CN117579629A - Calculation force matching method for digital networking - Google Patents
Calculation force matching method for digital networking Download PDFInfo
- Publication number
- CN117579629A CN117579629A CN202311427995.5A CN202311427995A CN117579629A CN 117579629 A CN117579629 A CN 117579629A CN 202311427995 A CN202311427995 A CN 202311427995A CN 117579629 A CN117579629 A CN 117579629A
- Authority
- CN
- China
- Prior art keywords
- node
- power
- computing power
- participating
- hub
- 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
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000006855 networking Effects 0.000 title claims abstract description 32
- 238000004364 calculation method Methods 0.000 title claims abstract description 24
- 230000004044 response Effects 0.000 claims abstract description 22
- 238000012544 monitoring process Methods 0.000 claims description 21
- 230000008569 process Effects 0.000 claims description 10
- 238000009826 distribution Methods 0.000 claims description 5
- 230000007423 decrease Effects 0.000 claims description 4
- 230000009467 reduction Effects 0.000 claims description 3
- 238000003860 storage Methods 0.000 description 9
- 238000002955 isolation Methods 0.000 description 7
- 230000008859 change Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 230000010354 integration Effects 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000004590 computer program Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000007726 management method Methods 0.000 description 4
- 238000013468 resource allocation Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000003068 static effect Effects 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 206010000117 Abnormal behaviour Diseases 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000012550 audit Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 238000013524 data verification Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
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
- H04L67/104—Peer-to-peer [P2P] networks
- H04L67/1059—Inter-group management mechanisms, e.g. splitting, merging or interconnection of groups
-
- 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
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
-
- 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
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the disclosure provides a computing power matching method for digital networking. The digital network includes a plurality of subnets. Each sub-network includes a hub node and a plurality of participating nodes directly connected to the hub node. The hub nodes in the plurality of subnetworks are directly connected to each other. The calculation force matching method comprises the following steps: each hub node obtains a digital networking computing power resource panorama; each hub node obtains a taxi calculation power panorama updated in real time by the digital network; in response to a first hub node in a first subnet receiving a purchase order of a computing power resource submitted by a participating node in the first subnet as a computing power renter, the first hub node: determining candidate participation nodes with computing power resources matched with the computing power resource buying list in the digital network according to the computing power resource panorama; determining a target computing power renter matched with the computing power resource buying list from the candidate participating nodes in real time according to the renting power panorama; and the target hub node directly connected with the target power calculation renter constructs a power calculation renting relation in a signing mode.
Description
Technical Field
The embodiment of the disclosure relates to the technical field of digital networking, in particular to a computing power matching method for digital networking.
Background
Data sharing and the circulation of data value are increasingly important to the current advance of digital economic transformation. With the continuous rise in the demands for intercommunication, sharing, and exchange of data, the "digital networking" has also grown following the Personal Computer (PC) internet, mobile internet, industrial internet, and internet of things. The new "digital networking" referred to herein is an infrastructure that enables data elements to circulate across industries, regions, and organizations, in contrast to traditional narrow digital networking (the internet used only for data buying and selling transactions in digital offices or similar scenarios). The narrow digital networking is simply the "digital-to-digital connected internet" and involves only transparent transmission of data in the digital networking. The application end can directly acquire data to generate various business applications based on the data.
In the new "digital networking", the transfer of data usage rights should be more reasonably ordered. In such digital networking, the very critical underlying capability is computing power. The calculation force is loose in distribution. The digital networking is used as an open network system, and can support a mass main body to access the digital networking in an allowed mode. The body accessing the data network constitutes one data node in the data network. The node has a certain computational power resource. When the local computing tasks of some nodes are finished, the computing power resources of the nodes are idle to a certain extent, so that the computing power resources are wasted. If the idle computing power resources are leased to other nodes, the computing power resources of the digital network can be fully utilized, so that the nodes participating in the digital network obtain values except data circulation.
Disclosure of Invention
Embodiments described herein provide a method of computing power matching for digital networking.
According to a first aspect of the present disclosure, a method of computing power matching for digital networking is provided. The digital network includes a plurality of subnets. Each sub-network includes a hub node and a plurality of participating nodes directly connected to the hub node. The hub nodes in the plurality of subnetworks are directly connected to each other. The calculation force matching method comprises the following steps: each hub node obtains a digital networking computing power resource panorama, wherein the computing power resource panorama comprises computing power resource information of all participating nodes in the digital networking; each hub node obtains a leasing power panoramic table updated in real time by the digital network, wherein the leasing power panoramic table comprises all the leasing power leasing information issued by the participating nodes serving as the leasing power parties in the digital network; in response to a first hub node in a first subnet receiving a purchase order of a computing power resource submitted by a participating node in the first subnet as a computing power renter, the first hub node performs the following operations: determining candidate participation nodes with computing power resources matched with the computing power resource buying list in the digital network according to the computing power resource panorama; determining a target computing power renter matched with the computing power resource buying list from the candidate participating nodes in real time according to the renting power panorama; and constructing a calculation force renting relation by signing with a target hub node directly connected with the target calculation force renter.
In some embodiments of the present disclosure, the force matching method further comprises: each computing power renter isolates idle computing power resources to be rented in a container mode; in response to the number of target power lessors being greater than 1, grouping containers in all target power lessors into a container cluster; selecting a main hub node from target hub nodes directly connected by each target computing lessor; in the process of executing the instruction of the computing power renter, the main hub node monitors the load condition of the container cluster in real time; in response to the load reduction, the primary hub node downscales the cluster of containers; and in response to the load increase, the primary hub node expands the size of the container cluster.
In some embodiments of the present disclosure, the force matching method further comprises: in response to the number of target power lessors being equal to 1, the target hub node to which the target power lessors are directly connected collecting and managing log data of containers in the target power lessors, and restarting the containers in case of failure or unavailability of the containers; and responsive to the number of targeted power renters being greater than 1, the primary hub node collecting and managing log data for individual containers in the cluster of containers and restarting the containers in the event that any of the containers fails or is unavailable.
In some embodiments of the present disclosure, each hub node obtaining a digital networked real-time updated rental force panorama comprises: collecting, by a hub node in each subnet, computing power lease information published by a plurality of participating nodes in the subnet; all hub nodes in the Internet of things forward the collected calculation power renting information to each other so that each hub node forms a calculation power panoramic table of the Internet of things respectively; in response to the first participating node's power lease information having been updated, the first participating node immediately issues updated power lease information to its directly connected designated hub node, the designated hub node updates the power panorama in real time using the updated power lease information of the first participating node, and the designated hub node forwards the updated power lease information of the first participating node to other hub nodes in the digital network so that the other hub nodes update the power panorama in real time.
In some embodiments of the present disclosure, the method of power matching further includes each participating node performing the following: predicting the resource demand of the participating node in a future period according to the historical computing power resource expense of the participating node; determining idle computing power resources of the participating node in a period of time in the future according to computing power resources of the participating node and the predicted resource requirements; the participating nodes isolate idle computing resources in the form of containers; releasing unused resources in the container in response to the predicted increase in resource demand; and increasing the resources in the container in response to the predicted decrease in the resource demand.
In some embodiments of the present disclosure, the method of power matching further includes each participating node performing the following: the method comprises the steps of monitoring the use condition of the computing power resource of the participating node in real time to determine the idle computing power resource of the participating node in real time, wherein the use condition comprises the distribution condition, the use rate and the idle time of the computing power resource; the participating nodes isolate idle computing resources in the form of containers; responsive to an increase in the real-time usage of the computational resources of the participating node, immediately releasing unused resources in the container; and in response to the real-time usage of the computational resources of the participating node decreasing, immediately increasing the resources in the container.
In some embodiments of the present disclosure, the force matching method further comprises: the target hub node immediately freezes the computing power resources leased by the target computing power leasing party upon completion of the subscription with the first hub node.
In some embodiments of the present disclosure, the force matching method further comprises: the second hub node calculates the calculation force contribution degree of all second participation nodes directly connected with the second hub node; the second hub node calculates the contribution degree duty ratio of each second participation node according to the calculated force contribution degree of the second participation node; and the second hub node distributes rewards to the second participation nodes according to the contribution ratio of each second participation node so as to improve the participation degree of the second participation nodes in the process of computing the power match.
Wherein the computational power contribution of the j-th node in the second participating nodes is calculated as:
wherein score j Representing the calculation force contribution degree of the j-th node, R ai Representing the number, W, of ith computational resource in the n computational resources of the jth node rai Representing weights set according to the class of the ith computational resource, R qi An ith quality parameter, W, of m quality parameters representing a jth node rqi Represents the weight set for the ith quality parameter, W rt Representing weights determined from time of leasing power of jth node, W rc Representing the weight determined according to the completion rate of the task in which the j-th node participates.
Wherein the contribution duty cycle of the j-th node is calculated as:
wherein ratio is j Represents the contribution degree duty ratio of the j-th node, K represents the number of nodes of the second participating node, score j And the calculated force contribution degree of the j-th node is represented.
In some embodiments of the present disclosure, obtaining a digital networked computing power resource panorama from each hub node comprises: collecting computing power resource information of a plurality of participating nodes in each subnet by a hub node in the subnet; all hub nodes in the Internet of things forward the collected computing power resource information to each other so that each hub node forms a computing power resource panorama of the Internet of things.
In some embodiments of the present disclosure, determining, in real-time from the candidate participating nodes, a target computing power lessor matching the computing power resource buyer according to the lessor panorama comprises: in response to the first hub node receiving the plurality of computing power resource buyers, determining a target computing power lessor matching each computing power resource buyer by jointly solving:
wherein x is ij Indicating whether the j candidate participating node is selected by the i-th task, x ij =1 indicates selected, x ij =0 denotes unselected, c ij Cost of computing power resource allocation to ith task of jth candidate participating node, d i Representing the total computational resource demand, w, of an ith task ij Representing the amount of computational power resources allocated to the ith task by the jth candidate participating node, R j And the total leasing power resource of the j candidate participating node is represented, and the i-th task is a task corresponding to the buying order of the i-th computing power resource.
In some embodiments of the present disclosure, the power resource buying order comprises: price, quantity, performance requirements, task descriptions for the desired computational resources.
In some embodiments of the present disclosure, the power rental information includes a power resource sales order. The sales order of the computing power resources comprises: description of price, quantity, performance, executable tasks for the computing power resources to rent.
According to a second aspect of the present disclosure, there is provided a computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the steps of the algorithm matching method according to the first aspect of the present disclosure.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the following brief description of the drawings of the embodiments will be given, it being understood that the drawings described below relate only to some embodiments of the present disclosure, not to limitations of the present disclosure, in which:
FIG. 1 is a schematic topology of a digital network;
fig. 2 is a schematic flow diagram of a method of computing power matchmaking for digital networking according to an embodiment of the present disclosure.
It is noted that the elements in the drawings are schematic and are not drawn to scale.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present disclosure more apparent, the technical solutions of the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings. It will be apparent that the described embodiments are some, but not all, of the embodiments of the present disclosure. All other embodiments, which can be made by those skilled in the art based on the described embodiments of the present disclosure without the need for creative efforts, are also within the scope of the protection of the present disclosure.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the presently disclosed subject matter belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the specification and relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein. In addition, terms such as "first" and "second" are used merely to distinguish one component (or portion of a component) from another component (or another portion of a component).
Fig. 1 shows a schematic topology of a digital network. The digital network may include a plurality of subnets 10. Each sub-network 10 comprises a hub node 11 and a plurality of participating nodes 12 directly connected to the hub node. The hub nodes 11 in the plurality of subnetworks 10 are directly connected to each other. The hub node 11 and the hub node 11 may be interconnected by a private network. The hub node 11 performs functions such as information aggregation, address navigation, etc. on the participating nodes 12. The participating nodes 12 may be various government bodies, industry bodies, corporate bodies, institutional bodies, and the like. The participating nodes 12 directly connected to the same hub node 11 communicate through the hub node 11. The participating nodes 12 directly connected to the different hub nodes 11 communicate through their respective directly connected hub nodes 11. That is, the participating nodes 12 only directly communicate with the hub nodes 11 directly connected thereto, the hub nodes 11 can directly communicate with each other, and the participating nodes 12 need to communicate with each other via the corresponding hub nodes 11.
In practice, there may be a large number of subnets 10 in the digital network. There may be a large number of participating nodes 12 in a single subnet 10. Thus, the number of participating nodes 12 in the digital network can be very large. When the local computing tasks of some participating nodes 12 are finished, the computing resources of the participating nodes 12 have a certain idle state, which results in the waste of the computing resources. If the participating nodes 12 lease idle computing power resources to the participating nodes 12 requiring additional computing power, the participating nodes 12 requiring additional computing power do not need to purchase additional hardware devices, and can complete computing tasks by only leasing idle computing power resources in the digital network, so that the whole computing power resources of the digital network can be well integrated and utilized.
The demand for additional computing power by a computing power renter is typically real-time, and thus there is a need to quickly find available unused computing power resources to assist the computing power renter in completing a computing task. While the distribution of idle computing resources may change in real-time as the local computing tasks of participating nodes 12 change. Therefore, in the digital networking, timeliness needs to be considered when matching the computing power of a large number of participating nodes 12.
The disclosure provides a computing force matching method for digital networking. Fig. 2 shows a schematic flow diagram of a computing force matching method 200 for digital networking according to an embodiment of the disclosure.
At block S202 of fig. 2, each hub node 11 obtains a digital networked computing resource panorama. The computing power resource panorama includes computing power resource information for all participating nodes 12 in the digital network. The computing power resource information may include: computing resources and storage resources, such as processors, memory, hard disk storage, graphics processors, distributed computing resources, server clusters, intranet bandwidth, and other configuration information and model numbers. The participating nodes 12 in the digital network are typically heterogeneous, i.e. the computing power capabilities provided by the different participating nodes 12 are different. Some of the participating nodes 12 are capable of providing distributed computing resources and some of the participating nodes 12 are capable of providing only stand-alone computing resources. Different levels of computing resource provisioning can affect the computational efficiency and cost settlement after subsequent combinations of computing forces.
In some embodiments of the present disclosure, the computational resource information of the plurality of participating nodes 12 in each sub-network 10 is collected by the hub node 11 in that sub-network 10. In one example, each participating node 12 may actively report the computational resource information to its directly connected hub node 11 when registering to join the digital network. In another example, each participating node 12 generates a local resource configuration information table and then passively waits for the hub node 11 to collect the resource configuration information table. After the hub node 11 collects Ji Tongyi the computing resource information of all the participating nodes 12 in the subnet 10, the collected computing resource information may be forwarded to other hub nodes. In one example, all hub nodes 11 in the digital network forward the collected computing power resource information to each other such that each hub node 11 individually forms a digital network computing power resource panorama. In another example, one master hub node may be elected from all hub nodes 11 in the digital network. The other hub nodes forward the collected computing power resource information to the primary hub node. The main hub node forms a digital networking computing power resource panorama according to the computing power resource information collected by each hub node 11, and then forwards the computing power resource panorama to other hub nodes.
The computational resources of the participating nodes 12 depend on their hardware configuration, and thus the computational resources of each participating node 12 may be considered relatively static (as opposed to being fixed with respect to the time required to complete a computational power match), and thus the computational resource panorama may be considered relatively static.
When the computational resources of the participating node 12 change, the participating node 12 may actively report new computational resource information to the hub node 11 to which it is directly connected, so that the hub node 11 updates the computational resource panorama and synchronizes this change to other hub nodes.
At block S204, each hub node 11 obtains a real-time updated rental power panorama for the digital network. The rental forces panorama includes all of the computerized rental information published by the participating nodes 12 in the digital network as the computerized renters. In some embodiments of the present disclosure, the power rental information includes a power resource sales order. The sales order for the computing resources may include: description of price, quantity, performance, executable tasks for the computing power resources to rent.
In some embodiments of the present disclosure, the power lease information published by the plurality of participating nodes 12 in each subnet 10 is collected by the hub node 11 in that subnet 10. The hub node 11 may collect the power rental information in the same manner as the power resource information. In one example, all hub nodes 11 in the digital network forward the collected power rental information to each other such that each hub node 11 individually forms a digital network power rental panorama. In another example, one master hub node may be elected from all hub nodes 11 in the digital network. The other hub nodes forward the collected power lease information to the primary hub node. The main hub node forms a rental power panorama of the Internet of numbers according to the collected power renting information of each hub node 11, and then forwards the rental power panorama to other hub nodes.
If the first participating node (the first participating node may be any participating node 12 in the digital network) has updated the power lease information, the first participating node immediately issues updated power lease information to its directly connected designated hub node, the designated hub node uses the updated power lease information of the first participating node to update the power panorama in real time, and the designated hub node forwards the updated power lease information of the first participating node to other hub nodes in the digital network so that the other hub nodes update the power panorama in real time.
In some embodiments of the present disclosure, each participating node 12 may determine idle computing resources (also interchangeably referred to herein as "idle computing resources" in the context) by dynamically analyzing the predictions. Each participating node 12 may predict the resource demand of the participating node 12 itself for a period of time in the future based on the historical computational power resource overhead of the participating node 12. The participating node 12 determines idle computing resources of the participating node 12 for a period of time in the future based on its computing resources and the predicted resource demand. In one example, the idle computing resources may be equal to the computing resources of the participating node 12 minus the predicted resource demand. In another example, to ensure that local computing tasks can be performed in time, the idle computing resources may be equal to the computing resources of the participating node 12 minus the sum of the predicted resource requirements and the preset redundancy value. Thus, even when the computational power resources required for a local computational task vary slightly, there are sufficient computational power resources to support the computational task.
The participating nodes 12 may isolate the idle computing resources in the form of containers. Container technology is widely used in cloud computing environments as a lightweight virtualization technology. The container technology can realize quick deployment, resource isolation and elastic expansion of application programs, and greatly improves portability and flexibility of the application. The participating nodes 12 may perform container isolation assessment to assess the degree of isolation between containers, including file system isolation, process isolation, network isolation, and the like. In assessing the isolation of containers, it is possible to observe whether other containers are affected by testing the degree of interaction and interference between the containers, such as by performing a network scan in one container. The participating nodes 12 may monitor and analyze the container in real-time by logging system logs and behavior logs of the container, and applying security audit and monitoring tools, monitor the operational status of the container, detect container vulnerabilities and abnormal behaviors, discover and respond to security events in time, and the like. The participating nodes 12 may evaluate the security of the container image, including the trustworthiness of the image source, verification of the image signature, vulnerability and malware detection of the image, and so forth. The participating nodes 12 may use image scanning tools to perform vulnerability scanning and malware detection on the container images to ensure security of the container images. The participating nodes 12 may additionally evaluate the security of the container network, including the security of communications between containers and the security of the connection of the containers to external networks. The participating nodes 12 may detect and prevent potential network attacks and intrusions by means of network traffic monitoring and firewall rule verification, etc. Thus, by employing container techniques, the participating nodes 12 are able to safely and efficiently isolate the idle computing resources so that other participating nodes 12 use the idle computing resources.
If the predicted resource demand by the participating node 12 increases, the participating node 12 may release unused resources in the container. If the resource demand predicted by the participating node 12 decreases, the participating node 12 may increase the resources in the container. In this way, the participating nodes 12 can flexibly configure the idle computing resources, ensuring that local computing tasks can be performed in time and that local idle computing resources are not idle.
In other embodiments of the present disclosure, each participating node 12 may determine idle force resources by way of factual monitoring. Each participating node 12 monitors the usage of the participating node's 12 computing resources in real-time to determine the participating node's 12 real-time idle computing resources. The usage may include allocation of computing resources, usage, and idle time. In one example, resources such as a Central Processing Unit (CPU) and a memory can be monitored in real time by using a monitoring tool, various resource use conditions of a server can be recorded, a flexible data model and a query language can be provided, and the resource use conditions such as the CPU and the memory can be monitored and analyzed. In addition, on the basis of the monitoring tool, an efficient resource management system can be developed to track and manage the computing power resources, including tracking the distribution condition, the utilization rate and the idle time of the computing power resources, and determining which resources and how many resources are in an idle state by analyzing the data of the resource management system.
The participating nodes 12 may isolate the idle computing resources in the form of containers. If the real-time usage of the computational resources of the participating node 12 increases, the participating node 12 may immediately release unused resources in the container. If the real-time usage of the computational resources of the participating node 12 decreases, the participating node 12 may immediately increase the resources in the container. In this way, the participating nodes 12 can flexibly configure the idle computing resources, ensuring that local computing tasks can be performed in time and that local idle computing resources are not idle.
In some embodiments of the present disclosure, a participating node 12 that is a computing force renter may submit a computing force resource purchase order to its directly connected hub node 11. The power resource purchase order may include: price, quantity, performance requirements, task descriptions for the desired computational resources.
If the first hub node in the first subnet receives a power resource purchase order submitted by a participating node 12 in the first subnet as a power renter ("yes" at block S206), the first hub node may perform operations at blocks S208 through S212. In this context, the first subnet refers to any one of the subnets 10 in the digital network. The first hub node refers to the hub node 11 in any one of the sub-networks 10.
At block S208, the first hub node determines candidate participating nodes in the digital network having computing power resources matching the computing power resource purchase order from the computing power resource panorama. As mentioned above, the computing resources panorama is relatively static and does not change much. Therefore, the computing power resource panorama can be used for preliminary screening, the participating nodes 12 with computing power resources matched with the computing power resource buying list are determined as candidate participating nodes, and the participating nodes 12 without computing power resources matched with the computing power resource buying list are eliminated. If none of the total computational resources of a participating node 12 is sufficient to match a computational resource purchase order, then it is more difficult for the idle computational resources of the participating node 12 to match the computational resource purchase order. For the digital networking with a large number of participating nodes 12, this step can eliminate some impossible participating nodes 12 in advance, thereby reducing the operand of subsequent operations.
At block S210, the first hub node determines, from the candidate participating nodes, a target computing power lessor matching the computing power resource buyer in real-time according to the lessor panorama. The first hub node can comprehensively consider the price of the computing power resource buying list and available resources for matching according to the bid and the resource requirement. The matching algorithm may take the price into account preferentially, then take available resources into account again, and may also perform matching according to the resource performance priority principle. The computing power renter can set a matching priority in the computing power resource ticket.
The matching system based on the bidding and the resource matching needs flexible algorithms and rules to balance the priorities of the bidding and the resource, comprehensively considers factors such as market demands, resource management, matching algorithms, transaction execution and the like, and achieves fair, efficient and optimized transaction matching.
In some embodiments of the present disclosure, the first hub node may receive a plurality of computational power resource buyers from a plurality of participating nodes. Each power resource purchase order corresponds to 1 task. The target power lessors that match each power resource ticket may be determined by combining the solutions:
wherein x is ij X represents whether the j-th candidate participating node is selected by the i-th task (i.e., whether the j-th candidate participating node is the target power lessor of the i-th task) ij =1 indicates selected, x ij =0 denotes unselected, c ij Cost of computing power resource allocation to ith task of jth candidate participating node, d i Representing the total computational resource demand, w, of an ith task ij Representing the amount of computational power resources allocated to the ith task by the jth candidate participating node, R j And the total leasing power resource of the j candidate participating node is represented, and the i-th task is a task corresponding to the buying order of the i-th computing power resource. I is more than or equal to 1 and less than or equal top, 1.ltoreq.j.ltoreq.q, p representing the total number of the computational power resource buyers received by the first hub node, q representing the total number of candidate participating nodes.
min∑(c ij ×x ij ) For minimizing the total cost of the computer rental. Sigma (w) ij ×x ij )≥d i The computational power resource requirements for each task are guaranteed to be met. Sigma (w) ij ×x ij )≤R j The total lease power resource for each candidate participating node is not exceeded.
At block S212, the first hub node establishes a power lease relationship with a target hub node to which the target power lease is directly connected by signing up. In some embodiments of the present disclosure, the target hub node immediately freezes the computing power resource leased by the target computing power lessor upon completion of the subscription with the first hub node, in order to avoid allocation of the computing power resource to other computing power lessors.
After signing, the power lessor executes the current transaction and performs power resource allocation according to the transaction result. Execution of the transaction may include transferring funds, transferring the right to use the computing asset, and recording completion of the transaction. Blockchain-based ledger administration may be performed. The first hub node and the target hub node may maintain an order ledger based on blockchain technology, bill the transaction information of the transaction (the order ledger cannot be tampered), and transfer the temporary calculation power usage right until the transaction is completed. The computing resource allocation process may involve computing inventory adjustments, funds transfer, or reallocation of other resources.
In some embodiments of the present disclosure, each computing power lessor isolates idle computing power resources to be leased in the form of containers. If the number of target power renters is greater than 1, the containers in all target power renters are grouped into a container cluster (the container cluster may be composed of multiple hosts, each host running multiple containers on which images may be deployed using a container orchestration tool), and one primary hub node is elected from among the target hub nodes to which each target power renter is directly connected. The primary hub node monitors the load conditions of the container clusters in real time during execution of instructions by the computing force renter (during the transaction). If the load of the container cluster is reduced, the primary hub node scales down the container cluster, stops the container instances that are no longer needed, and releases the computing power resources to reduce costs. If the load of the container cluster increases, the primary hub node expands the size of the container cluster. This may enable the container orchestration tool to automatically launch new container instances on available hosts by increasing the number of container instances. Thus, the elastic integration of the computing power resources can be realized.
In some embodiments of the present disclosure, the container is restarted in the event that the container fails or is unavailable, ensuring the availability of the application. If the number of target power renters is equal to 1, the target hub node to which the target power renter is directly connected collects and manages log data of the containers in the target power renter, and restarts the containers in case the containers fail or are not available. If the number of targeted power renters is greater than 1, the primary hub node collects and manages log data for each container in the container cluster and restarts the containers in the event that any of the containers fails or is unavailable. The container logs are collected and managed in a centralized mode, and fault removal, monitoring and analysis can be conveniently conducted. The log data may help identify performance problems, abnormal behavior, potential faults, and the like.
By using container technology in combination with container orchestration tools and automated management systems, a container-based elastic integration effort can be achieved. The method can rapidly and flexibly expand and reduce the computational power resources, improve the usability and the performance of the application program and reduce the waste of the computational power resources. Meanwhile, due to the lightweight and portability of the container, the containerized application program can be deployed and operated in different environments, so that the cross-platform elastic integration force is realized.
In some embodiments of the present disclosure, the hub node 11 may also perform computational resource monitoring. The computing power resource monitoring refers to real-time monitoring and recording of the use condition of computing resources so as to perform performance analysis, resource optimization, fault removal and other operations. In the context of elastic integration of computing forces, computing force resource monitoring may help to know the load condition of a container cluster or server cluster in real time.
The computing power resource monitoring metrics may include: CPU utilization, memory consumption, network traffic, storage space usage, container or process status. The monitoring of CPU usage can be used to understand the load condition of the computing resources. Monitoring the memory usage (memory consumption) of computing resources ensures that sufficient memory is available for use by applications. Monitoring network traffic (including ingress traffic and egress traffic) of computing resources can optimize network performance and bandwidth requirements. Monitoring the storage space usage of a computing resource may include the space usage and available space of a hard disk or storage volume. Monitoring the state of a container or process (including running state, start time, stop time, etc.) can facilitate timely discovery of anomalies and failures.
To enable computing resource monitoring, various monitoring tools and platforms may be used. These tools can collect, store and present monitoring data.
Blockchain-based certification combines the storage and verification of data with transaction records on the blockchain using the non-tamper-evident and de-centralised features of blockchain technology. Blockchain-based certification can ensure the integrity, credibility and invariance of the data so that the source and content of the data can be effectively verified and traced.
During the validation process, a digest or hash value of the data is typically calculated and stored in the transaction on the blockchain. Any attempt to alter or tamper with the data will change the digest of the data, leaving a trace on the blockchain, allowing the integrity of the data to be verified. In addition, the time stamp of the certification may also be confirmed by the block time of the blockchain.
To improve engagement of the engaged node 12 during a computational effort match, some embodiments of the present disclosure also present concepts of computational effort contribution. The computational effort contribution degree refers to the contribution degree of each participating node 12 to the computing power participating in the computing task in the computational effort sharing network. The computational effort contribution is measured in terms of the amount and quality of computational resources provided by the participating nodes 12.
In some embodiments of the present disclosure, the second hub node calculates the calculated force contribution of all second participating nodes directly connected to the second hub node. The second hub node is referred to herein as any one of the hub nodes 11 in the digital network. The second participating node refers to the participating node 12 directly connected to the arbitrary hub node 11. The second hub node calculates a contribution duty cycle of each second participating node according to the calculated force contribution of the second participating node. And then, the second hub node distributes rewards to each second participation node according to the contribution ratio of the second participation node so as to improve the participation of the second participation node in the process of computing the power match.
In one example, the calculated force contribution of the j-th node of the second participating nodes is calculated as:
wherein score j Representing the calculation force contribution degree of the j-th node, R ai Representing the number, W, of ith computational resource in the n computational resources of the jth node rai Representing weights set according to the class of the ith computational resource (which may be specifically set according to the actual application), R qi An ith quality parameter, W, of m quality parameters representing a jth node rqi Represents the weight set for the ith quality parameter, W rt Representing weights determined from time of leasing power of jth node, W rc Representing the weight determined according to the completion rate of the task in which the j-th node participates.
The computing resources mainly refer to a CPU, a GPU, storage, bandwidth and the like. W (W) rai The value of (c) can be specifically set according to practical application. The quality parameters may include performance metrics and reliability metrics. Performance metrics may include response time, throughput, task processing power, concurrent processing power, and the like. The reliability index may include availability, fault tolerance, reliability, and fault handling capability. Participants who participate in and provide computing resources for a long period of time are more contributing. Thus the longer the participating node 12 rents out the computing power, the more W rt The larger. In one example, any of the times approximately 5The sum of the time spent on the business accounts for the proportion of the total time spent on the business to calculate the time to lease the power of the participating nodes 12. The completion rate of the task in which the node participates may refer to a task completion rate of approximately 5 times. The higher the completion rate of a certain participating node 12 in the recent history task, the more the calculated contribution degree can be added. Default W when first engaged in task rc With a value of 0, if the completion rate is lower than 0.5, W rc -0.5; otherwise W rc 0.5.
In one example, the contribution duty cycle of the jth node is calculated as:
Wherein ratio is j Represents the contribution degree duty ratio of the j-th node, K represents the number of nodes of the second participating node, score j And the calculated force contribution degree of the j-th node is represented.
The contribution duty cycle of the participating nodes 12 is related to the rewarding mechanism. The higher the contribution of the participating nodes 12, the more rewards will be obtained. For example, the higher the contribution of a participating node 12, the participating node 12 may preferentially match a computational resource buy order or a computational resource sell order, or the participating node 12 may preferentially match a more priced computational resource buy order or computational resource sell order. This rewarding mechanism may motivate the participating nodes 12 to provide more computing resources, thereby maintaining the stability and security of the network, creating a virtuous circle of the internet of things transaction ecology, helping the healthy operation of the internet of things.
In addition, in the process of carrying out the power transaction between the power renter and the target power renter, the communication between the power renter and the first hub node, the first hub node and the target hub node, and the communication between the target hub node and the target power renter all follow the principle of safe calculation, and privacy protection calculation is carried out on data involved in the power transaction.
In other embodiments of the present disclosure, there is also provided a computer readable storage medium storing a computer program, wherein the computer program is capable of implementing the steps of the method as shown in fig. 2 when being executed by a processor.
In summary, according to the method for matching the computing power of the digital networking according to the embodiment of the disclosure, idle computing power resources in the digital networking can be found, efficient and elastic integration of the idle computing power resources is performed, computing power matching is rapidly and accurately achieved in the digital networking with massive participating nodes 12, and computing power resources of a plurality of participating nodes 12 can be efficiently combined to meet the requirement of single computing power resource buying list under the condition that the requirement of single computing power resource buying list is large. In addition, the power matching method for the digital networking according to the embodiment of the disclosure sets an encouraging mechanism based on the power contribution degree, so that the participation degree of the participation node 12 in the power matching process is improved, the stability and the safety of the network are maintained, the digital networking transaction ecology of virtuous circle is formed, and the digital networking healthy operation is helped.
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus and methods according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
As used herein and in the appended claims, the singular forms of words include the plural and vice versa, unless the context clearly dictates otherwise. Thus, when referring to the singular, the plural of the corresponding term is generally included. Similarly, the terms "comprising" and "including" are to be construed as being inclusive rather than exclusive. Likewise, the terms "comprising" and "or" should be interpreted as inclusive, unless such an interpretation is expressly prohibited herein. Where the term "example" is used herein, particularly when it follows a set of terms, the "example" is merely exemplary and illustrative and should not be considered exclusive or broad.
Further aspects and scope of applicability will become apparent from the description provided herein. It should be understood that various aspects of the present application may be implemented alone or in combination with one or more other aspects. It should also be understood that the description and specific examples are intended for purposes of illustration only and are not intended to limit the scope of the present disclosure.
While several embodiments of the present disclosure have been described in detail, it will be apparent to those skilled in the art that various modifications and variations can be made to the embodiments of the present disclosure without departing from the spirit and scope of the disclosure. The scope of the present disclosure is defined by the appended claims.
Claims (10)
1. A power calculation matching method for a digital network, wherein the digital network comprises a plurality of subnets, each subnet comprises a hub node and a plurality of participation nodes directly connected with the hub node, the hub nodes in the subnets are directly connected with each other, and the power calculation matching method comprises the following steps:
each hub node obtains a computing power resource panorama of the digital network, wherein the computing power resource panorama comprises computing power resource information of all participating nodes in the digital network;
each hub node obtains a renting power panorama updated in real time by the digital network, wherein the renting power panorama comprises all the information of the renting power issued by the participating nodes serving as the renting parties in the digital network;
in response to a first hub node in a first subnet receiving a power resource purchase order submitted by a participating node in the first subnet as a power lease, the first hub node performs the following operations:
determining candidate participation nodes with computing power resources matched with the computing power resource buying list in the digital network according to the computing power resource panorama;
determining a target computing power renter matching the computing power resource buyer from the candidate participating nodes in real time according to the renting computing power panorama; and
And the target hub node directly connected with the target power calculation renter constructs a power calculation renting relation in a signing mode.
2. The method of computing power matching of claim 1, further comprising:
each computing power renter isolates idle computing power resources to be rented in a container mode;
in response to the number of target power lessors being greater than 1, grouping containers in all target power lessors into a container cluster;
selecting a main hub node from target hub nodes directly connected by each target computing lessor;
during the execution of the instructions of the computing power renter, the main hub node monitors the load condition of the container cluster in real time;
responsive to the load reduction, the primary hub node downscaling the container cluster; and
in response to the load increase, the primary hub node expands the size of the container cluster.
3. The method of computing power matching of claim 2, further comprising:
in response to the number of target power renters being equal to 1, the target hub node to which the target power renters are directly connected collecting and managing log data of containers in the target power renters, and restarting the containers if the containers fail or are unavailable; and
In response to the number of targeted power renters being greater than 1, the primary hub node collects and manages log data for individual containers in the cluster of containers and restarts the containers if any of the containers fails or is unavailable.
4. The computing power matching method of claim 1, wherein each hub node obtaining the internet-of-numbers real-time updated taxi power panorama comprises:
collecting, by the hub node in each sub-network, the calculated lease information issued by the plurality of participating nodes in the sub-network;
all hub nodes in the digital network mutually forward the collected calculation power renting information so that each hub node respectively forms a renting power panorama of the digital network;
in response to the first participating node's calculated rental information having been updated, the first participating node immediately issues updated calculated rental information to its directly connected designated hub node, the designated hub node uses the first participating node's updated calculated rental information to update the rental panorama in real time, and the designated hub node forwards the first participating node's updated calculated rental information to other hub nodes in the digital network for the other hub nodes to update the rental panorama in real time.
5. The method of computing power matching according to any one of claims 1 to 4, further comprising each participating node:
predicting the resource demand of the participating node in a future period according to the historical computing power resource expense of the participating node;
determining idle computing power resources of the participating node in a period of time in the future according to computing power resources of the participating node and the predicted resource requirements;
the participating node isolates the idle computing resources in the form of a container;
releasing unused resources in the container in response to the predicted increase in resource demand; and
in response to the predicted decrease in resource demand, the resources in the container are increased.
6. The method of computing power matching according to any one of claims 1 to 4, further comprising each participating node:
the method comprises the steps of monitoring the use condition of the computing power resource of the participating node in real time to determine the idle computing power resource of the participating node in real time, wherein the use condition comprises the distribution condition, the use rate and the idle time length of the computing power resource;
the participating node isolates the idle computing resources in the form of a container;
Responsive to an increase in the real-time usage of the computational resources of the participating node, immediately releasing unused resources in the container; and
in response to the reduction in the real-time usage of the computational resources of the participating node, the resources in the container are immediately increased.
7. The method of computing force matching according to any one of claims 1 to 4, further comprising: and the target hub node immediately freezes the leased computing resources of the target computing power leasing party when the target hub node completes the subscription with the first hub node.
8. The method of computing force matching according to any one of claims 1 to 4, further comprising:
the second hub node calculates the calculation force contribution degree of all second participation nodes directly connected with the second hub node;
the second hub node calculates the contribution degree duty ratio of each second participation node according to the calculated force contribution degree of the second participation node; and
the second hub node distributes rewards to each second participation node according to the contribution ratio of the second participation node so as to improve the participation degree of the second participation node in the process of computing the power match;
Wherein the calculated force contribution of a j-th node of the second participating nodes is calculated as:
wherein score j Representing the calculated force contribution degree of the j-th node, R ai Representing the number, W, of ith computational resource of the n computational resources of the jth node rai Representing weights set according to the class of the ith computing power resource, R qi An ith quality parameter, W, of m quality parameters representing the jth node rqi Represents the weight, W, set for the ith quality parameter rt Representing weights determined from the time of the j-th node leasing power, W rc Representing a weight determined according to the completion rate of the task in which the j-th node participates;
wherein the contribution duty cycle of the j-th node is calculated as:
wherein ratio is j Representing the contribution degree duty ratio of the j-th node, K representing the number of nodes of the second participating node, score j And representing the calculated force contribution degree of the j-th node.
9. The computing power match method of any one of claims 1 to 4, wherein each hub node obtaining the digital networking computing power resource panorama comprises:
collecting, by the hub node in each sub-network, computing power resource information for the plurality of participating nodes in the sub-network;
All hub nodes in the digital network forward the collected computing power resource information to each other so that each hub node forms a computing power resource panorama of the digital network.
10. The method of computing power matching of any one of claims 1 to 4, wherein determining, from the candidate participating nodes, in real-time from the rental forces panorama, a target computing force renter that matches the computing force resource ticket comprises:
in response to the first hub node receiving a plurality of computing power resource buyers, determining a target computing power lessor matching each computing power resource buyer by jointly solving:
wherein x is ij Indicating whether the j candidate participating node is selected by the i-th task, x ij =1 indicates selected, x ij =0 denotes unselected, c ij Representing the cost of allocation of computational resources of the j-th candidate participating node to the i-th task, d i Representing the total computational power resource requirements, w, of the ith task ij Representing the amount of computational power resources allocated to the ith task by the jth candidate participating node, R j And the j candidate participating nodes are indicated to be total leased computing power resources, and the i task is a task corresponding to the buying list of the i computing power resources.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311427995.5A CN117579629A (en) | 2023-10-31 | 2023-10-31 | Calculation force matching method for digital networking |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311427995.5A CN117579629A (en) | 2023-10-31 | 2023-10-31 | Calculation force matching method for digital networking |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117579629A true CN117579629A (en) | 2024-02-20 |
Family
ID=89888981
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311427995.5A Pending CN117579629A (en) | 2023-10-31 | 2023-10-31 | Calculation force matching method for digital networking |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117579629A (en) |
-
2023
- 2023-10-31 CN CN202311427995.5A patent/CN117579629A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10841241B2 (en) | Intelligent placement within a data center | |
| Kim et al. | CometCloud: An autonomic cloud engine | |
| US10769687B2 (en) | Migrating computing environment entitlement contracts between a seller and a buyer | |
| US20130151712A1 (en) | Managing Organizational Computing Resources in Accordance with Computing Environment Entitlement Contracts | |
| US7894357B2 (en) | Capability-based testing and evaluation of network performance | |
| US8812679B2 (en) | Managing computing environment entitlement contracts and associated resources using cohorting | |
| Wen et al. | Quality model for evaluating SaaS service | |
| CN108733509A (en) | Method and system for data to be backed up and restored in group system | |
| CN104679591A (en) | Method and device for distributing resource in cloud environment | |
| Heilig et al. | Location-aware brokering for consumers in multi-cloud computing environments | |
| CN112997469A (en) | Intelligent, decentralized and autonomous marketplace for distributed computing and storage | |
| Thakur et al. | A robust reputation management mechanism in the federated cloud | |
| Eyckerman et al. | Requirements for distributed task placement in the fog | |
| Mahato et al. | Load balanced scheduling and reliability modeling of grid transaction processing system using colored Petri nets | |
| Andrade et al. | Performance and availability trade-offs in fog–cloud iot environments | |
| Melo et al. | Distributed application provisioning over Ethereum-based private and permissioned blockchain: availability modeling, capacity, and costs planning | |
| Kim et al. | Online risk analytics on the cloud | |
| Song et al. | Smart contract-based computing resources trading in edge computing | |
| CN105827744A (en) | Data processing method of cloud storage platform | |
| Mills et al. | Can economics-based resource allocation prove effective in a computation marketplace? | |
| CN117579629A (en) | Calculation force matching method for digital networking | |
| Burakowski et al. | Traffic Management for Cloud Federation. | |
| Correia et al. | Blockchain as a service environment: a dependability evaluation | |
| Nzanzu et al. | Monitoring and resource management taxonomy in interconnected cloud infrastructures: a survey | |
| Huang et al. | Kubebench: A benchmarking platform for ml workloads |
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 |