Optical data center network-oriented algorithm network fusion protection method and storage medium

Abstract

The application relates to the technical field of resource allocation, and discloses an all-optical data center network-oriented algorithm network fusion protection method and a storage medium, which comprises the following steps: mapping an all-optical data center network to obtain virtual servers, virtual wavelength division optical switching devices and virtual links, and obtaining a corresponding computing power wave plane list; selecting target working servers and current working paths for to-be-transmitted services based on the computing power wave plane list; based on backup working servers and backup paths of services that have been successfully deployed in time periods that are time-crossed with actual transmission time periods, constructing a shared protection server list and a shared protection path list, and selecting backup working servers and backup paths for to-be-transmitted services; the application adopts a shared protection strategy at the path and computing power resource level, simultaneously considers the survivability of servers and paths, and when a server or a path fails, the backup working servers and the backup paths can quickly take over tasks, reduce downtime, and ensure service survivability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of resource allocation, in particular to a network and computing integrated protection method for an all-optical data center network and a storage medium. BACKGROUND

[0002] With the popularization and application of computing-intensive businesses such as meta-universe, AR / VR and machine learning, data centers need to guarantee that such businesses can use computing resources on demand, which also puts higher requirements on the computing resource scheduling capability of data centers. In terms of computing resource scheduling, traditional networks use a communication and computing separation framework design and only serve as a pipeline for information transmission, which cannot perceive the needs of upper-layer business applications and thus cannot coordinate to guarantee computing resource scheduling needs.

[0003] At present, according to application technology, data center networks can be divided into electrical switching, optical-electric hybrid and all-optical data center networks. Compared with the three kinds of data center networks, all-optical data center networks avoid the electronic bottleneck in electrical switching technology, can upgrade electrical switching and optical-electric hybrid data center networks, and can be completely built by optical switching, with the advantages of low energy consumption, high bandwidth and low latency.

[0004] In an all-optical data center network, once key components such as ports, optical switches, links and computing servers fail during business scheduling, serious problems such as business interruption and even loss of business data will be caused. At present, research on survivability is generally based on a single failure model. For link failure, a strategy of allocating two disjoint paths for each optical path is often used to ensure 100% survivability. Among these strategies, 1+1 protection and SBPP (Shared Backup Path Protection) protection mechanisms are particularly common, which mainly focus on the impact of path breakage on business operation. 1+1 protection is a typical protection mechanism, in addition to the working link, it also sets up a standby fiber link; when data is transmitted, it is simultaneously sent to the main link and the standby link; at the receiving end, the system detects the data of the main link and the standby link, selects the correct data packet and discards the damaged data packet to ensure the continuity of data transmission. And the protection link of each business cannot be shared. The shared backup path protection SBPP uses a resource sharing method, by setting up a standby link shared by multiple businesses; when a main link fails, the standby link will be activated to take over the data transmission of the link; because the standby link is shared, it is only used when a fault occurs, so the bandwidth resource utilization is higher and the cost is lower.

[0005] However, in the survivability research of high-performance computing services, if a server node fails, such a simple path protection mechanism is not enough. When a server node fails, the following problems may occur: ① Service interruption: the service on the server node will be immediately interrupted and cannot process requests or perform computing tasks; ② Data loss: if there is no backup data on the failed server, data loss may occur; ongoing transactions or data processing may be interrupted, resulting in inconsistent or lost data; ③ Business continuity: affecting business continuity and reliability. Therefore, only considering path protection cannot completely guarantee the survivability of the service, and the protection of the full-optical data center "path + computing node" must be considered to more comprehensively guarantee the survivability of AI services. SUMMARY

[0006] To this end, the technical problem to be solved by the present application is to overcome the problem of low service terminal, data loss and low service survivability caused by only considering path protection in the prior art when a server node fails.

[0007] To solve the above technical problems, the present application provides a method for protecting the fusion of computing and networking for a full-optical data center network, comprising:

[0008] Constructing a full-optical data center network composed of multiple servers and multiple wavelength division optical switching devices;

[0009] Obtaining the source server, the alternative source server, the computing resources required by the service, the time required for service transmission, and the start time of the service to be transmitted;

[0010] Based on the start time of the service to be transmitted and the time required for service transmission, the actual transmission period of the service to be transmitted is obtained;

[0011] Mapping the servers and wavelength division optical switching devices in the full-optical data center network to virtual servers and virtual wavelength division optical switching devices, respectively; mapping the physical links with available wavelengths in the actual transmission period to virtual links; based on the virtual servers and their computing resources, the virtual wavelength division optical switching devices and their wavelength connection tables, and the virtual links and their occupancy configuration information, constructing an algorithmic wave plane list of the full-optical data center network in the actual transmission period;

[0012] Based on the algorithmic wave plane list, a plurality of alternative servers that meet the computing resources required by the service to be transmitted are obtained; for each alternative server, the shortest route between it and the source server of the service to be transmitted is obtained; based on the computing resources of each alternative server and the shortest route, and the reconfiguration number of the wavelength division optical switching devices on the shortest route, a target working server is selected, and the target wavelength and the current working path from the source server to the target working server are obtained;

[0013] Based on the standby working servers and standby paths of the successfully deployed services in the time period intersecting with the actual transmission time period, a shared protection server list and a shared protection path list are constructed;

[0014] It is judged whether there is a protection server in the shared protection server list, and a protection path from the candidate source server of the to-be-transmitted service to the protection server and not intersecting with the current working path can be obtained in the shared protection path list:

[0015] If there is, the protection server is obtained as the standby working server of the to-be-transmitted service; and based on the protection path corresponding to the protection server, a standby path of the to-be-transmitted service is constructed;

[0016] If not, the shortest paths from the candidate source server of the to-be-transmitted service to each protection server in the shared protection server list are calculated respectively, and the shortest path not intersecting with the current working path is selected from all the shortest paths as the protection path and added to the shared protection path list; based on the protection path, a standby path of the to-be-transmitted service is constructed, and the standby path and the protection server corresponding to the standby path are used as the standby path and the standby working server of the to-be-transmitted service;

[0017] Based on the target working server, the current working path, the standby working server and the standby path of the to-be-transmitted service, the to-be-transmitted service is transmitted.

[0018] Preferably, an algorithmic power wave plane list of the all-optical data center network in the actual transmission time period is constructed, comprising:

[0019] The servers in the all-optical data center network are mapped as virtual servers, and the computing resources of each virtual server in the actual transmission time period are recorded in a server resource list;

[0020] The wavelength division optical switching devices in the all-optical data center network are mapped as virtual wavelength division optical switching devices, and a wavelength connection table corresponding to each virtual wavelength division optical switching device is created based on the wavelength connection information of each port in each virtual wavelength division optical switching device in the actual transmission time period;

[0021] The physical links in the all-optical data center network having available wavelengths in the actual transmission time period are mapped as virtual links, and the occupation configuration information of the virtual links in the actual transmission time period is obtained; the occupation configuration information includes the occupation of the receiving end server, the occupation of the transmitting end server, and the port configuration of the wavelength division optical switching device;

[0022] The virtual server and server resource list in the actual transmission time period, the virtual wavelength division optical switch device and its corresponding wavelength connection table, and the virtual link and its occupation configuration information of the all-optical data center network are taken as the computing power wave plane list information in the actual transmission time period, and the computing power wave plane list is obtained.

[0023] Preferably, after obtaining the current working path, the method further comprises:

[0024] Based on the configuration information and configuration capability of the wavelength division optical switch device port in the current working path of the to-be-transmitted service, it is judged whether the port of the wavelength division optical switch device in the current working path needs to be reconfigured:

[0025] If reconfiguration is not needed, the service start time of the to-be-transmitted service is taken as the transmission start time, and the to-be-transmitted service is transmitted along the current working path;

[0026] If reconfiguration is needed, the service start time of the to-be-transmitted service is added to the wavelength division optical switch device port reconfiguration time to obtain the transmission start time, and the to-be-transmitted service is transmitted along the current working path.

[0027] Preferably, after obtaining the current working path, the method further comprises:

[0028] The idle start time of the source server on the current working path and the idle start time of the target working server are obtained, and the later time is taken as the earliest idle time;

[0029] The earliest idle time is compared with the service start time of the to-be-transmitted service:

[0030] If the earliest idle time is earlier than the service start time of the to-be-transmitted service, the service start time of the to-be-transmitted service is updated to the earliest idle time.

[0031] Preferably, based on the standby working servers and standby paths of the services that have been successfully deployed in the time period that exists time intersection with the actual transmission time period, a shared protection server list and a shared protection path list are constructed, comprising:

[0032] The standby working servers of all the services that have been successfully deployed in the time period that exists time intersection with the actual transmission time period of the to-be-transmitted service are obtained, and a shared protection server list is constructed;

[0033] The standby paths corresponding to all the standby working servers in the shared protection server list are obtained, and the optical channel construction order between all the wavelength division optical switch devices on the standby paths is taken as the protection path, and a shared protection path list is constructed;

[0034] delete the backup working server in the shared protection server list which does not meet the required computing resource of the service of the to-be-transmitted service, and update the shared protection server list;

[0035] and delete the protection path corresponding to the deleted backup working server from the shared protection path list, and update the shared protection path list.

[0036] Preferably, the judgment of whether there is a protection server in the shared protection server list can obtain a protection path from the candidate source server of the to-be-transmitted service to the protection server in the shared protection path list, including:

[0037] The source wavelength division optical switching device directly connected with the source candidate server is a source wavelength division optical switching device;

[0038] The wavelength division optical switching device directly connected with the protection server in the shared protection server list is a destination wavelength division optical switching device;

[0039] Judgment of whether there is a path from the source wavelength division optical switching device to the destination wavelength division optical switching device in the shared protection path list:

[0040] If there is, it means that there is a protection server in the shared protection server list, and a protection path from the candidate source server of the to-be-transmitted service to the protection server can be obtained in the shared protection path list, so as to build a backup path of the to-be-transmitted service.

[0041] If not, it means that the protection servers in the shared protection server list cannot obtain a protection path from the candidate source server of the to-be-transmitted service to the protection server in the shared protection path list.

[0042] Preferably, based on the computing power wave plane list, a plurality of candidate servers meeting the required computing resource of the service of the to-be-transmitted service are obtained, including: obtaining the computing resource of the virtual server in each computing power wave plane, and obtaining a plurality of servers meeting the required computing resource of the service of the to-be-transmitted service as candidate servers.

[0043] Preferably, when there are a plurality of to-be-transmitted services, and the current protection paths of the plurality of to-be-transmitted services do not intersect, the plurality of to-be-transmitted services are allowed to share the same backup working server and the same backup path.

[0044] Preferably, the two paths do not intersect, including that any link and any wavelength division optical switching device in the two paths do not overlap.

[0045] The embodiment also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the all-optical data center network oriented computing network fusion protection method.

[0046] The above technical scheme of the present application has the following beneficial effects compared with the prior art:

[0047] The all-optical data center network algorithm network fusion protection method provided by the present application maps the servers, wavelength division optical switching devices and physical links in the all-optical data center network, and obtains corresponding computing power wave planes; based on the computing resources, wavelength connection tables and occupation configuration information recorded in the computing power wave planes, a target working server and a current working path are selected. In order to further guarantee the service survivability, the present application adopts a shared protection strategy at the path level and the computing resource level, and considers the survivability of the servers and the paths, and selects a standby working server and a standby path for the to-be-transmitted service through the constructed shared protection server list and the shared protection path list; the path protection ensures the stability and continuity of data transmission, and improves the security of data transmission; when a server or a path fails, the standby working server and the standby path can quickly take over the task, reduce the downtime, and ensure the service survivability; the present application balances the cost benefit and the protection capability through efficient resource management and scheduling, optimizes the resource utilization rate, and improves the network performance.

[0048] The present application provides a standby source server for the to-be-transmitted service, ensures that the data can be protected during storage and processing, and avoids data loss and damage;

[0049] The present application allows more than two link disjoint services to share the computing power protection resources on the same server and share the same standby path, optimizes the resource utilization rate, improves the overall system performance, and provides a flexible, efficient and comprehensive survivability solution for the to-be-transmitted service in the all-optical data center network. BRIEF DESCRIPTION OF DRAWINGS

[0050] In order to make the content of the present application more easily understood, the present application will be further described in detail below according to specific embodiments of the present application and in conjunction with the drawings, in which:

[0051] Figure 1 is a step flow chart of the all-optical data center network algorithm network fusion protection method provided by the present application;

[0052] Figure 2 is a physical structure schematic diagram of the all-optical data center network provided by the present application;

[0053] Figure 3 is a computing power wave plane list of the all-optical data center network provided by the present application;

[0054] Figure 4 is a resource deployment flow chart provided by the present application;

[0055] Figure 5 is a flowchart based on the computing power wavelength plane algorithm provided by the present application. DETAILED DESCRIPTION

[0056] The present application will be further described below in conjunction with the drawings and specific embodiments, so that those skilled in the art can better understand the present application and implement it. The embodiments are not intended to limit the present application.

[0057] Referring to Figure 1 The step flowchart of the all-optical data center network oriented algorithm network fusion protection method of the present application is shown in the figure, and the specific steps include:

[0058] S101: Construct an all-optical data center network composed of multiple servers and multiple wavelength division optical switching devices;

[0059] S102: Obtain the source server, the alternative source server, the computing resources required by the service, the time required for service transmission, and the service start time of the service to be transmitted;

[0060] S103: Based on the service start time and the time required for service transmission of the service to be transmitted, obtain the actual transmission period of the service to be transmitted;

[0061] S104: Map the servers and wavelength division optical switching devices in the all-optical data center network to virtual servers and virtual wavelength division optical switching devices, respectively; map the physical links with available wavelengths in the actual transmission period to virtual links; based on the virtual servers and their computing resources, the virtual wavelength division optical switching devices and their wavelength connection tables, and the virtual links and their occupancy configuration information, construct the computing power wavelength plane list of the all-optical data center network in the actual transmission period;

[0062] S105: Based on the computing power wavelength plane list, obtain multiple alternative servers that meet the computing resources required by the service to be transmitted; for each alternative server, obtain the shortest route between it and the source server of the service to be transmitted; based on the computing resources of each alternative server and the shortest route, and the reconfiguration times of the wavelength division optical switching devices on the shortest route, select a target working server and obtain the target wavelength and the current working path from the source server to the target working server;

[0063] Among them, the selection of the alternative server includes: obtaining the computing resources of the virtual servers in each computing power wavelength plane, obtaining multiple servers whose computing resources meet the computing resources required by the service to be transmitted, as alternative servers;

[0064] S106: Based on the standby working servers and standby paths of the services that have been successfully deployed in the time period that overlaps with the actual transmission period, construct a shared protection server list and a shared protection path list;

[0065] S107: Determine if a protection server exists in the shared protection server list, and obtain a protection path from the alternative source server of the service to be transmitted to the protection server that does not intersect with the current working path from the shared protection path list:

[0066] If it exists, the protection server is acquired and used as a backup working server for the service to be transmitted; and a backup path for the service to be transmitted is constructed based on the protection path corresponding to the protection server.

[0067] If not, calculate the shortest path from the alternative source server of the service to be transmitted to each protection server in the shared protection server list, and select the shortest path that does not intersect with the current working path from all the shortest paths as the protection path and add it to the shared protection path list; construct the backup path for the service to be transmitted based on the protection path, and use the backup path and the protection server corresponding to the backup path as the backup path and backup working server for the service to be transmitted.

[0068] S108: Based on the target working server, current working path, backup working server and backup path of the service to be transmitted, transmit the service to be transmitted.

[0069] Specifically, refer to Figure 2 The diagram shown is a schematic of the physical structure of an all-optical data center network; refer to... Figure 3 The image shows a list of computing power waveplanes for an all-optical data center network. A list of computing power waveplanes for the all-optical data center network during the actual transmission time period is constructed, including:

[0070] S104-1: Map the servers in the all-optical data center network to virtual servers, and record the computing resources of each virtual server in the server resource list during the actual transmission time period;

[0071] S104-2: Map the wavelength division multiplexing (WDM) switching devices in the all-optical data center network to virtual WDM switching devices, and create a wavelength connection table corresponding to each virtual WDM switching device based on the wavelength connection information of each port in each virtual WDM switching device during the actual transmission time period.

[0072] S104-3: Map the physical links with available wavelengths in the all-optical data center network during the actual transmission time period to virtual links, and obtain the occupancy configuration information of the virtual links during the actual transmission time period; the occupancy configuration information includes the occupancy status of the receiving server, the occupancy status of the transmitting server, and the port configuration status of the wavelength division multiplexing (WDM) switching device.

[0073] Among them, the transmitting end server refers to the source server and the alternative source server, and the receiving end server refers to the target working server and the backup working server.

[0074] S104-4: obtaining the list of computing power wave planes as the list of computing power wave planes in the actual transmission time period, by taking the list of virtual servers and server resources in the actual transmission time period, the list of virtual wavelength division optical switching devices and the corresponding wavelength connection table, and the list of virtual links and the occupied configuration information thereof in the actual transmission time period.

[0075] The embodiment adds mapping of server available states and computing resources on a conventional wavelength plane, so that unified management and control of network resources and computing resources are realized based on the computing power wave plane, the path allocation efficiency is improved, and service survivability is ensured.

[0076] Specifically, in the embodiment of the application, after the current working path is obtained, the method further includes: judging whether the ports of the wavelength division optical switching device in the current working path need to be reconfigured based on the configuration information and configuration capability of the ports of the wavelength division optical switching device in the current working path of the to-be-transmitted service; if the ports of the wavelength division optical switching device do not need to be reconfigured, taking the service start time of the to-be-transmitted service as the transmission start time, and transmitting the to-be-transmitted service along the current working path; if the ports of the wavelength division optical switching device need to be reconfigured, taking the service start time of the to-be-transmitted service plus the reconfiguration time of the ports of the wavelength division optical switching device as the transmission start time, and transmitting the to-be-transmitted service along the current working path.

[0077] Specifically, in the embodiment of the application, after the current working path is obtained, the method further includes: obtaining the idle start time of the source server and the idle start time of the target working server on the current working path, taking the later time as the earliest idle time; comparing the earliest idle time with the service start time of the to-be-transmitted service; and if the earliest idle time is earlier than the service start time of the to-be-transmitted service, updating the service start time of the to-be-transmitted service to the earliest idle time.

[0078] The embodiment adjusts the service start time of the to-be-transmitted service in combination with the actual network situation of the all-optical data center network, so as to realize more efficient resource management and scheduling.

[0079] Specifically, in step S106, the shared protection server list and the shared protection path list are built, including:

[0080] S106-1: obtaining the standby working servers of all successfully deployed services in a time period that exists time intersection with the actual transmission time period of the to-be-transmitted service, and constructing a shared protection server list;

[0081] S106-2: obtaining the standby paths corresponding to all standby working servers in the shared protection server list, obtaining the optical channel construction sequence between all wavelength division optical switching devices on the standby paths as protection paths, and constructing a shared protection path list;

[0082] S106-3: deleting the backup working server in the shared protection server list which does not meet the required computing resource of the service to be transmitted, updating the shared protection server list;

[0083] S106-4: deleting the protection path corresponding to the deleted backup working server from the shared protection path list, updating the shared protection path list.

[0084] Specifically, in step S107, it is judged whether there is a protection server in the shared protection server list and whether a protection path from the candidate source server of the service to be transmitted to the protection server can be obtained in the shared protection path list, including:

[0085] The source wavelength division optical switching device directly connected with the source candidate server is a source wavelength division optical switching device;

[0086] The destination wavelength division optical switching device directly connected with the protection server in the shared protection server list is a destination wavelength division optical switching device;

[0087] It is judged whether there is a path from the source wavelength division optical switching device to the destination wavelength division optical switching device in the shared protection path list:

[0088] If there is, it means that there is a protection server in the shared protection server list and a protection path from the candidate source server of the service to be transmitted to the protection server can be obtained in the shared protection path list, so as to construct a backup path of the service to be transmitted.

[0089] If not, it means that the protection servers in the shared protection server list cannot obtain a protection path from the candidate source server of the service to be transmitted to the protection server in the shared protection path list.

[0090] If there is a path from the source wavelength division optical switching device to the destination wavelength division optical switching device in the shared protection path list, it is judged whether the source wavelength division optical switching device and the destination wavelength division optical switching device need to be ported, if yes, the corresponding protection path is obtained after the porting is completed, if not, the path at the current time is directly obtained as the protection path.

[0091] In this embodiment, when selecting a backup path, a protection path from one wavelength division optical switching device to another wavelength division optical switching device is selected, and the two wavelength division optical switching devices at both ends of the protection path are directly connected with the candidate source server and the protection server respectively to obtain the corresponding backup path.

[0092] In the embodiment, when there are multiple services to be transmitted, and the current protection paths of the multiple services to be transmitted do not intersect, the multiple services to be transmitted are allowed to share the same backup working server and the same backup path. Wherein, the two paths do not intersect, including any link and any wave division optical switching device in the two paths do not overlap. The application allows more than two links of the services to share the computing power protection resources on the same server and share the same backup path, optimizes the resource utilization, improves the overall performance of the system, and provides a flexible, efficient and comprehensive survivability solution for the services to be transmitted in the all-optical data center network

[0093] The embodiment inherits the traditional survivability algorithm, and additionally introduces the survivability of computing power resources. Referring to the 1+1 protection and shared protection concept in path protection, the embodiment designs a corresponding shared backup computing power protection strategy (Shared Backup Computing Power Protection, SBCPP) for computing power resources, which is a shared protection strategy for computing power resources, similar to the SBPP (Shared Backup Path Protection) strategy of path protection, but the SBCPP strategy focuses on computing power and allows more than two services to share the computing power protection resources on the same server. In the service deployment process, the application comprehensively considers the dual protection of path survivability and computing power survivability. For path protection, the application uses the traditional SBPP strategy, combined with the SBCPP strategy of computing power protection, that is, combines SBPP path protection and SBCPP computing power protection to efficiently complete the allocation of working paths and protection paths of services.

[0094] Based on the above embodiment, in the embodiment of the application, the service sequence is transmitted by using the all-optical data center network computing network fusion protection method provided by the application, as shown in Figure 4 , which is a resource deployment flowchart, and the specific steps include:

[0095] S201: generating a service sequence of intensive computing;

[0096] Specifically, given a set of intensive computing requirements, each service requirement is represented as (s, d, s p , d p , T, r, r p , t). It is known that s, d, s p , d p , T, t respectively represent the source server, the working server, the backup source server, the protection server, the time required for service transmission, and the time when the service starts to be deployed. All data is randomly generated. r represents the working route of the service, r pThe protection route of the service, generally empty. d represents the working server of the service, d p The protection server of the service, which is preset to be empty.

[0097] For each service requirement, the destination server d is obtained according to an algorithm, then the path r between the server s and d is found, and the wavelength available in T consecutive time slots along the path is found.

[0098] And the protection server d is allocated to the service p , then the path r between the server s p and d p is found p , and the services are deployed in sequence according to the generation order.

[0099] S202: Constructing the computing wave plane, the shared protection server list and the shared protection path list;

[0100] According to the number of available wavelengths N, the computing wave planes ω1, ω2, …, ω N of the all-optical data center network within [t, t+T-1] are generated According to the time, the link occupation, the receiver transmitter occupation and the switch port configuration of the successfully deployed services are updated on the computing wave plane corresponding to the service, and the server computing resource list is updated.

[0101] According to the time [t, t+T-1] and the successfully deployed services that cross the to-be-transmitted service, it is determined whether the protection server corresponding to the service exists in the shared protection server list. If the protection server exists, it is determined whether the shared protection computing resource of the corresponding server in the shared server list is greater than the protection computing resource occupied by the current service. If yes, the protection computing resource number is updated, otherwise no processing is performed. If the protection server does not exist, the protection server corresponding to the service and the protection computing resource occupied on the server are added to the shared protection server list. Meanwhile, it is determined whether the protection optical channel corresponding to the service exists in the shared protection path list. If not, the protection optical channel corresponding to the service is recorded in the shared protection path list.

[0102] S203: Allocating the working server, the working path, the protection server and the protection path to the service;

[0103] The work server and work path are deployed using the computing power wave plane algorithm: for a high-performance computing service, the system takes the service arrival time as the service start scheduling deployment time, first obtains the service deployment work server according to the computing power wave plane algorithm. Then, the all-optical data center network in the next T continuous time slots is used to find the shortest path by Dijkstra and judge whether there is corresponding transceiver, wavelength and switch port configuration in T continuous time slots to perform service deployment. If yes, the optical channel is established for the service to transmit data, the current service work path deployment is successful; and the link occupation, transceiver and transmitter occupation, switch port configuration on the work path corresponding computing power wave plane are updated, and the computing power resource usage of the work server is updated.

[0104] The embodiment uses the SBPP path protection + SBCPP computing power protection strategy to complete the deployment of the protection server and the protection path: this strategy adopts a shared protection strategy at the link and computing power resource level, aiming to balance cost effectiveness and protection capability through efficient resource management and scheduling. In this strategy, when there is no intersection link between the main paths of multiple services, multiple services are allowed to share the same protection path, and the computing resources on the same server are allowed to be shared.

[0105] First, check whether the current shared protection server list is empty. If it is empty, the protection server and protection path are selected using the computing power wave plane algorithm, and the protection server and protection path are updated to the shared protection server list and shared protection path list. If it is not empty, check whether the shared protection path list has an optical channel connecting s p and the protection server d p If yes, the protection path and protection server deployment is completed. If all protection servers in the shared protection server list do not have available protection optical channels with s p , the shortest path is found using Dijkstra and it is judged whether there is corresponding transceiver, wavelength and switch port configuration in T continuous time slots to perform deployment, and finally the found optical channel is recorded to the shared protection path list.

[0106] When the work path, work server, standby path and protection server are deployed, the current service deployment is completed, otherwise it will be waiting, repeating the above process until the service deployment is successful.

[0107] Specifically, the definition of the computing wave plane is that the computing wave plane can be understood as a new definition extended from the concept of the traditional wave plane; the traditional wave plane only considers the mapping of network devices (switches) and links (fiber links), and is not aware of computing power. The computing wave plane adds the mapping of server availability and computing resources on the server to the concept of the traditional wave plane, realizing unified management and control of network resources and computing resources.

[0108] Referring to Figure 5 Fig. 1 shows a flowchart based on the computing wavelength plane algorithm, and the main steps are as follows:

[0109] S301: Construct the computing wave plane in the corresponding time period;

[0110] For each wavelength, a computing wave plane is generated, which is copied from the physical network topology, wherein each virtual server (FPGA) and virtual wavelength division optical switching device are mapped from the physical server node and physical wavelength division optical switching device, and each virtual link corresponds to whether the corresponding wavelength of the physical link is available.

[0111] When the corresponding wavelength of the physical link is available within the service duration, the link is added to the corresponding computing wave plane.

[0112] When the corresponding computing power is available on the physical server node to provide services within a given time, a virtual server is added, and the computing resources of the server are recorded in the server resource list on the computing wave plane.

[0113] In addition, a wavelength connection table is created for each wavelength division optical switching device to store the wavelength connection information of each port therein; using the availability of each wavelength and the computing power on each node, the algorithm can generate a list of computing wave planes ω1, ω2, …, ω N of the all-optical DCN in the corresponding time period; and after each service is completed, the network and server resource mapping list on the computing wave plane is updated;

[0114] S302: Select a candidate destination server node for the server to obtain a candidate server;

[0115] Find the server computing resource status on the constructed computing wave plane, and select k servers with sufficient computing resources for the service, i.e., candidate servers d1, d2, …, d k , wherein 0 < k < N, and N represents the total number of servers;

[0116] S303: Find the computing wave plane list and select the route;

[0117] For each candidate server d1, d2, …, d kCorresponding to each wavelength of the all-optical data center network in [t, t+T-1] The computing power wave plane ω1, ω2, …, ω N Query one by one, try to find the shortest route between the source server and the candidate server on each computing power wavelength plane by Dijkstra algorithm, and the wavelength corresponding to the current computing power wave plane is the wavelength used by the route r;

[0118] Scan to get the earliest idle time t of the transmitter and receiver at both ends of the route x , t r Update the service time;

[0119] S304: Joint optimization of server computing resources and route selection by algorithm, get the optimal server as the destination server D, and get the corresponding shortest route information R and wavelength ω*;

[0120] The joint optimization includes: in all reachable candidate servers that can find corresponding routes, select the corresponding route R and server with the least number of wavelength division optical switching device switch reconfiguration as the optimal server D according to the descending order of computing resources.

[0121] S305: Check the switch port;

[0122] Check the configuration information of the switch port through which the route R passes, and determine whether reconfiguration is needed according to whether the wavelength is configured between the ports and whether the switch has reconfiguration capability;

[0123] If reconfiguration is needed, the transmission time is extended by ts, where ts is the switch port reconfiguration time, that is, T=T+ts, and the service is deployed; If reconfiguration is not needed, the service is directly deployed;

[0124] S306: Transmission is completed, and the service is processed at the server end;

[0125] During the service computing time, the corresponding computing resources of the server are occupied, and the computing resource list is updated.

[0126] Based on the above embodiment, in this embodiment, the all-optical data center network is composed of servers and switches. A total of 12 servers are arranged in the network, and the computing capacity of the servers is randomly generated in the range of [50TFLOPS / s, 500TFLOPS / s], where FLOPS / s represents the floating point operation number per second. The total number of service requests in the network is fixed, and the time required for service transmission is randomly generated in the range of [20, 50] time slots TS (Time Slot), and the computing requirement of the service is randomly generated in the computing power requirement interval [2TGFLOPS, 20TFLOPS], wherein the unit FLOPS is the floating point operation number. The service arrival time is initialized to 0 TS. No wavelength conversion is allowed in the service configuration process, so each optical path is subject to the constraint of wavelength continuity. In addition, in order to prevent the risk of simultaneous failure of working paths and protection paths caused by access switch failure, a preventive measure is taken in this embodiment: the service data is backed up to the servers connected by two different input ports in advance, so as to ensure the accessibility of the data and the continuity of the service. The protection algorithm is used to allocate working servers and working paths, protection servers and protection paths for the service respectively, to obtain the maximum number of wavelengths used for task completion and the redundancy of backup resources.

[0127] The embodiment of the application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the steps of the all-optical data center network oriented algorithm network fusion protection method.

[0128] The all-optical data center network-oriented algorithm network fusion protection method maps servers, wavelength division optical switching devices and physical links in the whole network data center network, and obtains corresponding computing power wave planes; based on the computing resources, wavelength connection table and occupation configuration information recorded in the computing power wave plane, a target working server and a current working path are selected. In order to further guarantee the service survivability, the sharing protection strategy is adopted at the path level and the computing resource level, and the survivability of the server and the path is considered, and the sharing protection server list and the sharing protection path list are constructed to select the standby working server and the standby path for the to-be-transmitted service; the path protection ensures the stability and continuity of data transmission, and improves the security of data transmission; when the server or the path fails, the standby working server and the standby path can quickly take over the task, reduce the downtime, and ensure the service survivability; the application realizes the balance between cost benefit and protection capability through efficient resource management and scheduling. The application provides a standby source server for the to-be-transmitted service, ensures that the data can be protected in the storage and processing process, avoids data loss and damage, and optimizes the resource utilization rate and improves the overall performance of the system. The application provides a flexible, efficient and comprehensive survivability solution for the to-be-transmitted service in the all-optical data center network.

[0129] Those skilled in the art will appreciate that embodiments of the application can be provided as methods, systems, or computer program products. Accordingly, the application can be embodied in the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the application can be embodied in the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk memory, CD-ROM, optical memory, etc.) having computer usable program code embodied thereon.

[0130] The application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of the embodiments of the application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one flow or multiple flows and / or blocks Figure 1 The device that implements the functions specified in one block or multiple blocks.

[0131] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 The flow or flows and / or blocks Figure 1 The flow or flows and / or blocks

[0132] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions that execute on the computer or other programmable apparatus provide steps for implementing the flow Figure 1 The flow or flows and / or blocks Figure 1 The flow or flows and / or blocks

[0133] Obviously, the above-described embodiments are only examples and are not intended to limit the present application. Based on the above description, one of ordinary skill in the art can make other variations and changes without departing from the present application. It is not necessary or possible to enumerate all the embodiments. The obvious variations and changes derived therefrom are still within the scope of the present application.

Claims

1. An algorithm-network fusion protection method for an all-optical data center network, characterized in that, The application relates to a method for constructing a full-optical data center network. The method comprises the following steps: constructing a full-optical data center network composed of multiple servers and multiple wavelength division optical switching devices; obtaining a source server, an alternative source server, required computing resources of a service, a service transmission required time and a service start time of the service to be transmitted; obtaining an actual transmission time period of the service to be transmitted based on the service start time and the service transmission required time of the service to be transmitted; mapping the servers and the wavelength division optical switching devices in the full-optical data center network into virtual servers and virtual wavelength division optical switching devices respectively, mapping the physical links with available wavelengths in the actual transmission time period into virtual links, and constructing an algorithm power wave plane list of the full-optical data center network in the actual transmission time period based on the virtual servers and the computing resources thereof, the virtual wavelength division optical switching devices and the wavelength connection tables thereof, and the virtual links and the occupied configuration information thereof; obtaining multiple alternative servers satisfying the required computing resources of the service to be transmitted based on the algorithm power wave plane list; obtaining the shortest route between each alternative server and the source server of the service to be transmitted; selecting a target working server, obtaining a target wavelength and a current working path from the source server to the target working server based on the computing resources of each alternative server, the shortest route and the reconfiguration times of the wavelength division optical switching devices on the shortest route; constructing a shared protection server list and a shared protection path list based on the standby working servers and the standby paths of the services successfully deployed in the time period intersecting with the actual transmission time period; judging whether there is a protection server in the shared protection server list, and whether a protection path intersecting with the current working path and from the alternative source server of the service to be transmitted to the protection server can be obtained in the shared protection path list: if yes, obtaining the protection server as a standby working server of the service to be transmitted; and constructing a standby path of the service to be transmitted based on the protection path corresponding to the protection server; if not, calculating the shortest paths from the alternative source server of the service to be transmitted to each protection server in the shared protection server list, and selecting the shortest path intersecting with the current working path as the protection path, and adding the protection path into the shared protection path list; constructing the standby path of the service to be transmitted based on the protection path, and taking the standby path and the protection server corresponding to the standby path as the standby path and the standby working server of the service to be transmitted; 2. The all-optical data center network oriented algorithm and network fusion protection method of claim 1, wherein, transmitting the service to be transmitted based on the target working server, the current working path, the standby working server and the standby path of the service to be transmitted. The method for constructing the algorithm power wave plane list of the full-optical data center network in the actual transmission time period comprises the following steps: mapping the servers in the full-optical data center network into virtual servers, and recording the computing resources of each virtual server in the actual transmission time period in a server resource list; mapping the wavelength division optical switching devices in the full-optical data center network into virtual wavelength division optical switching devices, and creating a wavelength connection table corresponding to each virtual wavelength division optical switching device based on the wavelength connection information of each port in each virtual wavelength division optical switching device in the actual transmission time period. The physical link with available wavelengths in the actual transmission time period in the all-optical data center network is mapped to a virtual link, and occupation configuration information of the virtual link in the actual transmission time period is obtained; the occupation configuration information includes receiving end server occupation, transmitting end server occupation, and wave division optical switching device port configuration; The virtual server and server resource list in the actual transmission time period, the virtual wave division optical switching device and the corresponding wavelength connection table, and the virtual link and the occupation configuration information thereof in the all-optical data center network are taken as the actual transmission time period, and an algorithmic wave plane list is obtained.

3. The all-optical data center network oriented algorithm and network fusion protection method of claim 1, wherein, After obtaining the current working path, the following steps are further included: Based on the configuration information and configuration capability of the wave division optical switching device port in the current working path of the to-be-transmitted service, it is judged whether the port of the wave division optical switching device in the current working path needs to be reconfigured: If reconfiguration is not needed, the service start time of the to-be-transmitted service is taken as the transmission start time, and the to-be-transmitted service is transmitted along the current working path; If reconfiguration is needed, the service start time of the to-be-transmitted service is added to the wave division optical switching device port reconfiguration time to obtain the transmission start time, and the to-be-transmitted service is transmitted along the current working path.

4. The all-optical data center network oriented algorithm and network fusion protection method of claim 1, wherein, After obtaining the current working path, the following steps are further included: The idle start time of the source server and the idle start time of the target working server on the current working path are obtained, and the later time is taken as the earliest idle time; The earliest idle time is compared with the service start time of the to-be-transmitted service: If the earliest idle time is earlier than the service start time of the to-be-transmitted service, the service start time of the to-be-transmitted service is updated to the earliest idle time.

5. The all-optical data center network oriented algorithm and network fusion protection method of claim 1, wherein, Based on the standby working servers and standby paths of the services successfully deployed in the time period intersecting with the actual transmission time period, a shared protection server list and a shared protection path list are constructed, including: All standby working servers of the services successfully deployed in the time period intersecting with the actual transmission time period of the to-be-transmitted service are obtained to construct a shared protection server list; The standby paths corresponding to all standby working servers in the shared protection server list are obtained, and the optical channel construction sequence between all wave division optical switching devices on the standby paths is obtained as protection paths to construct a shared protection path list; The standby working servers in the shared protection server list whose computing resources do not meet the required computing resources of the to-be-transmitted service are deleted, and the shared protection server list is updated; The protection paths corresponding to the deleted standby working servers are deleted from the shared protection path list, and the shared protection path list is updated.

6. The all-optical data center network oriented algorithm and network fusion protection method of claim 1, wherein, It is judged whether there is a protection server in the shared protection server list, and a protection path from the standby source server of the to-be-transmitted service to the protection server can be obtained in the shared protection path list, including: The wave division optical switching device directly connected to the source standby server is taken as the source wave division optical switching device; The wave division optical switching device directly connected to the protection server in the shared protection server list is taken as the destination wave division optical switching device; determining whether there is a path from the source Wavelength Division Multiplexing (WDM) switching device to the destination WDM switching device in the shared protection path list: if yes, it indicates that there is a protection server in the shared protection server list, and a protection path from the candidate source server of the to-be-transmitted service to the protection server can be obtained in the shared protection path list, so as to construct a backup path of the to-be-transmitted service; if no, it indicates that none of the protection servers in the shared protection server list can obtain a protection path from the candidate source server of the to-be-transmitted service to the protection server in the shared protection path list.

7. The all-optical data center network oriented algorithm and network fusion protection method of claim 1, wherein, Based on the computing power wave plane list, a plurality of candidate servers satisfying the required computing resources of the to-be-transmitted service are obtained, including: obtaining the computing resources of the virtual servers in each computing power wave plane, and obtaining a plurality of servers satisfying the required computing resources of the to-be-transmitted service as the candidate servers.

8. The all-optical data center network oriented algorithm and network fusion protection method of claim 1, wherein, When there are a plurality of to-be-transmitted services, and the current protection paths of the plurality of to-be-transmitted services do not intersect with each other, the plurality of to-be-transmitted services are allowed to share the same backup working server and the same backup path. 9.The all-optical data center network oriented algorithm and network fusion protection method according to claim 1, characterized in that, The two paths do not intersect, including that any link and any WDM switching device in the two paths do not coincide.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the all-optical data center network oriented computing network fusion protection method according to any one of claims 1 to 9.

Images