduling board card exists in the N board cards with the lowest current weight, selecting the scheduling board card, and distributing the service request to the scheduling board card for processing;
and if the scheduling board card does not exist in the N board cards with the lowest current weight, randomly selecting one board card from the N board cards with the lowest current weight, or selecting one board card with the lowest current weight, and distributing the service request to the selected board card for processing.
Preferably, after the service request is allocated, the processing unit 31 is further configured to:
and updating the current weight value corresponding to each local board card.
In summary, in the embodiment of the present invention, when receiving a service request, a network element device determines a real-time load state of each board card according to a current weight of each board card, and allocates the service request to a board card with a lightest current load for processing, where a current weight of one board card is obtained by calculation according to a current weight of at least one load state parameter corresponding to the board card. Therefore, the network element equipment can update the current weight of each board card in real time by adopting a dynamic weighting mode according to the quantized load state parameters, so as to master the real-time load performance of each board card, so as to select the board card resource with the optimal performance to process the service request, and because the real-time influence degree of each load performance parameter on the board card performance is considered, the board card with the optimal current performance can be quickly selected, and the processing speed and the response time of the service are improved; and under the condition of single board resource shortage, multi-board resource sharing is realized, and board resources are optimized, so that the existing platform evolution thought is reserved, a new board is not required to be developed, and the operation and maintenance cost is effectively saved.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various modifications and variations can be made in the embodiments of the present invention without departing from the spirit or scope of the embodiments of the invention. Thus, if such modifications and variations of the embodiments of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to encompass such modifications and variations.
Claims (10)
1. A load balancing method for board resources is characterized by comprising the following steps:
the network element equipment receives a service request;
the network element device determines a current weight of each local board card, wherein the calculating, by the network element device, the current weight of any local board card specifically includes: the network element equipment respectively calculates the difference value between the current value and the appointed historical value of each load state parameter corresponding to any one board card; the network element equipment obtains the weight adjustment step length of each load state parameter according to the corresponding difference value of each load state parameter; the network element equipment calculates and obtains the current weight of each load state parameter based on the weight initial value and the weight scheduling step length corresponding to each load state parameter; the network element equipment adds the current weight of each load state parameter to obtain the current weight of any one board card, the current weight of any one board card represents the current load state of any one board card, and the current weight is obtained according to the current weight of at least one load state parameter corresponding to any one board card;
the load state parameter corresponding to one board card comprises one or any combination of the following:
the CPU occupancy rate of the central processing unit is higher, wherein the higher the CPU occupancy rate is, the higher the corresponding weight is; the physical memory occupancy rate is higher, and the corresponding weight is higher; the number of the bearing cells is greater, and the corresponding weight is higher; if the scheduling board card is not the scheduling board card, the corresponding weight is higher;
and if the scheduling board card exists in the N board cards with the lowest current weight, the network element equipment selects the scheduling board card and distributes the service request to the scheduling board card for processing.
2. The method of claim 1, wherein the determining, by the network element device, the current weight value of each locally configured board separately includes:
before receiving a service request, the network element equipment respectively calculates the current weight of each local board card and updates the current weight periodically, and after receiving the service request, the latest current weight calculated for each board card is read; or,
after receiving the service request, the network element device calculates the current weight of each local board card in real time.
3. The method of claim 1, wherein the historical value of one load state parameter is an initial default value set for the one load state parameter, or a value recorded for the one load state parameter when the weight is calculated last time.
4. The method according to any of claims 1-3, wherein the network element device selects one board from the N boards with the lowest current weight, further comprising:
and if the scheduling board card does not exist in the N board cards with the lowest current weight, randomly selecting one board card from the N board cards with the lowest current weight, or selecting one board card with the lowest current weight, and distributing the service request to the selected board card for processing.
5. The method of claim 4, wherein assigning the service request is further performed by:
and updating the current weight value corresponding to each local board card.
6. A board resource load balancing device is characterized by comprising:
a communication unit for receiving a service request;
the processing unit is configured to determine a current weight of each local board, where calculating the current weight of any local board is specifically used for: respectively calculating the difference value between the current value and the appointed historical value of each load state parameter corresponding to any one board card; respectively obtaining the weight adjustment step length of each load state parameter according to the corresponding difference value of each load state parameter; calculating to obtain the current weight of each load state parameter based on the weight initial value and the weight scheduling step length corresponding to each load state parameter; adding the current weight of each load state parameter to obtain the current weight of any one board card, wherein the current weight of any one board card represents the current load state of any one board card, and the current weight is obtained according to the current weight of at least one load state parameter corresponding to any one board card;
the load state parameter corresponding to one board card comprises one or any combination of the following:
the CPU occupancy rate of the central processing unit is higher, wherein the higher the CPU occupancy rate is, the higher the corresponding weight is; the physical memory occupancy rate is higher, and the corresponding weight is higher; the number of the bearing cells is greater, and the corresponding weight is higher; if the scheduling board card is not the scheduling board card, the corresponding weight is higher;
and the allocating unit is used for selecting the scheduling board card by the network element equipment and allocating the service request to the scheduling board card if the scheduling board card exists in the N board cards with the lowest current weight.
7. The apparatus of claim 6, wherein when determining the current weight for each of the locally configured boards, the processing unit is configured to:
before a service request is received, the current weight of each local board card is calculated respectively and is updated periodically, and after the service request is received, the latest current weight calculated aiming at each board card is read; or,
and after receiving the service request, respectively calculating the current weight of each local board card in real time.
8. The apparatus of claim 6, wherein the historical value of one load state parameter is an initial default value set for the one load state parameter, or a value recorded for the one load state parameter when the weight is calculated last time.
9. The apparatus of any of claims 6-8, wherein one card is selected from the N cards with the lowest current weight, and the allocation unit is further configured to:
and if the scheduling board card does not exist in the N board cards with the lowest current weight, randomly selecting one board card from the N board cards with the lowest current weight, or selecting one board card with the lowest current weight, and distributing the service request to the selected board card for processing.
10. The apparatus as claimed in claim 9, wherein said processing unit is further configured to, after assigning said service request:
and updating the current weight value corresponding to each local board card.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610046025.4A CN106998340B (en) | 2016-01-22 | 2016-01-22 | Load balancing method and device for board resources |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610046025.4A CN106998340B (en) | 2016-01-22 | 2016-01-22 | Load balancing method and device for board resources |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106998340A CN106998340A (en) | 2017-08-01 |
CN106998340B true CN106998340B (en) | 2020-10-30 |
Family
ID=59428799
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610046025.4A Active CN106998340B (en) | 2016-01-22 | 2016-01-22 | Load balancing method and device for board resources |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106998340B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3761609A4 (en) * | 2019-05-09 | 2021-01-13 | Wangsu Science & Technology Co., Ltd. | Load adjustment method and system for resource server, and device |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108055333A (en) * | 2017-12-21 | 2018-05-18 | 创新科存储技术(深圳)有限公司 | A kind of NAS-CIFS cluster load balancing methods based on UFS |
CN112312202B (en) * | 2020-08-10 | 2023-02-28 | 浙江宇视科技有限公司 | Decoding splicing processing equipment |
CN116155438A (en) * | 2021-11-23 | 2023-05-23 | 中国电信股份有限公司 | Optical multiplexing segment creation method and device, network management system and box type wavelength division device |
CN115048206B (en) * | 2022-08-15 | 2022-12-27 | 阿里巴巴(中国)有限公司 | Resource scheduling method and server |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102882959A (en) * | 2012-09-21 | 2013-01-16 | 国电南瑞科技股份有限公司 | Load balancing mechanism for WEB server in electric power scheduling system |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101860486A (en) * | 2010-06-07 | 2010-10-13 | 北京邮电大学 | Dynamic load balancing mechanism based on leaky bucket algorithm |
CN104423529B (en) * | 2013-08-30 | 2018-08-21 | 华为技术有限公司 | Central processing unit state adjustment method and device |
CN103501509B (en) * | 2013-09-04 | 2016-08-17 | 大唐移动通信设备有限公司 | A kind of method and device equalizing radio network controller load |
CN104702521A (en) * | 2013-12-06 | 2015-06-10 | 中兴通讯股份有限公司 | Load balance method and device |
EP2947908A1 (en) * | 2014-05-23 | 2015-11-25 | Alcatel Lucent | Management method for use by a management system |
-
2016
- 2016-01-22 CN CN201610046025.4A patent/CN106998340B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102882959A (en) * | 2012-09-21 | 2013-01-16 | 国电南瑞科技股份有限公司 | Load balancing mechanism for WEB server in electric power scheduling system |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3761609A4 (en) * | 2019-05-09 | 2021-01-13 | Wangsu Science & Technology Co., Ltd. | Load adjustment method and system for resource server, and device |
Also Published As
Publication number | Publication date |
---|---|
CN106998340A (en) | 2017-08-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106998340B (en) | Load balancing method and device for board resources | |
CN110474966B (en) | Method for processing cloud platform resource fragments and related equipment | |
EP2193472B1 (en) | Method and apparatus for managing license capacity in a telecommunication network | |
CN109960585A (en) | A kind of resource regulating method based on kubernetes | |
CN107908457B (en) | Containerized cloud resource allocation method based on stable matching | |
CN111694663B (en) | Load balancing method, device and system for server cluster | |
CN103701916A (en) | Dynamic load balancing method of distributed storage system | |
CN108628674A (en) | Method for scheduling task, cloud platform based on cloud platform and computer storage media | |
CN103401947A (en) | Method and device for allocating tasks to multiple servers | |
CN107562545A (en) | A kind of container dispatching method based on Docker technologies | |
CN104619029B (en) | It is a kind of centralization cellular network architecture under baseband pool resource allocation methods and device | |
CN113115252B (en) | Delay sensitive task distributed mobile edge computing resource scheduling method and system | |
WO2022257631A1 (en) | Mobile edge computing-oriented distributed special protection service scheduling method | |
CN103731372A (en) | Resource supply method for service supplier under hybrid cloud environment | |
CN102110014A (en) | Method for balancing loads of virtual machine (VM) | |
CN102664814A (en) | Grey-prediction-based adaptive dynamic resource allocation method for virtual network | |
CN104980935A (en) | Method, device and system for sharing network | |
CN104917839A (en) | Load balancing method used in cloud computing environment | |
CN114625500A (en) | Method and application for scheduling micro-service application based on topology perception in cloud environment | |
CN112559122A (en) | Virtualization instance management and control method and system based on electric power special security and protection equipment | |
CN110167031B (en) | Resource allocation method, equipment and storage medium for centralized base station | |
CN113672391A (en) | Parallel computing task scheduling method and system based on Kubernetes | |
CN107203256B (en) | Energy-saving distribution method and device under network function virtualization scene | |
CN107491352B (en) | Resource scheduling method and device | |
CN103458520A (en) | Allocation method and device of uplink frequency domain resources |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |