CN115038123A - UPF signaling surface load allocation method, system, electronic equipment and storage medium - Google Patents
UPF signaling surface load allocation method, system, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN115038123A CN115038123A CN202210451016.9A CN202210451016A CN115038123A CN 115038123 A CN115038123 A CN 115038123A CN 202210451016 A CN202210451016 A CN 202210451016A CN 115038123 A CN115038123 A CN 115038123A
- Authority
- CN
- China
- Prior art keywords
- upf
- load
- service instance
- service
- cluster
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
- H04W28/08—Load balancing or load distribution
- H04W28/088—Load balancing or load distribution among core entities
-
- 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
- H04L67/1004—Server selection for load balancing
- H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
- H04W28/08—Load balancing or load distribution
- H04W28/09—Management thereof
- H04W28/0925—Management thereof using policies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/02—Hierarchically pre-organised networks, e.g. paging networks, cellular networks, WLAN [Wireless Local Area Network] or WLL [Wireless Local Loop]
- H04W84/04—Large scale networks; Deep hierarchical networks
- H04W84/08—Trunked mobile radio systems
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Telephonic Communication Services (AREA)
Abstract
The invention provides a UPF signaling surface load allocation method, a system, an electronic device and a storage medium, comprising the following steps: when PFCP signaling is transmitted, allocating UPF service instance load in the UPF cluster based on UPF signaling surface load allocation algorithm; and processing the PFCP message between the SMF and the UPF based on the deployed UPF service instance. According to the invention, by applying a dynamic load allocation algorithm of a cloud native UPF signaling surface, when PFCP signaling is transmitted, dynamic load reallocation is carried out when the load of service instances in a UPF cluster is overlarge, so that normal work of all service instances in the cluster is ensured, and the requirements of high reliability and autonomous fault migration of the cloud native UPF cluster are met.
Description
Technical Field
The present invention relates to the field of mobile communications technologies, and in particular, to a method and a system for allocating a load on a UPF signaling plane, an electronic device, and a storage medium.
Background
In 5G (5th Generation Mobile Communication Technology, fifth Generation Mobile Communication Technology), a 5G core network (5GC) is entirely divided into two categories, namely, CPF (Control Plane Function) and UPF (User Plane Function), as shown in fig. 1.
A Session Management Function (SMF) of a Control plane configures AN UPF through a Packet Forwarding Control Protocol (PFCP), the UPF establishes a tunnel between AN Access Network (AN) and a Data Network (DN) according to Packet information carried by the PFCP, and performs operations such as Forwarding, discarding, caching, and quality of service (QoS) on Data of a User Equipment (UE) according to the tunnel information and a Forwarding rule configured by the PFCP. In a cloud native core network, after a UPF (User Plane Function) is decomposed into a plurality of running service instances, different session establishment exists on the plurality of service instances, and when a load of a certain service instance is too high or a fault occurs, session information established on the service instance needs to be subjected to failover.
Therefore, a new method for load scheduling on the cloud-native UPF signaling plane needs to be proposed.
Disclosure of Invention
The invention provides a method, a system, electronic equipment and a storage medium for allocating UPF signaling surface loads, which are used for solving the defect that in the prior art, a plurality of running service instances in a cloud native core network lack dynamic session message allocation, so that part of the service instances are excessively loaded or have faults.
In a first aspect, the present invention provides a method for allocating a load on a UPF signaling plane, including:
when PFCP signaling is transmitted, allocating UPF service instance load in the UPF cluster based on UPF signaling surface load allocation algorithm;
and processing the PFCP message between the SMF and the UPF based on the deployed UPF service instance.
According to the method for allocating the UPF signaling surface load provided by the invention, the user plane function-based UPF signaling surface load allocation algorithm is used for allocating the UPF service instance load in the UPF cluster, and the method comprises the following steps:
acquiring container pod state information of the service instance through a Kubernets API;
if the number of the service instances corresponding to the pod state information is smaller than the total number of the service instances and any service instance is in an operating state, calculating to obtain the number of the session information to be distributed according to the operating state information of any service instance;
allocating the number of the session information to be distributed to any service instance with the average load less than the number of the session information to be distributed based on a first preset distribution proportion;
if the number of the service instances corresponding to the pod state information is smaller than the total number of the service instances and any service instance is in a fault state, allocating a difference value between the load of any service instance and the average load of the UPF cluster to other service instances based on a second preset distribution proportion;
if the number of the service instances corresponding to the pod state information is determined to be more than or equal to the total number of the service instances, determining that the state updating of all the service instances is completed;
and updating the SEID table record of the session endpoint identification according to the deployed service instance load.
According to the method for allocating the load of the UPF signaling plane provided by the present invention, before acquiring the container pod status information of the service instance through the kubernets API, the method further includes:
calling a metrics interface, and acquiring the CPU occupancy rate, the memory occupancy rate, the service instance capacity and the current session number of the service instance through the metrics interface;
determining the UPF service instance load based on the CPU occupancy rate, the memory occupancy rate, the service instance capacity and the number of the current sessions of the service instance;
acquiring the total number of the service instances in the UPF cluster, and calculating to obtain the average load of the UPF cluster according to the load of the UPF service instances and the total number of the service instances;
obtaining a plurality of service instances with the average load less than the UPF cluster, calculating the difference between each service instance and the average load of the UPF cluster, and calculating the sum of all the differences.
According to the method for allocating the load of the UPF signaling plane provided by the invention, the number of the session information to be distributed is calculated according to the running state information of any service instance, and the method comprises the following steps:
determining that the CPU occupancy rate is greater than a first threshold value, or the memory occupancy rate is greater than a second threshold value, or the ratio of the service instance capacity to the number of the current sessions of the service instance is greater than the first threshold value;
and obtaining the number of the session information to be distributed based on the current session number of any service instance, the difference value between the current session number and the UPF cluster average load and the current load value.
According to the method for allocating the load of the UPF signaling plane provided by the invention, the allocating the number of the session information to be distributed to any service instance with the average load is performed based on the first preset distribution proportion, and the method comprises the following steps:
determining the first preset shunt ratio based on the difference between each service instance and the average load of the UPF cluster and the sum of all the differences;
and multiplying the number of the session information to be distributed by the first preset distribution proportion to obtain the number of the session information distributed to any service instance with the average load smaller than the average load.
According to the method for allocating the load of the UPF signaling plane provided by the invention, the allocating the load of any service instance to other service instances based on the second preset split ratio comprises the following steps:
determining the second preset shunting proportion based on the sum of the current conversation number of any service instance and all difference values;
and multiplying the difference value between the load of any service instance and the UPF cluster average load by the second preset shunting proportion to obtain the number of session information distributed from any service instance by any service instance with the load smaller than the average load.
In a second aspect, the present invention further provides a system for allocating a load on a UPF signaling plane, including:
the dispatching module is used for dispatching UPF service instance loads in the UPF cluster based on a UPF signaling surface load dispatching algorithm when the PFCP signaling is transmitted;
and the processing module is used for processing the PFCP message between the SMF and the UPF based on the allocated UPF service instance.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the method for allocating the load of the UPF signaling plane.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method for UPF signaling plane load balancing as described in any of the above.
The present invention also provides a computer program product, including a computer program, where the computer program, when executed by a processor, implements the method for allocating a load on a UPF signaling plane as described in any of the above.
According to the UPF signaling surface load allocation method, the UPF signaling surface load allocation system, the electronic equipment and the storage medium, a cloud native UPF signaling surface dynamic load allocation algorithm is applied, and when the load of service instances in a UPF cluster is overlarge during PFCP signaling transmission, dynamic load reallocation is carried out, so that all the service instances in the cluster can be guaranteed to normally work, and the requirements of high reliability and autonomous fault migration of the cloud native UPF cluster are met.
Drawings
In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a structure diagram of a user plane function and a control plane function of a core network provided in the prior art;
fig. 2 is a schematic flow chart of a method for scheduling a load on a UPF signaling plane according to the present invention;
fig. 3 is a diagram of a cloud-native UPF signaling plane load balancer architecture provided by the present invention;
fig. 4 is a schematic diagram of a format of a PFCP message header provided by the present invention;
fig. 5 is a second flowchart of a method for allocating a load on a UPF signaling plane according to the present invention;
fig. 6 is a schematic structural diagram of a UPF signaling plane load scheduling system provided in the present invention;
fig. 7 is a schematic structural diagram of an electronic device provided by the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without inventive step based on the embodiments of the present invention, are within the scope of protection of the present invention.
Fig. 2 is a schematic flow chart of a method for scheduling a load on a UPF signaling plane according to the present invention, as shown in fig. 2, the method includes:
step 100: when PFCP signaling is transmitted, allocating UPF service instance load in the UPF cluster based on UPF signaling surface load allocation algorithm;
step 200: and processing the PFCP message between the SMF and the UPF based on the deployed UPF service instance.
It should be noted that, in the core network cloud native evolution process, the key point is to split the network function into a plurality of fine-grained micro services according to the function and other dimensions, and combine virtualization and micro service management to meet the requirements of cloud native elasticity, reliability and expandability.
The invention provides a cloud native UPF signaling surface load allocation algorithm, which is deployed on a UPF signaling surface load balancer device, and the device dynamically reallocates loads when the loads of service instances in a UPF cluster are overlarge during PFCP signaling transmission according to the cloud native UPF signaling surface dynamic load allocation algorithm so as to ensure that all the service instances in the cluster normally work. By the algorithm, the requirements of high reliability and autonomous fault migration of the cloud native UPF cluster can be met. A cloud-native UPF signaling plane load balancer architecture diagram is shown in fig. 3.
Session Endpoint Identifier (SEID) handling is a unique identification of the PFCP session context between PF control plane network element entities (i.e., SMF and UPF). The F-SEID comprises the IP address of the PFCP entity and the distributed SEID, the CP entity and the UPF entity are respectively and independently distributed and sent to the opposite side in the PFCP message, and the PFCP entity of the opposite side takes the identification as the unique identification for identifying the PFCP session.
The PFCP header may be an optional SEID, taking 8 bytes. For node-related messages, the PFCP header does not contain the SEID field; for the session related message, the SEID must be included, and in the session establishment request, the SEID is set to all 0 s. The SEID field is used in the SEID-UPF table to indicate the correspondence of session information to service instances, and the general format of the PFCP message header is shown in fig. 4.
The cloud native UPF signaling plane dynamic load allocation algorithm provided by the invention aims to solve the problem that when a UPF service instance fails or has a high load, the algorithm can dynamically allocate session information on the UPF service instance to other service instances with low loads. The whole algorithm is executed according to periodicity, and the period information can be input in a parameter or configuration item mode when the algorithm is started, namely the algorithm is executed according to the input periodicity.
When PFCP signaling is transmitted, the dynamic load redistribution is carried out when the load of the service instances in the UPF cluster is overlarge, so that the normal work of all the service instances in the cluster is ensured, and the requirements of high reliability and autonomous fault migration of the cloud native UPF cluster are met.
Based on the above embodiments, the allocating the UPF service instance load in the UPF cluster based on the user plane function UPF signaling plane load allocation algorithm includes:
acquiring container pod state information of the service instance through a Kubernetes API;
if the number of the service instances corresponding to the pod state information is smaller than the total number of the service instances and any service instance is in an operation state, calculating to obtain the number of the session information to be distributed according to the operation state information of any service instance;
allocating the number of the session information to be distributed to any service instance with the average load less than the number of the session information to be distributed based on a first preset distribution proportion;
if the number of the service instances corresponding to the pod state information is smaller than the total number of the service instances and any service instance is in a fault state, allocating a difference value between the load of any service instance and the average load of the UPF cluster to other service instances based on a second preset distribution proportion;
if the number of the service instances corresponding to the pod state information is determined to be more than or equal to the total number of the service instances, determining that the state updating of all the service instances is completed;
and updating the SEID table record of the session endpoint identification according to the deployed service instance load.
Specifically, as shown in fig. 5, the pod status is first obtained by calling pod information of the service instance through the kubernets API.
Setting the current service instance number in the UPF cluster as i, comparing the current service instance number i with the service instance total number n, if the service instance number is less than the service instance total number and any service instance is running, further calculating the number of session information to be distributed on the service instance, and distributing the number of the session information to be distributed to other service instances with low load according to a first preset distribution proportion.
And if the number of the service instances is less than the total number of the service instances and any service instance is in a fault state, further allocating the difference value between the load of the service instance and the average load of the UPF cluster to other service instances according to a second preset distribution proportion.
If the number of the current service instances is determined to exceed the total number of the service instances from the beginning, the states of all the service instances in the UPF cluster are considered to be checked and updated, and the algorithm is restarted after waiting for 10 minutes;
and when the deployment is finished, synchronously updating the record in the SEID table, and updating the service example corresponding to the SEID.
On the basis of a cloud native core network, the invention solves the problem that when a UPF service instance has a fault or has a high load, the algorithm can dynamically allocate the session information on the UPF service instance to other service instances with low loads through the UPF signaling surface load allocation algorithm, so that the containerized UPF cluster can meet the requirements of high reliability and autonomous fault migration.
Based on any of the above embodiments, before acquiring the container pod status information of the service instance through the kubernets API, the method further includes:
calling a metrics interface, and acquiring the CPU occupancy rate, the memory occupancy rate, the service instance capacity and the current session number of the service instance through the metrics interface;
determining the UPF service instance load based on the CPU occupancy rate, the memory occupancy rate, the service instance capacity and the number of the current sessions of the service instance;
acquiring the total number of the service instances in the UPF cluster, and calculating to obtain the average load of the UPF cluster according to the load of the UPF service instances and the total number of the service instances;
obtaining a plurality of service instances with the average load less than the UPF cluster, calculating the difference between each service instance and the average load of the UPF cluster, and calculating the sum of all the differences.
Specifically, as shown in fig. 5, a metrics interface is called, a CPU occupancy CPU, a memory occupancy memory, a service instance capacity and a current session number seNumber of the service instance of the UPF service instance are obtained, and a load of the UPF service instance is obtained as follows:
and setting the number of UPF cluster service instances as n, calculating to obtain the average load of the UPF cluster as bLoad:
obtaining service instances with Load smaller than bLoad, and calculating to obtain that the difference value between the bLoad and each service instance is delta Load:
ΔLoad＝bLoad-load
calculating the sum of the differences of all delta loads as sLoad:
according to the invention, by calculating the UPF cluster average load, the load condition of each service instance and the difference value between the load of a single service instance and the cluster average load, the real-time load condition of each service instance can be dynamically obtained, and the subsequent allocation calculation is facilitated.
Based on any of the above embodiments, the calculating the number of session information to be offloaded according to the running state information of any service instance includes:
determining that the CPU occupancy rate is greater than a first threshold value, or the memory occupancy rate is greater than a second threshold value, or the ratio of the service instance capacity to the number of the current sessions of the service instance is greater than the first threshold value;
and obtaining the number of the session information to be distributed based on the current session number of any service instance, the difference value between the current session number and the UPF cluster average load and the current load value.
Specifically, the ith service instance in the UPF cluster is obtained and is set as instance [ i ].
Checking whether the instance of the service is instance [ i ] cpu >0.8 or instance [ i ] memory >0.7 or instance [ i ] capacity/instance [ i ] seNumber >0.8, namely when any rule of the above conditions is met, calculating the number of session information needing to be distributed out by the service instance with higher load as s:
the invention solves the problem that when the load of the UPF service instance is higher, the algorithm can dynamically allocate the session information on the UPF service instance to other service instances with lower load through the UPF signaling surface load allocation algorithm.
Based on any of the above embodiments, the allocating, based on the first preset offloading ratio, the number of session information to be offloaded to any service instance with a load smaller than the average load includes:
determining the first preset shunting proportion based on the difference value of each service instance and the average load of the UPF cluster and the sum of all the difference values;
and multiplying the number of the session information to be distributed by the first preset distribution proportion to obtain the number of the session information distributed to any service instance with the load less than the average load.
Specifically, as shown in fig. 5, the service instances with positive Δ Load in the UPF cluster are selected, the session information of each service instance with positive Δ Load divided from the service instance with higher Load is Δ s,
the invention solves the problem that when the load of the UPF service instance is higher, the algorithm can dynamically allocate the session information on the UPF service instance to other service instances with lower load through the UPF signaling surface load allocation algorithm.
Based on any of the above embodiments, the allocating the load of any service instance to other service instances based on the second preset offloading proportion includes:
determining the second preset shunting proportion based on the sum of the current conversation number of any service instance and all difference values;
and multiplying the difference value between the load of any service instance and the UPF cluster average load by the second preset shunting proportion to obtain the number of session information distributed from any service instance by any service instance with the load smaller than the average load.
Specifically, as shown in fig. 5, similarly, a service instance with a positive Δ Load in the UPF cluster is selected, and session information obtained from a failed service instance by each service instance with a positive Δ Load is calculated as Δ s:
the invention solves the problem that when the UPF service instance has a fault, the algorithm can dynamically allocate the session information on the UPF service instance to other service instances with lower loads through the UPF signaling surface load allocation algorithm.
The following is a description of a complete flow of the cloud-native UPF signaling plane load allocation method shown in fig. 5, and the specific steps include:
(1) the algorithm starts.
(2) And calling a metrics interface to obtain the CPU occupancy rate CPU, the memory occupancy rate memory, the capacity of the service instance and the current session number seNumber of the service instance of the UPF service instance.
(3) Let the load of the UPF service instance be load:
(4) and setting the number of UPF cluster service instances as n, calculating to obtain the average load of the UPF cluster as bLoad:
(5) obtaining service instances with Load smaller than bLoad, and calculating to obtain that the difference value between the bLoad and each service instance is delta Load:
ΔLoad＝bLoad-load
(6) calculating the sum of the differences of all delta loads as sLoad:
(7) and calling the pod information of the service instance through a Kubernetes API to acquire the pod state.
(8) Let i equal 0.
(9) Judging whether i < n is true, if so, proceeding to the step (10); otherwise, go to step (21).
(10) And acquiring the ith service instance in the UPF cluster, and setting the ith service instance as instance [ i ].
(11) Check if the pod status of the service instance [ i ] is running. If the pod state is running, going to the step (12); otherwise, go to step (19).
(12) Checking service instance [ i ]. cpu >0.8 or instance [ i ]. memory >0.7 or instance [ i ]. capacity/instance [ i ]. seNumber > 0.8. If one of the rules is satisfied, proceeding to step (13); otherwise, go to step (18).
(13) The number of session information required to be dropped by the service instance with higher load is calculated as s,
(14) and selecting the service example with the delta Load being positive in the UPF cluster.
(15) The session information of each service instance with positive Δ Load, which is obtained from the service instance with higher Load, is Δ s:
(16) and transferring the session information on the service instance with higher fault or load.
(17) And updating the SEID table, and updating the service example corresponding to the SEID.
(18) Let i be i +1, go to step (9).
(19) And selecting the service example with the delta Load being positive in the UPF cluster.
(20) And (5) calculating the session information of each service instance with positive DeltaLoad, which is obtained from the failed service instance, to be Deltas, and proceeding to the step (17).
(21) And after the state of all the service instances in the UPF cluster is checked and updated, the algorithm is restarted after waiting for 10 minutes.
(22) The algorithm ends.
The present invention provides a system for allocating UPF signaling plane loads, and the system for allocating UPF signaling plane loads and the method for allocating UPF signaling plane loads described above may be referred to in correspondence.
Fig. 6 is a schematic structural diagram of a UPF signaling plane load scheduling system provided in the present invention, as shown in fig. 6, including: a blending module 61 and a processing module 62, wherein:
the allocating module 61 is configured to allocate, when the PFCP signaling is to be transmitted, the UPF service instance load in the UPF cluster based on the UPF signaling plane load allocating algorithm; the processing module 62 is configured to process the PFCP message between the SMF and the UPF based on the provisioned UPF service instance.
The cloud native UPF signaling plane dynamic load allocation algorithm provided by the invention aims to solve the problem that when a UPF service instance fails or has a high load, the algorithm can dynamically allocate session information on the UPF service instance to other service instances with a low load, the whole algorithm is executed periodically, and the period information can be input in a parameter or configuration item mode when the algorithm is started, namely the algorithm is executed according to the input periodicity.
Fig. 7 illustrates a physical structure diagram of an electronic device, and as shown in fig. 7, the electronic device may include: a processor (processor)710, a communication Interface (Communications Interface)720, a memory (memory)730, and a communication bus 740, wherein the processor 710, the communication Interface 720, and the memory 730 communicate with each other via the communication bus 740. Processor 710 may invoke logic instructions in memory 730 to perform a UPF signaling plane load scheduling method comprising: when PFCP signaling is transmitted, allocating UPF service instance load in the UPF cluster based on UPF signaling surface load allocation algorithm; and processing the PFCP message between the SMF and the UPF based on the allocated UPF service instance.
In addition, the logic instructions in the memory 730 can be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In another aspect, the present invention further provides a computer program product, where the computer program product includes a computer program, the computer program may be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, a computer can execute the UPF signaling plane load allocation method provided by the above methods, where the method includes: when PFCP signaling is transmitted, allocating UPF service instance load in the UPF cluster based on UPF signaling surface load allocation algorithm; and processing the PFCP message between the SMF and the UPF based on the deployed UPF service instance.
In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program is implemented to, when executed by a processor, perform the method for UPF signaling plane load allocation provided by the above methods, where the method includes: when PFCP signaling is transmitted, allocating UPF service instance load in the UPF cluster based on UPF signaling surface load allocation algorithm; and processing the PFCP message between the SMF and the UPF based on the deployed UPF service instance.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. Based on the understanding, the above technical solutions substantially or otherwise contributing to the prior art may be embodied in the form of a software product, which may be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method according to the various embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A UPF signaling plane load allocating method is characterized by comprising the following steps:
when a message forwarding control protocol PFCP signaling is transmitted, allocating UPF service instance loads in the UPF cluster based on a user plane function UPF signaling plane load allocation algorithm;
and processing the PFCP message between the SMF and the UPF based on the allocated UPF service instance.
2. The method for allocating UPF signaling plane load according to claim 1, wherein the allocating UPF service instance load in a UPF cluster based on a user plane function UPF signaling plane load allocation algorithm comprises:
acquiring container pod state information of the service instance through a Kubernetes API;
if the number of the service instances corresponding to the pod state information is smaller than the total number of the service instances and any service instance is in an operation state, calculating to obtain the number of the session information to be distributed according to the operation state information of any service instance;
allocating the number of the session information to be distributed to any service instance with the average load less than the number of the session information to be distributed based on a first preset distribution proportion;
if the number of the service instances corresponding to the pod state information is smaller than the total number of the service instances and any service instance is in a fault state, allocating the difference value between the load of any service instance and the average load of the UPF cluster to other service instances based on a second preset distribution ratio;
if the number of the service instances corresponding to the pod state information is determined to be more than or equal to the total number of the service instances, determining that the state updating of all the service instances is completed;
and updating the SEID table record of the session endpoint identification according to the deployed service instance load.
3. The method for allocating UPF signaling plane load according to claim 2, wherein before acquiring the container pod status information of the service instance through kubernets API, the method further comprises:
calling a metrics interface, and acquiring the CPU occupancy rate, the memory occupancy rate, the service instance capacity and the current session number of the service instance through the metrics interface;
determining the UPF service instance load based on the CPU occupancy rate, the memory occupancy rate, the service instance capacity and the number of the current sessions of the service instance;
acquiring the total number of the service instances in the UPF cluster, and calculating to obtain the average load of the UPF cluster according to the load of the UPF service instances and the total number of the service instances;
obtaining a plurality of service instances with the average load less than the UPF cluster, calculating the difference between each service instance and the average load of the UPF cluster, and calculating the sum of all the differences.
4. The method for allocating UPF signaling plane load according to claim 3, wherein the calculating the number of sessions to be offloaded according to the running state information of any service instance includes:
determining that the CPU occupancy rate is greater than a first threshold value, or the memory occupancy rate is greater than a second threshold value, or the ratio of the service instance capacity to the number of the current sessions of the service instance is greater than the first threshold value;
and obtaining the number of the session information to be distributed based on the current session number of any service instance, the difference value between the current session number and the UPF cluster average load and the current load value.
5. The method for allocating the load on the UPF signaling plane according to claim 3, wherein the allocating the number of the session information to be offloaded to any service instance with a load smaller than an average load based on a first preset offloading ratio includes:
determining the first preset shunting proportion based on the difference value of each service instance and the average load of the UPF cluster and the sum of all the difference values;
and multiplying the number of the session information to be distributed by the first preset distribution proportion to obtain the number of the session information distributed to any service instance with the load less than the average load.
6. The method for allocating the load of the UPF signaling plane according to claim 3, wherein the allocating the load of the any service instance to other service instances based on the second predetermined split ratio includes:
determining the second preset shunting proportion based on the sum of the current conversation number of any service instance and all difference values;
and multiplying the difference value between the load of any service instance and the average load of the UPF cluster by the second preset split ratio to obtain the number of session information distributed from any service instance by any service instance with the load smaller than the average load.
7. A system for scheduling a load on a UPF signaling plane, comprising:
the dispatching module is used for dispatching UPF service instance loads in the UPF cluster based on a UPF signaling surface load dispatching algorithm when the PFCP signaling is transmitted;
and the processing module is used for processing the PFCP message between the SMF and the UPF based on the allocated UPF service instance.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the method for UPF signaling plane load balancing according to any one of claims 1 to 6.
9. A non-transitory computer-readable storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the method for UPF signaling plane load balancing according to any one of claims 1 to 6.
10. A computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the UPF signaling plane load scheduling method according to any of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210451016.9A CN115038123A (en) | 2022-04-26 | 2022-04-26 | UPF signaling surface load allocation method, system, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210451016.9A CN115038123A (en) | 2022-04-26 | 2022-04-26 | UPF signaling surface load allocation method, system, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115038123A true CN115038123A (en) | 2022-09-09 |
Family
ID=83119725
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210451016.9A Pending CN115038123A (en) | 2022-04-26 | 2022-04-26 | UPF signaling surface load allocation method, system, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115038123A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115567986A (en) * | 2022-11-29 | 2023-01-03 | 阿里巴巴（中国）有限公司 | Communication network based on load balancing |
-
2022
- 2022-04-26 CN CN202210451016.9A patent/CN115038123A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115567986A (en) * | 2022-11-29 | 2023-01-03 | 阿里巴巴（中国）有限公司 | Communication network based on load balancing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105610632B (en) | Virtual network equipment and related method | |
| US11909663B1 (en) | Methods and apparatus for efficient use of link aggregation groups | |
| CN107026907A (en) | A kind of load-balancing method, load equalizer and SiteServer LBS | |
| CN112703774B (en) | Method for managing processing resources in a telecommunication network, telecommunication network and storage medium | |
| CN109246833B (en) | Bearer configuration determination method, bearer configuration information sending method, bearer configuration determination device, bearer configuration information sending device, main base station and auxiliary base station | |
| WO2023124309A1 (en) | Cloud native upf signaling plane load balancing selection method and system | |
| CN109391691A (en) | The restoration methods and relevant apparatus that NAS is serviced under a kind of single node failure | |
| CN112350952A (en) | Controller distribution method and network service system | |
| CN112671928A (en) | Equipment centralized management architecture, load balancing method, electronic equipment and storage medium | |
| CN112351083A (en) | Service processing method and network service system | |
| CN115038123A (en) | UPF signaling surface load allocation method, system, electronic equipment and storage medium | |
| CN105450540A (en) | Load balancing method and device as well as DHCP (dynamic host configuration protocol) server | |
| CN102612165A (en) | Method and device for releasing resource | |
| CN109922012B (en) | Data transmission method and related equipment | |
| CN109526032B (en) | Method and device for modifying network slice example | |
| CN117354312A (en) | Access request processing method, device, system, computer equipment and storage medium | |
| WO2016206433A1 (en) | Method and apparatus for balancing server load | |
| CN113873008B (en) | Connection reconfiguration method, device, system and medium for RDMA network node | |
| CN107566475B (en) | Session failover method and device | |
| CN112671914B (en) | IOT (Internet of things) equipment communication method and system based on actor model | |
| CN113473571B (en) | Cell access control method, device and base station | |
| CN108200185A (en) | A kind of method and device for realizing load balancing | |
| CN113535402A (en) | Load balancing processing method and device based on 5G MEC and electronic equipment | |
| US11363494B2 (en) | Method and apparatus for session offloading, device and storage medium | |
| KR20210096267A (en) | Data offloading transmission method, network master node (MN) and network secondary node (SN) |
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 |