CDN provider switching method and system, and device, product and storage medium

Abstract

The present disclosure relates to the technical field of communications. Disclosed are a CDN provider switching method and system, and a device, a product and a storage medium. The method comprises: acquiring a first CDN provider currently used by a target domain name, and a first CDN node under the first CDN provider for providing an acceleration service for the target domain name; searching a CDN node repository for a second CDN node matching the first CDN node, wherein the CDN node repository is used for storing, in historical domain name accesses, CDN nodes for providing the acceleration service for one or more domain names and node states of the CDN nodes when providing the acceleration service; on the basis of the node state of the second CDN node in the CDN node repository, determining the provider state of the first CDN provider; and if the provider state indicates that the first CDN provider is abnormal, switching the first CDN provider, which is currently used by the target domain name, to a second CDN provider.

Description

Method, system, device, product and storage medium for switching CDN provider

[0001] Cross-reference to Related Applications

[0002] The present application claims priority to the Chinese patent application No. 202410865741.X, filed on June 28, 2024, and entitled "Method, system, device, product and storage medium for switching CDN provider", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0003] The present disclosure relates to the technical field of communication, and particularly relates to a method, system, device, product and storage medium for switching CDN provider. BACKGROUND

[0004] In order to provide better access experience for end users, many Internet service providers usually use the acceleration service provided by CDN (Content Delivery Network) to deliver various network resources. However, this CDN-based resource delivery method is often affected by various faults of CDN, such as service exception of CDN edge node, network failure of CDN node in a certain region, etc. SUMMARY

[0005] Therefore, the present disclosure provides a method, system, electronic device, computer program product and computer readable storage medium for switching CDN provider.

[0006] In one aspect, the present disclosure provides a method for switching CDN provider, comprising:

[0007] obtaining a first CDN provider currently used by a target domain name, and a first CDN node under the first CDN provider for providing acceleration service for the target domain name;

[0008] finding a second CDN node matched with the first CDN node from a CDN node library, wherein the CDN node library is used to save CDN nodes providing acceleration service for one or more domain names in historical domain name access and node states of each CDN node when providing acceleration service;

[0009] determining a provider state of the first CDN provider according to the node state of the second CDN node in the CDN node library;

[0010] if the provider state represents that the first CDN provider is abnormal, switching the first CDN provider currently used by the target domain name to a second CDN provider.

[0011] The present disclosure provides, in one aspect, a CDN provider switching system, the system comprising:

[0012] a CDN node acquisition module, configured to acquire a first CDN provider currently used by a target domain name, and a first CDN node under the first CDN provider for providing acceleration service for the target domain name;

[0013] a CDN node screening module, configured to find a second CDN node matching the first CDN node from a CDN node library, wherein the CDN node library is configured to save CDN nodes providing acceleration service for one or more domain names in historical domain name access, and node states of the respective CDN nodes when providing acceleration service;

[0014] a provider state determination module, configured to determine a provider state of the first CDN provider according to the node state of the second CDN node in the CDN node library;

[0015] a provider switching module, configured to switch the first CDN provider currently used by the target domain name to a second CDN provider if the provider state represents that the first CDN provider is abnormal.

[0016] The present disclosure provides, in another aspect, a computer readable storage medium for storing a computer program, wherein the computer program is executed by a processor to implement the method described above.

[0017] The present disclosure provides, in another aspect, an electronic device comprising a processor and a memory, wherein the memory is configured to store a computer program, and the computer program is executed by the processor to implement the method described above.

[0018] The present disclosure provides, in another aspect, a computer program product comprising a computer program, wherein the computer program is executed by a processor to implement the method described above. BRIEF DESCRIPTION OF DRAWINGS

[0019] The features and advantages of the present disclosure will be more clearly understood through the following detailed description taken in conjunction with the accompanying drawings, which are shown by way of illustration and not by way of limitation, wherein:

[0020] Fig. 1 shows a schematic diagram of disaster recovery switching of CDN providers in some technologies;

[0021] Fig. 2 shows a domain name access process in a CDN scenario;

[0022] Fig. 3 shows a schematic diagram of the relationship between a domain name and CDN nodes under a CDN provider;

[0023] FIG. 4 shows a flowchart of a switching method according to an embodiment of the present disclosure;

[0024] FIG. 5 shows an interaction diagram of a disaster recovery switching of a CDN provider used by a newly configured domain name according to an embodiment of the present disclosure;

[0025] FIG. 6 shows a module diagram of a switching system according to an embodiment of the present disclosure;

[0026] FIG. 7 shows a diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0027] For the purpose of making the objects, technical solutions and advantages of embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only some of the embodiments of the present disclosure, rather than all the embodiments. Based on the embodiments in the present disclosure, any other embodiments obtained by a person of ordinary skill in the art without creative work should be within the protection scope of the present disclosure.

[0028] Embodiments of the present disclosure will be described in greater detail below with reference to the drawings. Although some embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms, and should not be interpreted as being limited to the embodiments set forth herein. Rather, these embodiments are provided so as to make the present disclosure more thorough and complete. It should be understood that the drawings and embodiments of the present disclosure are merely for exemplary purposes, and are not intended to limit the protection scope of the present disclosure.

[0029] As described above, in order to provide better access experience for end users, many Internet service providers usually use the acceleration service provided by CDN (Content Delivery Network) to deliver various network resources. However, this CDN-based resource delivery method is often affected by various faults of CDN, such as service exception of CDN edge node, network failure of CDN node in a certain region, etc.

[0030] In some current technologies, multiple probing nodes are deployed in different regions, and periodic probing access is initiated by the probing nodes to detect whether the CDN nodes under each CDN vendor currently used are available. If a CDN node under a CDN vendor is detected to be faulty, the CDN vendor is switched to an alternative CDN vendor, so as to ensure that the acceleration service can be normally provided for end users. However, in this probing method, additional traffic is often generated, increasing the switching cost.

[0031] Therefore, there is an urgent need for a method that can reduce switching cost, and when switching management is performed on CDN providers, the additional traffic generated can be reduced, and the switching cost can be reduced.

[0032] In the technical scheme of some embodiments of the present disclosure, since the CDN node library stores the CDN nodes that provide acceleration services for one or more domain names in historical domain name access and the node state of each CDN node when providing acceleration services, after obtaining the first CDN node for providing acceleration services for the target domain name, the provider state of the first CDN provider can be determined according to the historical node state of the second CDN node matched with the first CDN node in the CDN node library, and then according to the provider state, it can be determined whether to switch the first CDN provider currently used by the target domain name. This is based on historical domain name access information to switch the CDN provider currently used by the target domain name, which reduces the number of probes, so that when switching management is performed on CDN providers, the additional traffic generated can be reduced, and the switching cost can be reduced.

[0033] Referring to FIG. 1, a schematic diagram of disaster recovery switching of CDN providers in some technologies is shown. In FIG. 1, the domain name refers to an identifier used to locate the location of a network resource when a client accesses the network resource. Each CDN provider includes multiple CDN nodes (also known as edge nodes), which can be used to provide acceleration services for domain names. The principle of CDN-based domain name acceleration is as follows: in the DNS (Domain Name System) setting, through alias record, the domain name is pointed to the CNAME provided by the CDN provider. In this way, the access request for the domain name can be introduced into the network of the CDN provider, so that the client can obtain the network resource under the domain name from one of the CDN nodes in the region, without obtaining the network resource from a source station far away. In this way, domain name acceleration is achieved.

[0034] In order to prevent the CDN node of the CDN provider from failure, usually, the CDN node of multiple CDN providers can be used to provide acceleration service for the domain name in a manner of disaster recovery switching. Specifically, the multiple CDN providers can be divided into a common CDN provider and a cold backup CDN provider. The disaster recovery switching system can issue a probing task to each probing node, and each probing node can periodically send a probing message to each CDN node of the common CDN provider and the cold backup CDN provider to obtain a communication index for representing communication quality, such as TCP connection time, etc. Each probing node can report the obtained communication index to the disaster recovery switching system as a probing result. The disaster recovery switching system can determine whether the CDN node of the common CDN provider and the cold backup CDN provider fails by analyzing the probing result reported by the probing node. When it is determined that the CDN node of the common CDN provider is normal, the domain name can be directed to the CNAME of the common CDN provider through DNS setting, so as to preferentially provide acceleration service for the domain name through the CDN node of the common CDN provider. However, when it is determined that the CDN node of the common CDN provider fails and the CDN node of the cold backup CDN provider is normal, the DNS setting can be modified to direct the domain name to the CNAME of the cold backup CDN provider, so as to provide acceleration service for the domain name through the CDN node of the cold backup CDN provider. In this way, the stability of the acceleration service can be ensured. However, the above scheme has the following problems:

[0035] 1) Each probing node needs to periodically probe the CDN node, and additional traffic will be generated in this process, thereby increasing the switching cost;

[0036] 2) A large number of probing nodes need to be deployed in multiple different regions, and the deployment cost of the probing nodes is high;

[0037] 3) Considering the cost, usually, only one region or one or several representative probing nodes are deployed, and the CDN nodes that can be detected by the few probing nodes are limited, that is, the probing result of the probing nodes cannot truly represent the availability of all CDN nodes in a region, and therefore, disaster recovery false switching or missed switching is prone to occur, and the precision is low;

[0038] 4) Since the network in which the probing node is located has jitter, in order to improve the detection precision, it is usually determined that the CDN node fails after the probing of one probing node fails for a preset number of times. Therefore, in the above scheme, a long time is needed from the real failure of the CDN node to the completion of the disaster recovery switching, and the timeliness is not high.

[0039] To solve the above problems, in combination with reference to FIG. 2, the present disclosure first analyzes the domain name access process under the CDN scenario. As can be seen from FIG. 2, the domain name access process under the CDN scenario can include five stages of DNS resolution, TCP connection and SSL connection, data sending and data response. In the five stages, the main failures that can occur include:

[0040] 1) DNS resolution failure of the domain name;

[0041] 2) TCP network connection failure of the CDN node;

[0042] 3) Service level failure, such as HTTP response returning a status code of 500, etc.

[0043] Among them, the above-mentioned TCP network connection failure is the most frequent failure, which is a four-layer network failure.

[0044] Based on the above description, the present disclosure further analyzes the TCP network connection failure under the CDN scenario. In combination with reference to FIG. 3, it is a schematic diagram of the relationship between the domain name and the CDN node under the CDN provider. As can be seen from FIG. 3, the CDN nodes under different CDN providers can provide acceleration services for the same domain name, and the same CDN node can provide acceleration services for different domain names. Among them, when a CDN node simultaneously provides acceleration services for different domain names, if the operator networks used are the same, then the four-layer network data information between the client initiating an access request for these domain names and the CDN node is consistent.

[0045] Based on the above description, it is assumed that the CDN node providing acceleration services for domain name A and domain name B is the same. First, since the four-layer network failure is the most frequent failure, if the client fails to access domain name A, it can be reasonably inferred that a four-layer network failure has occurred. Secondly, since the four-layer network data information when the client accesses domain name A and domain name B is consistent, therefore, in the case that the client fails to access domain name A, the client will also fail to access domain name B.

[0046] In short, when a same CDN node provides acceleration services for multiple different domain names based on a same operator's network, if the client fails to access one of the domain names, then the client will also fail to access the other domain names.

[0047] According to the above principle, the disclosure provides a CDN provider switching method, which can solve the problems existing in the scheme shown in FIG. 1. The switching method can be implemented based on a CDN node library, so before the switching method of the disclosure is executed, the construction of the CDN node library can be completed first. Specifically, the CDN node library can be used to save the CDN nodes that provide acceleration services for one or more domain names and the node states of each CDN node when providing acceleration services in the historical domain name access. The node state refers to the communication quality index when the CDN node communicates with the client, such as the connection time and connection success rate between the CDN node and the client, the HTTP status code returned by the CDN node to the client, etc. The connection time can also be referred to as TCP connection time, and the connection success rate can also be referred to as TCP connection success rate. Based on the node state of the CDN node when providing acceleration services, the communication quality between the CDN node and the client can be judged, and then whether the CDN node fails can be judged. The following describes how to construct the CDN node library.

[0048] 1) For the historical domain name access, the following information of each domain name access can be obtained respectively: the region where the client initiating the domain name access is located, the operator used by the client, the domain name accessed, the IP of the CDN node providing acceleration services for the domain name, the CDN provider to which the CDN node belongs, the region where the CDN node is located, the operator used by the CDN node, the connection time of the CDN node and the client, whether the CDN node and the client are connected successfully, the HTTP status code returned by the CDN node, etc.

[0049] 2) The domain name access information obtained is data cleaned to ensure data accuracy. For example, in the obtained domain name access information, the connection time of a domain name access is several hours or negative, which is obviously not in line with the actual situation, so the data of this domain name access can be excluded to ensure data accuracy.

[0050] 3) The obtained domain name access information is counted according to the specified counting period and counting dimension to obtain multiple domain name access records in the CDN node library.

[0051] Specifically, in the embodiment shown in the disclosure, the counting period can be 1 minute, and the counting dimension can include the region where the client initiating the domain name access is located, the operator used by the client, the domain name accessed, the IP of the CDN node providing acceleration services for the domain name, the CDN provider to which the CDN node belongs, the region where the CDN node is located, the operator used by the CDN node, etc. Based on the above description, the finally obtained CDN node library can be similar to Table 1 shown in the table:

[0052] Table 1 CDN node library

[0053] In Table 1, the CDN node library includes node states of CDN nodes in multiple statistical periods. The node states include connection durations of the CDN nodes with clients initiating domain name access, connection success rates of the CDN nodes with the clients, and request numbers of the CDN nodes. The connection duration can be obtained by averaging connection durations of multiple domain name accesses in a statistical period. The connection success rate can be obtained according to the number of successful connections and the total number of requests in a statistical period. For example, in a statistical period, a CDN node is requested 5 times, 3 times of which are successful and 2 times of which are failed, and the connection success rate is 60%. The request number can be obtained by counting the number of requests of the CDN node in a statistical period.

[0054] After the CDN node library is constructed, most or all of the CDN nodes under each CDN provider can be located in the CDN node library, and the node states of each CDN node in the CDN node library are dynamically updated according to statistical periods. The dynamic update means that information of domain name access that has occurred is counted in real time according to statistical periods.

[0055] At this point, the construction of the CDN node library is completed.

[0056] Based on the CDN node library that has been constructed, the switching method of the present disclosure can be executed. The switching method can be applied to the disaster recovery switching system in FIG. 1 or to an electronic device running the disaster recovery switching system. The electronic device includes but is not limited to a tablet computer, a notebook computer, a desktop computer, a server, and the like. In combination with FIG. 4, a flowchart of the switching method provided by an embodiment of the present disclosure is shown. In FIG. 4, the switching method includes the following steps:

[0057] In step S41, a first CDN provider currently used by a target domain name and a first CDN node under the first CDN provider for providing acceleration services for the target domain name are obtained.

[0058] Specifically, the target domain name can be a domain name pre-configured in the disaster recovery switching system. The disaster recovery switching system manages CDN providers used by the configured domain names for disaster recovery switching. The CDN node library can include domain name access records of the target domain name or can not include the domain name access records of the target domain name. When the CDN node library does not include the domain name access records of the target domain name, it indicates that the target domain name is a new domain name that has not been accessed or a domain name for which domain name access information has not been obtained.

[0059] DNS resolution is performed on the target domain name to obtain a CNAME pointed by the target domain name. The CDN provider represented by the CNAME is the first CDN provider currently used by the target domain name. DNS resolution is performed on the CNAME of the first CDN provider to obtain the IP of the first CDN node under the first CDN provider for providing acceleration services for the target domain name.

[0060] In this embodiment, the first CDN provider currently used by the target domain name can be divided by region+operator. The first CDN provider used by the target domain name can be different in different regions+operators. For example, in region A+operator A and region B+operator B, the first CDN provider used by the target domain name is CDN provider A; in region A+operator B, the first CDN provider used by the target domain name is CDN provider B.

[0061] In step S41, one or more first CDN providers can be screened, and there can be one or more first CDN nodes under each first CDN provider for providing acceleration services for the target domain name.

[0062] In step S42, a second CDN node matching the first CDN node is searched from the CDN node library.

[0063] In this embodiment, for any first CDN node, a CDN node having the same IP, CDN provider, region and operator as the first CDN node can be searched from the CDN node library, and the searched CDN node is taken as the second CDN node matching the first CDN node. For example, it is assumed that one of the first CDN nodes obtained in step S41 is first CDN node A located under CDN provider A+region A+operator A, and the IP is 1.1.1.1. It is assumed that in the CDN node library, there is a domain name access record similar to that shown in Table 2:

[0064] Table 2 CDN node library

[0065] In Table 2, for the domain name access record with the statistical period of “202209031208”, the operator, CDN node IP and CDN provider are all different from the first CDN node A, so the CDN node in this domain name access record cannot be taken as the second CDA node matching the first CDN node A. However, for the two domain name access records with the statistical periods of “202209031206” and “202209031207”, the IP, CDN provider, region and operator are the same as the first CDN node A, so the CDN nodes in these two records can be taken as the second CDA node matching the first CDN node A.

[0066] Since the CDN node library includes most CDN nodes under each CDN provider, for some or all of the first CDN nodes, a second CDN node matching the first CDN node can be found in the CDN node library. Of course, there can be some first CDN nodes that cannot find a second CDN node matching the first CDN node in the CDN node library.

[0067] In step S43, the provider state of the first CDN provider is determined according to the node state of the second CDN node in the CDN node library.

[0068] The provider state is used to represent whether the first CDN provider is in a normal state. If the first CDN provider is in a normal state, it means that the first CDN node under the first CDN provider does not fail and can normally provide acceleration services for the target domain name. If the first CDN provider is in an abnormal state, it means that the first CDN node of the first CDN provider fails and cannot provide acceleration services for the target domain name. As can be seen, to determine the provider state, it is still necessary to determine whether the first CDN node fails.

[0069] Specifically, through the above step S42, it can be understood that the first CDN node and the second CDN node can be understood as the same CDN node. Since the same CDN node provides acceleration services for different domain names based on the same operator network, if the access for one of the domain names fails, the access for other domain names will also fail. Therefore, whether the target domain name is in the CDN node library or not, the historical node state of the second CDN node in the CDN library can be used to measure the historical ability of the first CDN node to provide acceleration services for the target domain name. The historical ability of the first CDN node to provide acceleration services for the target domain name can be used to measure the current ability of the first CDN node to provide acceleration services for the target domain name. According to the current ability of the first CDN node to provide acceleration services for the target domain name, it can be determined whether the first CDN node fails, and the provider state of the first CDN provider can also be determined.

[0070] For example, according to the node state of the second CDN node in the CDN library, it is found that in the domain name access that has occurred before the current time point, the client and the second CDN node have both failed to connect, or the connection success rate is too low, then it can be determined that before the current time point, the first CDN node cannot provide acceleration services for the target domain name. It can be reasonably inferred that the first CDN node cannot currently provide acceleration services for the target domain name, that is, the first CDN node fails. In this way, it can be determined that the first CDN provider is abnormal.

[0071] In step S43, the supplier state of the first CDN supplier under each region+operator can be determined respectively according to the region+operator dimension. For example, the supplier state of the first CDN supplier under region A+operator A is determined according to the node state of the second CDN node under region A+operator A in the CDN node library; the supplier state of the first CDN supplier under region B+operator B is determined according to the node state of the second CDN node under region B+operator B in the CDN node library.

[0072] In step S44, if the supplier state represents that the first CDN supplier is abnormal, the first CDN supplier currently used by the target domain name is switched to the second CDN supplier.

[0073] Specifically, the first CDN supplier currently used by the target domain name can be switched according to the region+operator dimension. For example, if the first CDN supplier is abnormal under region A+operator A, the first CDN supplier currently used by the target domain name can be switched to the second CDN supplier under region A+operator A. In this way, the CDN supplier used by the target domain name is not switched in a large area, and the normal region of the first CDN supplier can still preferentially use the first CDN node to provide acceleration service for the target domain name.

[0074] In summary, in the technical scheme of some embodiments of the present disclosure, first, since the CDN node library stores the CDN nodes providing acceleration service for one or more domain names in the historical domain name access and the node state of each CDN node when providing acceleration service, after the first CDN node used to provide acceleration service for the target domain name is obtained, the supplier state of the first CDN supplier can be determined according to the historical node state of the second CDN node matched with the first CDN node in the CDN node library, and then according to the supplier state, it can be determined whether to switch the first CDN supplier currently used by the target domain name. This is based on the historical domain name access information to switch and manage the CDN supplier currently used by the target domain name, which reduces the number of probes, so that when the CDN supplier is switched and managed, the additional traffic generated can be reduced, and the switching cost is reduced.

[0075] Secondly, in the method of the present disclosure, even if the target domain name does not have corresponding domain name access information in the disaster recovery switching system, the CDN supplier used by the target domain name can be managed for disaster recovery switching, which has stronger applicability.

[0076] Again, the method of the present disclosure determines the provider state of the first CDN provider based on historical domain name access information, so that a few or no probe nodes need to be deployed, greatly reducing the deployment cost of probe nodes. At the same time, the problem of low timeliness of disaster recovery switching caused by network jitter is avoided, so that disaster recovery switching is more timely.

[0077] Finally, the CDN nodes in the CDN node library of the present disclosure can be more comprehensive, and for all or most of the first CDN nodes, a second CDN node matching the first CDN node can be found in the CDN node library, so that comprehensive node state analysis can be performed on the first CDN nodes under the first CDN provider, so that the determined provider state is more accurate.

[0078] The scheme of the present disclosure is further described below.

[0079] In some embodiments, the obtaining of the first CDN provider currently used by the target domain name and the first CDN node under the first CDN provider for providing acceleration services for the target domain name in step S41 can include:

[0080] 1) Obtain a client IP for accessing the target domain name.

[0081] Specifically, one or more IPs under each operator in each region can be obtained as the client IP according to the region+operator dimension. For example, IP1 and IP2 are obtained under operator A in region A, IP3 and IP4 are obtained under operator B in region A, IP5 and IP6 are obtained under operator A in region B, and so on.

[0082] 2) Based on the client IP and the target domain name, construct and send a domain name resolution request.

[0083] In this step, the domain name switching system simulates the scenario of the client accessing the target domain name based on the client IP and the target domain name. The domain name switching system can send a domain name resolution request to the domain name resolution server based on the HTTPDNS protocol or the EDNS protocol. The domain name resolution request can include the client IP and the target domain name. Based on the received client IP, the domain name resolution server can determine the region A and the operator network A where the client is located, and then determine the CNAME currently pointed to by the target domain name in the corresponding region A+operator A. The CNAME is the CNAME of the first CDN provider currently used by the target domain name in the region A+operator A. Resolving the CNAME can obtain the IP of one of the first CDN nodes under the first CDN provider for providing acceleration services for the target domain name.

[0084] 3) extracting the first CDN node and the first CDN provider from the response message of the domain name resolution request.

[0085] Specifically, the response message can include the CNAME of the first CDN provider parsed by the domain name resolution server and the IP of one of the first CDN nodes under the first CDN provider. Since a first CDN provider can include multiple first CDN nodes for providing acceleration services for a target domain name under a region + operator, and the disaster recovery switching system can only obtain the IP of one of the first CDN nodes for each domain name resolution request, it is obvious that the IP of the first CDN node obtained in this way is incomplete. Therefore, the disaster recovery switching system can repeatedly send domain name resolution requests based on the same client IP and target domain name to obtain all or most of the first CDN nodes of the first CDN provider under a region + operator for providing acceleration services for the target domain name. For example, assuming that a client IP1 is obtained under the operator A in the region A. Based on the IP1 and the target domain name, repeatedly constructing and sending domain name resolution requests can obtain the IP of multiple first CDN nodes of the first CDN provider under the operator A in the region A for providing acceleration services for the target domain name.

[0086] At this point, how to obtain the first CDN provider currently used by the target domain name and the first CDN node under the first CDN provider for providing acceleration services for the target domain name is described.

[0087] In some embodiments, according to the description in Table 1, the CDN node library can include the node status of each second CDN node in multiple statistical periods. Therefore, determining the provider status of the first CDN provider according to the node status of the second CDN node in the CDN node library in step S43 can include:

[0088] determining whether each second CDN node fails according to the node status of the second CDN node in the last statistical period, respectively;

[0089] If the proportion of the second CDN nodes that fail in the second CDN nodes exceeds the first proportion threshold, it is determined that the first CDN provider is abnormal.

[0090] Specifically, the last statistical period refers to the statistical period closest to the current time point. Since the last statistical period is closest to the current time point, the node status of the second CDN node in the last statistical period can be close to the node status of the second CDN node at the current time point. Therefore, the result of determining whether the second CDN node fails can also be used to measure whether the second CDN node fails at the current time point.

[0091] It should be noted that in the CDN node library, the second CDN nodes with the same region + CDN provider + operator + IP can be considered as the same second CDN node. If a second CDN node has multiple domain name access records in the last statistical period, the node status in these domain name access records can be calculated comprehensively (such as average calculation, weighted calculation), and according to the result of comprehensive calculation, it is determined whether the second CDN node fails. For ease of understanding, it is assumed that the last statistical period is “202209031206”, and the CDN node library in Table 3 is referred to:

[0092] Table 3 CDN node library

[0093] Assuming that the CDN nodes in Table 3 are all found second CDN nodes, the two second CDN nodes under the region A + CDN provider A + operator A + 1.1.1.1 can be considered as the same second CDN node A, and the second CDN node under the region A + CDN provider B + operator B + 1.1.1.1 can be considered as another second CDN node B. Taking the second CDN node A under the region A + CDN provider A + operator A + 1.1.1.1 as an example, the node status in the two domain name access records under the second CDN node A can be evaluated comprehensively (such as average calculation of each node status), and then it is determined whether the second CDN node A fails. If the proportion of the failed second CDN nodes exceeds the first proportion threshold, it can be determined that the first CDN provider is abnormal.

[0094] In some embodiments, according to the description of the CDN node library, the node status can include the connection time of the second CDN node and the client initiating the domain name access. Therefore, the above determination of whether each second CDN node fails according to the node status of the second CDN node in the last statistical period can include:

[0095] For one of the second CDN nodes, if the connection time of the CDN node and the client exceeds the time threshold in the last statistical period, it is determined that the CDN node fails.

[0096] Since the connection time is a communication quality indicator of the client and the CDN node, according to the connection time to determine whether the second CDN node fails, the result is more accurate.

[0097] Specifically, referring to the second CDN node under the region A+CDN provider A+operator A+1.1.1.1 in Table 3, if a second CDN node has multiple domain name access records in the last statistical period, the connection time of the multiple domain name access records can be calculated (such as average calculation, weighted calculation). If the result of the comprehensive calculation exceeds the time domain name, it is determined that the second CDN node is faulty.

[0098] In some embodiments, the time threshold described above can be obtained based on the following method:

[0099] For one of the second CDN nodes, the connection time of the CDN node with the client in multiple statistical periods is averaged, and the result of the calculation is used as the time threshold for fault judgment of the CDN node.

[0100] For example, the connection time of the second CDN node A in multiple statistical periods is averaged to obtain the time threshold for fault judgment of the second CDN node A; the connection time of the second CDN node B in multiple statistical periods is averaged to obtain the time threshold for fault judgment of the second CDN node B.

[0101] The multiple statistical periods for calculating the time threshold can be the last K-1, K-2, K-3 statistical periods. K is an integer.

[0102] In the above embodiments, the time threshold for fault judgment of each second CDN node is determined according to the connection time of each second CDN node in multiple statistical periods. In this way, the setting of the time threshold is more in line with the actual situation of each second CDN node, and the result of fault judgment is more accurate. For example, for the second CDN node A, the connection time is normally 0.5 seconds, but for the second CDN node B, the connection time is normally 0.3 seconds. If the second CDN node A and the second CDN node B share the same time threshold, it is likely that one of the second CDN nodes will be misjudged. For example, in the case of a time threshold of 0.4 seconds, the second CDN node B can be accurately judged to be faulty, but it is likely that the second CDN node A will be misjudged.

[0103] In some embodiments, according to the description of the CDN node library, the node status can include the connection success rate of the second CDN node with the client; therefore, the above-mentioned determination of whether each second CDN node fails according to the node status of the second CDN node in the last statistical period includes:

[0104] For one of the second CDN nodes, if the connection success rate of the CDN node with the client in the last statistical period is lower than the success rate threshold, it is determined that the CDN node fails.

[0105] The process is similar to the process of determining whether the second CDN node fails according to the connection duration described above, and is not described here.

[0106] In some embodiments, as can be known from the description of step S42, when searching for a second CDN node matching the first CDN node from the CDN node library, there can be some first CDN nodes that cannot find a second CDN node matching them from the CDN node library. It can be understood that if there are many first CDN nodes that cannot find a second CDN node matching them from the CDN node library, the node status of the second CDN node cannot objectively reflect the node status of the overall first CDN node, and further, the supplier status determined according to the node status of the second CDN node is also inaccurate. In view of this, the determination of the supplier status of the first CDN supplier according to the node status of the second CDN node in the CDN node library in step S43 can include:

[0107] In the case where some first CDN nodes match the CDN nodes in the CDN node library, the first number of first CDN nodes that have a matching relationship with the CDN nodes in the CDN node library is obtained, and the second number of first CDN nodes that do not have a matching relationship with the CDN nodes in the CDN node library is obtained.

[0108] In the case where the ratio of the first number and the second number reaches a second proportion threshold, the supplier status of the first CDN supplier is determined according to the node status of the second CDN node in the CDN node library.

[0109] The ratio of the first number and the second number can also be referred to as the coverage rate of the first CDN node in the CDN node library.

[0110] The second proportion threshold can be set according to actual conditions, such as 80% and the like. In a case where the ratio of the first quantity and the second quantity reaches the second proportion threshold, it is indicated that the first CDN nodes that have a matching relationship with the CDN nodes in the CDN node library are more, the coverage of the first CDN nodes in the CDN node library is higher, and the second CDN node can accurately determine the node states of the first CDN nodes as a whole. Therefore, the supplier state of the first CDN provider can be determined only according to the node states of the second CDN node in the CDN node library.

[0111] However, in a case where the ratio of the first quantity and the second quantity does not reach the second proportion threshold, the above determination of the supplier state of the first CDN provider can include:

[0112] sending a probing message to at least part of the first CDN nodes to obtain a communication index for representing a communication quality;

[0113] generating a probing result representing whether the first CDN nodes are faulty based on the communication index;

[0114] determining the supplier state of the first CDN provider according to the probing result and the node state of the second CDN node in the CDN node library.

[0115] Specifically, the communication index can refer to that described in FIG. 1, which is not described herein again. The probing result can be used to assist in determining the supplier state. Therefore, when the probing message is sent to the first CDN nodes, a small number of first CDN nodes can be selected for probing, and it is not necessary to probe all the first CDN nodes. It should be noted that the present disclosure scheme will only perform probing in a case where the ratio of the first quantity and the second quantity does not reach the second proportion threshold. Therefore, compared with the scheme shown in FIG. 1, the additional traffic generated during disaster recovery switching management can still be greatly reduced.

[0116] In the above embodiment, the probing result can include the node states of the first CDN nodes obtained by probing. The node states obtained by probing and the node state of the second CDN node in the CDN node library can be fused and calculated to determine the supplier state of the first CDN provider. In this way, the accuracy of the determined supplier state is improved.

[0117] Further, the ratio of the first quantity and the second quantity under each region + operator can be counted respectively according to the region + operator dimension, and the first CDN node is tested for the specific region and the specific operator whose ratio of the first quantity and the second quantity does not reach the second proportion threshold. For example, taking region A + operator A as an example. The first CDN node under region A + operator A can be obtained first. For the convenience of distinguishing description, the first CDN node under region A + operator A will be called regional CDN node in the following. Based on the regional CDN node, the first quantity of the regional CDN nodes that have a matching relationship with the CDN nodes in the CDN node library and the second quantity of the regional CDN nodes that do not have a matching relationship with the CDN nodes in the CDN node library can be obtained, and in the case that the ratio of the first quantity and the second quantity does not reach the second proportion threshold, part of the regional CDN nodes under region A + operator A are probed (i.e., the probe message is sent). In this way, the number of probes is reduced, and the additional traffic generated in the disaster recovery switching management process is reduced.

[0118] In some embodiments, according to the description of the CDN node library, the node status of the second CDN node in the CDN node library is dynamically updated according to the statistical period. After the first CDN provider currently used by the target domain name is switched to the second CDN provider, the method of the present disclosure can further include:

[0119] According to the latest node status of the second CDN node in the CDN node library, the provider status of the first CDN provider is re-determined.

[0120] If the re-determined provider status represents that the first CDN provider is normal, the second CDN provider currently used by the target domain name is switched to the first CDN provider.

[0121] According to the latest node status, the process of re-determining the provider status is similar to the above description, which is not described here.

[0122] Similarly, the provider status of the first CDN provider under each region + operator can be re-determined according to the region + operator dimension. If the provider status of the first CDN provider under one region + operator becomes normal, the second CDN provider currently used by the target domain name under the region + operator is switched to the first CDN provider. In this way, it is ensured that the target domain name under the region + operator can use the first CDN provider in time.

[0123] In some embodiments, the above re-determination of the provider status of the first CDN provider according to the latest node status of the second CDN node in the CDN node library can include:

[0124] Based on the latest node status of the second CDN node in the CDN node database, determine whether each second CDN node has failed.

[0125] If the proportion of second CDN nodes that have not experienced a failure exceeds the second proportion threshold, the first CDN provider is determined to be normal.

[0126] The relevant principles are similar to those described above regarding how to determine the anomaly of the first CDN provider, and will not be repeated here.

[0127] Furthermore, in some embodiments, switching the target domain's currently used second CDN provider to the first CDN provider may include:

[0128] In each statistical period after switching from the first CDN provider to the second CDN provider, the provider status of the first CDN provider is redefined.

[0129] If the redefined supplier status indicates that the first CDN supplier is functioning normally within multiple consecutive statistical periods, then the second CDN supplier currently used by the target domain name will be switched to the first CDN supplier.

[0130] For example, after switching from the first CDN provider to the second, if the re-established provider status indicates that the first CDN provider is functioning normally for five consecutive statistical periods, then the target domain's current second CDN provider will be switched back to the first CDN provider. This ensures that the first CDN node under the first CDN provider is stable before switching CDN providers, avoiding frequent switching between CDN providers.

[0131] Referring to Figure 5, this is an interactive diagram illustrating the disaster recovery switching of the CDN provider used by a newly configured domain name, according to an embodiment of this disclosure.

[0132] In Figure 5, "configured domain name" refers to a domain name configured in the disaster recovery failover system that has already obtained domain name access information. "Newly configured domain name" refers to a domain name configured during disaster recovery failover but for which domain name access information has not yet been obtained. The disaster recovery failover system provides quality center services and disaster recovery failover services.

[0133] The principle behind disaster recovery switching for the CDN provider used by the newly configured domain name can be summarized as follows:

[0134] 1) If a client initiates a domain access request for a configured domain, the client can obtain the domain access information of the configured domain, perform data cleaning and statistics on this domain access information, and then save it to the CDN node library.

[0135] 2) The quality center service calculates the communication quality indicators, such as connection time and connection success rate, of the client when communicating with each CDN node according to the domain name access information in the CDN node library;

[0136] 3) The quality center service obtains the first CDN provider currently used by the newly configured domain name and the first CDN node under the first CDN provider for providing acceleration services for the newly configured domain name, and determines the second CDN node matching the first CDN node in the CDN node library and the coverage of the first CDN node in the CDN node library;

[0137] 4) When the coverage of the first CDN node in the CDN node library is low, the quality center service can issue a probing task to the probing node through the disaster recovery switching service, and the probing node probes part of the first CDN node to obtain the communication quality indicators of the first CDN node;

[0138] 5) The quality center service receives the communication quality indicators through the disaster recovery switching service;

[0139] 6) If the coverage of the first CDN node in the CDN node library is low, the quality center service determines whether the first CDN provider currently used by the newly configured domain name is abnormal based on the communication quality indicators and the node state of the second CDN node in the CDN node library; if the coverage of the first CDN node in the CDN node library is high, the quality center service determines whether the first CDN provider currently used by the newly configured domain name is abnormal based on the node state of the second CDN node in the CDN node library;

[0140] 7) If the quality center service determines that the first CDN provider currently used by the newly configured domain name is abnormal, the quality center service switches the first CDN provider currently used by the newly configured domain name to the second CDN provider through the disaster recovery switching service.

[0141] Corresponding to the CDN provider switching method, the disclosure also provides a CDN provider switching system. In combination with FIG. 6, a module schematic diagram of the switching system provided by an embodiment of the disclosure is shown. In FIG. 6, the switching system includes:

[0142] The CDN node obtaining module is configured to obtain the first CDN provider currently used by the target domain name and the first CDN node under the first CDN provider for providing acceleration services for the target domain name;

[0143] The CDN node screening module is configured to find the second CDN node matching the first CDN node from the CDN node library, wherein the CDN node library is configured to save the CDN nodes providing acceleration services for one or more domain names in the historical domain name access and the node state of each CDN node when providing acceleration services.

[0144] a supplier state determining module configured to determine a supplier state of the first CDN supplier according to the node states of the second CDN nodes in the CDN node library;

[0145] a supplier switching module configured to switch the first CDN supplier currently used by the target domain name to the second CDN supplier if the supplier state indicates that the first CDN supplier is abnormal.

[0146] In some embodiments, the CDN node library comprises node states of the second CDN nodes in a plurality of statistical periods; and the supplier state determining module is specifically configured to:

[0147] determine whether each of the second CDN nodes fails according to the node state of the second CDN node in the last statistical period, respectively;

[0148] if the proportion of the second CDN nodes that fail in the second CDN nodes exceeds the first proportion threshold, determine that the first CDN supplier is abnormal.

[0149] In some embodiments, the node state comprises a connection duration of the second CDN node with a client that initiates a domain name access; and the supplier state determining module is specifically configured to:

[0150] for one of the second CDN nodes, if the connection duration of the CDN node with the client in the last statistical period exceeds a duration threshold, determine that the CDN node fails.

[0151] In some embodiments, the supplier state determining module is specifically configured to determine the duration threshold based on the following method:

[0152] for one of the second CDN nodes, average the connection durations of the CDN node with the client in the plurality of statistical periods, and take the result of the calculation as the duration threshold for judging the failure of the CDN node.

[0153] In some embodiments, the node state comprises a connection success rate of the second CDN node with the client; and the supplier state determining module is specifically configured to:

[0154] for one of the second CDN nodes, if the connection success rate of the CDN node with the client in the last statistical period is lower than a success rate threshold, determine that the CDN node fails.

[0155] In some embodiments, the supplier state determining module is specifically configured to:

[0156] In a case where part of the first CDN nodes match the CDN nodes in the CDN node library, a first quantity of the first CDN nodes that match the CDN nodes in the CDN node library is obtained, and a second quantity of the first CDN nodes that do not match the CDN nodes in the CDN node library is obtained;

[0157] In a case where a ratio of the first quantity and the second quantity reaches a second proportion threshold, a supplier state of the first CDN supplier is determined according to a node state of the second CDN node in the CDN node library.

[0158] In some embodiments, in a case where the ratio of the first quantity and the second quantity does not reach the second proportion threshold, the supplier state determination module is specifically configured to:

[0159] sending a probing message to at least part of the first CDN nodes to obtain a communication indicator for representing a communication quality;

[0160] generating a probing result representing whether the first CDN nodes are faulty based on the communication indicator;

[0161] determining a supplier state of the first CDN supplier according to the probing result and the node state of the second CDN node in the CDN node library.

[0162] In some embodiments, the node state of the second CDN node in the CDN node library is dynamically updated according to a statistical period; after the first CDN supplier currently used by the target domain name is switched to the second CDN supplier, the supplier switching module is specifically configured to:

[0163] redetermining the supplier state of the first CDN supplier according to a latest node state of the second CDN node in the CDN node library;

[0164] if the redetermined supplier state represents that the first CDN supplier is normal, switching the second CDN supplier currently used by the target domain name to the first CDN supplier.

[0165] In some embodiments, the supplier switching module is specifically configured to:

[0166] redetermining the supplier state of the first CDN supplier in each statistical period after the first CDN supplier is switched to the second CDN supplier;

[0167] if the redetermined supplier state represents that the first CDN supplier is normal in a plurality of continuous statistical periods, switching the second CDN supplier currently used by the target domain name to the first CDN supplier.

[0168] In some embodiments, the supplier switching module is specifically configured to:

[0169] determine whether each second CDN node fails according to the latest node status of the second CDN nodes in the CDN node library;

[0170] If the proportion of the second CDN nodes that do not fail in the second CDN nodes exceeds the second proportion threshold, it is determined that the first CDN provider is normal.

[0171] In some embodiments, the CDN node obtaining module is specifically configured to:

[0172] obtain a client IP for accessing the target domain name;

[0173] construct and send a domain name resolution request based on the client IP and the target domain name;

[0174] extract the first CDN node and the first CDN provider from a response message of the domain name resolution request.

[0175] In some embodiments, the CDN node screening module is specifically configured to:

[0176] For any first CDN node, find CDN nodes with the same IP, CDN provider, region and operator as the first CDN node from the CDN node library, and take the found CDN nodes as second CDN nodes matched with the first CDN node.

[0177] Referring to FIG. 7, a schematic diagram of an electronic device provided by an embodiment of the present disclosure is shown. The electronic device includes a processor and a memory, and the memory is configured to store a computer program, which is executed by the processor to implement the method described above.

[0178] The processor can be a central processing unit (CPU). The processor can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations thereof.

[0179] The memory, as a non-transitory computer readable storage medium, can be used to store non-transitory software programs, non-transitory computer executable programs and modules, such as program instructions / modules corresponding to the method in the embodiments of the present application. The processor executes various functions and data processing of the processor by running the non-transitory software programs, instructions and modules stored in the memory, i.e. implements the method in the above-mentioned method embodiments.

[0180] The memory can include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required by a function; and the data storage area can store data created by the processor and the like. In addition, the memory can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include a memory remotely disposed relative to the processor, and these remote memories can be connected to the processor through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0181] One embodiment of the present disclosure further provides a computer readable storage medium for storing a computer program, which, when executed by a processor, implements the above-mentioned method.

[0182] The present disclosure further provides a computer program product comprising a computer program, which, when executed by a processor, implements the above-mentioned method.

[0183] Although the embodiments of the present disclosure are described in conjunction with the accompanying drawings, various modifications and changes can be made by those skilled in the art without departing from the spirit and scope of the present disclosure, and such modifications and changes fall within the scope defined by the appended claims.

Claims

1. A method for switching CDN providers, comprising: Obtain the first CDN provider currently used by the target domain name, and the first CDN node under the first CDN provider used to provide acceleration services for the target domain name; The CDN node database is used to find a second CDN node that matches the first CDN node. The CDN node database is used to store CDN nodes that provide acceleration services for one or more domains in the historical domain access records, as well as the node status of each CDN node when providing acceleration services. The supplier status of the first CDN provider is determined based on the node status of the second CDN node in the CDN node database. If the supplier status indicates that the first CDN supplier is abnormal, the first CDN supplier currently used by the target domain name will be switched to the second CDN supplier.

2. The method as described in claim 1, wherein the CDN node library includes the node status of each of the second CDN nodes within multiple statistical periods; Determining the supplier status of the first CDN provider based on the node status of the second CDN node in the CDN node database includes: Based on the node status of the second CDN node in the last statistical period, determine whether each of the second CDN nodes has failed. If the proportion of faulty second CDN nodes in the second CDN nodes exceeds the first proportion threshold, the first CDN provider is determined to be abnormal.

3. The method as described in claim 2, wherein the node status includes the connection duration between the second CDN node and the client that initiated the domain name access; The step of determining whether each of the second CDN nodes has failed based on the node status of the second CDN node in the last statistical period includes: For one of the second CDN nodes, if the connection duration between the CDN node and the client exceeds the duration threshold in the last statistical period, then the CDN node is determined to have failed.

4. The method of claim 3, wherein the duration threshold is obtained based on the following method: For one of the second CDN nodes, the average connection time between the CDN node and the client over multiple statistical periods is calculated, and the calculated result is used as the time threshold for fault judgment of the CDN node.

5. The method as described in claim 2, wherein the node status includes the connection success rate between the second CDN node and the client; The step of determining whether each of the second CDN nodes has failed based on the node status of the second CDN node in the last statistical period includes: For one of the second CDN nodes, if the success rate of establishing a connection with the client is lower than the success rate threshold in the last statistical period, then the CDN node is determined to have failed.

6. The method of claim 1, wherein determining the supplier status of the first CDN provider based on the node status of the second CDN node in the CDN node library includes: In the case where some of the first CDN nodes match the CDN nodes in the CDN node library, a first number of first CDN nodes that have a matching relationship with the CDN nodes in the CDN node library is obtained, and a second number of first CDN nodes that do not have a matching relationship with the CDN nodes in the CDN node library is obtained. When the ratio of the first quantity to the second quantity reaches the second percentage threshold, the supplier status of the first CDN supplier is determined based on the node status of the second CDN node in the CDN node library.

7. The method of claim 6, wherein when the ratio of the first quantity to the second quantity does not reach the second percentage threshold, determining the supplier status of the first CDN supplier includes: Send test messages to at least some of the first CDN nodes to obtain communication metrics used to characterize communication quality; Based on the communication metrics, a test result is generated to characterize whether the first CDN node is faulty. Based on the test results and the node status of the second CDN node in the CDN node library, the supplier status of the first CDN provider is determined.

8. The method as described in claim 6 or 7, wherein the node status of the second CDN node in the CDN node library is dynamically updated according to a statistical period; After switching the first CDN provider currently used by the target domain name to the second CDN provider, the method further includes: Based on the latest node status of the second CDN node in the CDN node database, the supplier status of the first CDN provider is re-determined; If the redefined supplier status indicates that the first CDN supplier is functioning normally, then the second CDN supplier currently used by the target domain name will be switched to the first CDN supplier.

9. The method of claim 8, wherein switching the second CDN provider currently used by the target domain name to the first CDN provider includes: In each statistical period after the first CDN provider is switched to the second CDN provider, the provider status of the first CDN provider is re-determined. If the redefined supplier status indicates that the first CDN supplier is functioning normally within multiple consecutive statistical periods, then the second CDN supplier currently used by the target domain name will be switched to the first CDN supplier.

10. The method of claim 8, wherein re-determining the supplier status of the first CDN provider based on the latest node status of the second CDN node in the CDN node library comprises: Based on the latest node status of the second CDN node in the CDN node database, determine whether each of the second CDN nodes has failed. If the proportion of second CDN nodes that have not experienced a failure exceeds the second proportion threshold, the first CDN provider is determined to be normal.

11. The method of claim 1, wherein obtaining the first CDN provider currently used by the target domain name, and the first CDN node under the first CDN provider for providing acceleration services to the target domain name, comprises: Obtain the client IP address used to access the target domain name; Based on the client IP and the target domain name, construct and send a domain name resolution request; Extract the first CDN node and the first CDN provider from the response message of the domain name resolution request.

12. The method of claim 1, wherein finding a second CDN node matching the first CDN node from the CDN node library comprises: For any of the first CDN nodes, search the CDN node database for CDN nodes that have the same IP address, CDN provider, region, and carrier as the first CDN node, and use the found CDN nodes as the second CDN nodes that match the first CDN node.

13. A CDN provider switching system, comprising: The CDN node acquisition module is used to acquire the first CDN provider currently used by the target domain name, and the first CDN node under the first CDN provider used to provide acceleration services for the target domain name; The CDN node filtering module is used to find a second CDN node that matches the first CDN node from the CDN node library. The CDN node library is used to store CDN nodes that provide acceleration services for one or more domains in historical domain access records, as well as the node status of each CDN node when providing acceleration services. The supplier status determination module is used to determine the supplier status of the first CDN supplier based on the node status of the second CDN node in the CDN node library. The supplier switching module is used to switch the first CDN supplier currently used by the target domain name to a second CDN supplier if the supplier status indicates that the first CDN supplier is abnormal.

14. A computer-readable storage medium for storing a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 12.

15. An electronic device comprising a processor and a memory, the memory for storing a computer program that, when executed by the processor, implements the method as claimed in any one of claims 1 to 12.

16. A computer program product comprising a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 12.

Images