Cluster management mimicry transformation method based on cloudEon platform

Abstract

The invention relates to a cluster management mimicry transformation method based on a cloudEon platform, which comprises the following steps: a user initiates a new cluster request on a containerized big data management platform cloudEon, and a cluster distribution module receives the request and distributes the request to a heterogeneous kubernates cluster; the heterogeneous kubernates cluster outputs response information and returns the response information to the cluster distribution module; the cluster distribution module sends the response information to the cluster judgment module, and the cluster judgment module judges the response information; the cluster judgment module sends a judgment result to the judgment result output module, and the judgment result output module generates a final response; if the abnormal information exists, the abnormal information is sent to the negative feedback module, the negative feedback module reports the information to the cluster management module, and the cluster management module issues a related instruction to the cluster distribution module. The problem that a big data component managed by a big data management platform is poor in security when facing known or unknown vulnerabilities is solved, the defense capability of the big data component to the known or unknown vulnerabilities is improved, and the security of the component is improved.

Description

Technical Field

[0001] The present invention relates to a network space security technology, and in particular to a cluster management mimicry transformation method based on a containerized big data management platform CloudEon. Background Art

[0002] With the rapid development of information technology, big data components are applied to more and more fields. Behind the prosperous development, the means of attack against big data components are endless, and the impact caused has also risen from network security incidents to production safety accidents. How to ensure the information security and operation and maintenance of big data components has become an urgent problem that needs to be solved.

[0003] Mimicry defense is a new security defense technology proposed in recent years. It is an active defense behavior. Its core is a dynamic heterogeneous redundant structure based on the inherent security mechanism of cyberspace. It can mainly defend against attacks from known or unknown vulnerabilities in the system. The main working principle of mimicry defense technology is: a set F of functionally equivalent heterogeneous executors is used. At time t1, k executors are randomly selected from F to provide input and output services with a judgment mechanism.

[0004] The big data management platform improves the operation and maintenance efficiency and service stability of big data components. However, the current big data management platform has not yet used mimicry defense technology to improve the security of big data components, and faces poor security issues in the face of known or unknown vulnerabilities. Summary of the invention

[0005] Aiming at the problem of poor security of big data components managed by big data management platforms facing known or unknown vulnerabilities, a cluster management mimicry transformation method based on the cloudEon platform is proposed to form a containerized big data management platform with mimicry defense capabilities. Applying mimicry defense technology to the field of containerized big data management platforms improves the defense capabilities of big data components against known or unknown vulnerabilities and enhances the security of big data components.

[0006] The technical solution of the present invention is:

[0007] A cluster management mimetic transformation method based on the cloudEon platform, the mimetic transformation part includes a cluster distribution module, a cluster decision module, a decision result output module, a negative feedback module, a cluster management module, and N heterogeneous kubernates clusters; the method includes the following steps:

[0008] The user initiates a request to add a new cluster on the containerized big data management platform cloudEon. The cluster distribution module receives the request, pre-processes the request, creates N requests with the same requirements, and distributes the request to the heterogeneous kubernates clusters; the heterogeneous kubernates clusters output response information and return the response information to the cluster distribution module; the cluster distribution module sends the response information output by the heterogeneous kubernates cluster to the cluster decision module, and the cluster decision module makes a decision on the received response information according to a certain decision strategy; the cluster decision module sends the decision result to the decision result output module, and the decision result output module generates the final response; if there is abnormal information in the decision result, the cluster decision module sends the abnormal information to the negative feedback module, and the negative feedback module reports the information to the cluster management module, and the cluster management module issues relevant instructions to the cluster distribution module.

[0009] Furthermore, the mimicry transformation specifically includes:

[0010] Cluster distribution module: used to pre-process new cluster service requests initiated by the containerized big data management platform cloudEon, and distribute the requests to heterogeneous kubernates clusters;

[0011] Cluster decision module: used to make decisions on the output responses of heterogeneous kubernates clusters; send the decision results to the decision result output module, and send the decided abnormal information to the negative feedback module;

[0012] Decision result output module: used to receive the decision result of the cluster decision module and output the final response result according to the decision result;

[0013] Negative feedback module: receives the abnormal information determined by the cluster determination module and reports the information to the cluster management module;

[0014] Cluster management module: used to manage heterogeneous kubernates cluster kubeconfig, number and other information; used to receive abnormal information reported by the negative feedback module, and issue instructions to the cluster distribution module based on the abnormal information;

[0015] N heterogeneous kubernates clusters: N groups (N≥3) of computing nodes running containerized big data components; N groups of heterogeneous kubernates clusters are functionally equivalent and are managed by the containerized big data management platform cloudEon.

[0016] Furthermore, the specific steps of the method are as follows:

[0017] Step 1: The user initiates a request to add a new cluster on the containerized big data management platform CloudEon. After receiving the request, the cluster distribution module submits the request data to the buffer queue, takes out the request header, and adds a unified code and signature to the request header; the cluster distribution module obtains the kubeconfig information of the heterogeneous kubernates cluster from the cluster management module, and creates N new cluster requests based on the obtained kubeconfig information, marked as request1, request2, ... requestN respectively; the cluster distribution module sends N requests to N heterogeneous kubernates clusters and proceeds to step 2;

[0018] Step 2: N heterogeneous kubernates clusters generate N responses and return the N responses to the cluster distribution module; the cluster distribution module sends the received N response information to the cluster decision module, and the cluster decision module compares the N responses one by one according to the multi-mode consistency voting algorithm; then proceed to step 3;

[0019] Step 3: record the number of responses with consistent comparison results as M; if the comparison results are consistent, that is, M=N, go to step 4; if the comparison results are inconsistent, that is, M<N, when M≥3, go to step 5; when 0<M<3, go to step 6; when M=0, go to step 9;

[0020] Step 4: Store the information of N heterogeneous kubernates clusters into the relevant database table. The cluster_tag field is used to identify the association relationship of the N heterogeneous kubernates clusters in the database table. The cluster decision module uses a random selection strategy to randomly select a response result from the N responses with consistent comparison results, and sends it to the decision result output module, and then proceeds to step 10.

[0021] Step 5: Record the NM abnormal heterogeneous kubernates cluster numbers with inconsistent comparison results. The cluster decision module sends the abnormal information to the negative feedback module, and the negative feedback module reports the abnormal information to the cluster management module; store the M heterogeneous kubernates cluster information with consistent comparison results in the relevant database table, and the M heterogeneous clusters use the cluster_tag field in the database table to identify the cluster association relationship; the cluster decision module uses a random selection strategy to randomly select a response result from the M consistent responses, and send it to the decision result output module, and enter step 10;

[0022] Step 6: Store the information of the M heterogeneous Kubernetes clusters with consistent comparison results into the relevant database table. The cluster_tag field is used to identify the cluster association relationship of the M heterogeneous clusters in the database table; record the number of the abnormal heterogeneous Kubernetes cluster with inconsistent comparison results, send the abnormal information to the negative feedback module, and the negative feedback module reports the abnormal information to the cluster management module. The cluster management module sends instructions to the cluster distribution module, and then proceeds to step 7;

[0023] Step 7: The cluster distribution module receives the instruction issued by the cluster management module, counts the number of abnormal kubernates clusters L from the instruction, obtains the kubeconfig information of the new L heterogeneous kubernates clusters from the cluster management module, recreates L new cluster requests, and distributes the requests to the L heterogeneous kubernates clusters; proceed to step 8;

[0024] Step 8: L heterogeneous kubernates clusters generate L responses and return them to the cluster distribution module. The cluster distribution module sends the L response information to the cluster decision module. The cluster decision module compares and obtains the number K of consistent comparison results. If M+K≥3, the K heterogeneous kubernates cluster information is stored in the relevant database table. Cluster_tag is also used to identify the association between the K clusters and the M clusters. A response result is randomly selected from the K responses and sent to the decision result output module to enter step 10; if M+K<3, K is recorded as the new M value and step 6 is repeated;

[0025] Step 9: The cluster distribution module sends the specific identifier -1 to the decision result output module, and sends the N heterogeneous kubernates cluster numbers to the negative feedback module, and the negative feedback module reports the abnormal information to the cluster management module; proceed to step 10;

[0026] Step 10: The decision result output module determines whether the received content is a specific identifier-1 according to the received decision result. If so, a preset error response is returned to the user; if not, the received response result is returned to the user.

[0027] The beneficial effects of the present invention are:

[0028] (1) Improved the defense capabilities of big data components managed by the containerized big data management platform cloudEon against known or unknown vulnerabilities.

[0029] (2) Applying mimicry defense technology to containerized big data management platforms expands the application scope of mimicry defense technology.

[0030] (3) Cluster management mimicry transformation based on the containerized big data management platform cloudEon. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 The schematic diagram of cluster management mimicry transformation of the present invention

[0032] Figure 2 A cluster flow chart is newly added for the present invention. DETAILED DESCRIPTION

[0033] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0034] A cluster management mimetic transformation method based on the cloudEon platform, the mimetic transformation part includes a cluster distribution module, a cluster decision module, a decision result output module, a negative feedback module, a cluster management module, and N heterogeneous kubernates clusters.

[0035] Cluster distribution module: used to pre-process new cluster service requests initiated by the containerized big data management platform cloudEon, and distribute the requests to heterogeneous kubernates clusters.

[0036] Cluster decision module: used to make decisions on the output responses of heterogeneous kubernates clusters. The decision results are sent to the decision result output module, and the abnormal information is sent to the negative feedback module.

[0037] Decision result output module: used to receive the decision result of the cluster decision module and output the final response result according to the decision result.

[0038] Negative feedback module: receives abnormal information determined by the cluster determination module and reports the information to the cluster management module.

[0039] Cluster management module: used to manage heterogeneous kubernates cluster kubeconfig, number and other information; used to receive abnormal information reported by the negative feedback module, and issue instructions to the cluster distribution module based on the abnormal information.

[0040] N heterogeneous kubernates clusters: N groups (N≥3) of computing nodes running containerized big data components. N groups of heterogeneous kubernates clusters are functionally equivalent and are managed by the containerized big data management platform cloudEon.

[0041] This solution mainly performs mimicry transformation on the cluster management of cloudEon, a containerized big data management platform. Figure 1 As shown, it can be summarized as follows: the user initiates a request to add a new cluster on the containerized big data management platform cloudEon. The cluster distribution module receives the request, pre-processes the request, creates N requests with the same requirements, and distributes the request to the heterogeneous kubernates cluster. The heterogeneous kubernates cluster outputs response information and returns the response information to the cluster distribution module. The cluster distribution module sends the response information output by the heterogeneous kubernates cluster to the cluster decision module. The cluster decision module makes a decision on the received response information according to a certain decision strategy (this solution uses a majority consensus voting algorithm). The cluster decision module sends the decision result to the decision result output module, and the decision result output module generates the final response. If there is abnormal information in the decision result, the cluster decision module sends the abnormal information to the negative feedback module. The negative feedback module reports the information to the cluster management module, and the cluster management module sends relevant instructions to the cluster distribution module. As shown in Figure 2 As shown, the specific steps are as follows:

[0042] Step 1: The user initiates a request to add a new cluster on the containerized big data management platform CloudEon. After receiving the request, the cluster distribution module submits the request data to the buffer queue, takes out the request header, and adds a unified code and signature to the request header. The cluster distribution module obtains the kubeconfig information of the heterogeneous kubernates cluster from the cluster management module, and creates N new cluster requests based on the obtained kubeconfig information, marked as request1, request2, ... requestN. The cluster distribution module sends N requests to N heterogeneous kubernates clusters and proceeds to step 2.

[0043] Step 2: N heterogeneous kubernates clusters generate N responses and return them to the cluster distribution module. The cluster distribution module sends the received N response information to the cluster decision module, which compares the N responses one by one according to the multi-mode consistency voting algorithm. Go to step 3.

[0044] Step 3: The number of responses with consistent comparison results is recorded as M. If the comparison results are consistent, that is, M = N, go to step 4; if the comparison results are inconsistent, that is, M < N, when M ≥ 3, go to step 5; when 0 < M < 3, go to step 6; when M = 0, go to step 9.

[0045] Step 4: Store the information of N heterogeneous kubernates clusters in the relevant database table. The cluster_tag field is used to identify the association between the N heterogeneous clusters in the database table. The cluster decision module uses a random selection strategy to randomly select a response result from the N responses with consistent comparison results, and sends it to the decision result output module to enter step 10.

[0046] Step 5: Record the NM abnormal heterogeneous kubernates cluster numbers with inconsistent comparison results. The cluster decision module sends the abnormal information to the negative feedback module, and the negative feedback module reports the abnormal information to the cluster management module. The M heterogeneous kubernates cluster information with consistent comparison results is stored in the relevant database table. The cluster_tag field is used to identify the cluster association relationship in the database table. The cluster decision module uses a random selection strategy to randomly select a response result from the M consistent responses and sends it to the decision result output module to enter step 10.

[0047] Step 6: Store the M heterogeneous kubernates cluster information with consistent comparison results into the relevant database table. The cluster_tag field is used to identify the cluster association in the database table. Record the abnormal heterogeneous kubernates cluster number with inconsistent comparison results, send the abnormal information to the negative feedback module, and the negative feedback module reports the abnormal information to the cluster management module. The cluster management module sends instructions to the cluster distribution module and enters step 7.

[0048] Step 7: The cluster distribution module receives the instruction issued by the cluster management module, counts the number of abnormal kubernates clusters L from the instruction, obtains the kubeconfig information of the new L heterogeneous kubernates clusters from the cluster management module, recreates L new cluster requests, and distributes the requests to the L heterogeneous kubernates clusters. Go to step 8.

[0049] Step 8: L heterogeneous kubernates clusters generate L responses and return them to the cluster distribution module. The cluster distribution module sends the L response information to the cluster decision module. The cluster decision module compares and obtains the number K of consistent comparison results. If M+K≥3, the K heterogeneous kubernates cluster information is stored in the relevant database table. Cluster_tag is also used to identify the association between the K clusters and the M clusters. A response result is randomly selected from the K responses and sent to the decision result output module to enter step 10. If M+K<3, K is recorded as the new M value and step 6 is repeated.

[0050] Step 9: The cluster distribution module sends the specific identifier -1 to the decision result output module, and sends the N heterogeneous kubernates cluster numbers to the negative feedback module, and the negative feedback module reports the abnormal information to the cluster management module. Go to step 10.

[0051] Step 10: The decision result output module determines whether the received content is a specific identifier-1 according to the received decision result. If so, a preset error response is returned to the user; if not, the received response result is returned to the user.

[0052] The above-mentioned embodiment only expresses one implementation mode of the present invention, and its description is relatively specific and detailed, but it cannot be understood as limiting the scope of the invention patent. It should be pointed out that for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be based on the attached claims.

Claims

1. A cluster management virtualization transformation method based on the cloudEon platform, characterized in that: The mimicry transformation part includes cluster distribution module, cluster decision module, decision result output module, negative feedback module, cluster management module, and N heterogeneous kubernates clusters; The method comprises the following steps: The user initiates a request to add a new cluster on the containerized big data management platform cloudEon. The cluster distribution module receives the request, pre-processes the request, creates N requests with the same requirements, and distributes the request to the heterogeneous kubernates clusters; the heterogeneous kubernates clusters output response information and return the response information to the cluster distribution module; the cluster distribution module sends the response information output by the heterogeneous kubernates cluster to the cluster decision module, and the cluster decision module makes a decision on the received response information according to a certain decision strategy; the cluster decision module sends the decision result to the decision result output module, and the decision result output module generates the final response; if there is abnormal information in the decision result, the cluster decision module sends the abnormal information to the negative feedback module, and the negative feedback module reports the information to the cluster management module, and the cluster management module issues relevant instructions to the cluster distribution module.

2. According to claim 1, the cluster management virtualization transformation method based on the cloudEon platform is characterized in that: The mimicry transformation specifically includes: Cluster distribution module: used to pre-process new cluster service requests initiated by the containerized big data management platform cloudEon, and distribute the requests to heterogeneous kubernates clusters; Cluster decision module: used to make decisions on the output responses of heterogeneous kubernates clusters; send the decision results to the decision result output module, and send the decided abnormal information to the negative feedback module; Decision result output module: used to receive the decision result of the cluster decision module and output the final response result according to the decision result; Negative feedback module: receives the abnormal information determined by the cluster determination module and reports the information to the cluster management module; Cluster management module: used to manage kubeconfig and numbering information of heterogeneous kubernates clusters; used to receive abnormal information reported by the negative feedback module, and issue instructions to the cluster distribution module based on the abnormal information; N heterogeneous kubernates clusters: N groups of computing nodes running containerized big data components, N ≥ 3; N groups of heterogeneous kubernates clusters are functionally equivalent and are managed by the containerized big data management platform cloudEon.

3. According to claim 1, the cluster management virtualization transformation method based on the cloudEon platform is characterized in that: The specific steps of this method are as follows: Step 1: The user initiates a request to add a new cluster on the containerized big data management platform CloudEon. After receiving the request, the cluster distribution module submits the request data to the buffer queue, takes out the request header, and adds a unified code and signature to the request header; The cluster distribution module obtains the kubeconfig information of the heterogeneous kubernates cluster from the cluster management module, and creates N new cluster requests based on the obtained kubeconfig information, marked as request1, request2, ...requestN respectively; the cluster distribution module sends the N requests to the N heterogeneous kubernates clusters and goes to step 2; Step 2: N heterogeneous Kubernetes clusters generate N responses and return the N responses to the cluster distribution module; The cluster distribution module sends the received N response information to the cluster decision module, and the cluster decision module compares the N responses one by one according to the multi-mode consistency voting algorithm; then proceeds to step 3; Step 3: record the number of responses with consistent comparison results as M; if the comparison results are consistent, that is, M=N, go to step 4; if the comparison results are inconsistent, that is, M<N, when M≥3, go to step 5; when 0<M<3, go to step 6; when M=0, go to step 9; Step 4: Store the information of N heterogeneous Kubernetes clusters into the relevant database table. The cluster_tag field is used in the database table to identify the relationship between the N heterogeneous Kubernetes clusters. The cluster decision module uses a random selection strategy to randomly select a response result from N responses with consistent comparison results, and sends it to the decision result output module, and then goes to step 10; Step 5: Record the numbers of NM abnormal heterogeneous kubernates clusters with inconsistent comparison results. The cluster adjudication module sends the abnormal information to the negative feedback module, and the negative feedback module reports the abnormal information to the cluster management module. Store the information of the M heterogeneous kubernates clusters with consistent comparison results into the relevant database table. The cluster_tag field is used in the database table to identify the association relationship of the M heterogeneous clusters. The cluster decision module uses a random selection strategy to randomly select a response result from the M responses with the same result, and sends it to the decision result output module, and then goes to step 10; Step 6: Store the information of the M heterogeneous Kubernetes clusters with consistent comparison results into the relevant database table. The cluster_tag field is used to identify the cluster association relationship of the M heterogeneous clusters in the database table; record the number of the abnormal heterogeneous Kubernetes cluster with inconsistent comparison results, send the abnormal information to the negative feedback module, and the negative feedback module reports the abnormal information to the cluster management module. The cluster management module sends instructions to the cluster distribution module, and then proceeds to step 7; Step 7: The cluster distribution module receives the instruction issued by the cluster management module, counts the number of abnormal Kubernetes clusters L from the instruction, obtains the kubeconfig information of the new L heterogeneous Kubernetes clusters from the cluster management module, recreates L new cluster requests, and distributes the requests to the L heterogeneous Kubernetes clusters; Go to step 8; Step 8: L heterogeneous kubernates clusters generate L responses and return them to the cluster distribution module. The cluster distribution module sends the L response information to the cluster decision module. The cluster decision module compares and obtains the number K of consistent comparison results. If M+K≥3, the K heterogeneous kubernates cluster information is stored in the relevant database table. Cluster_tag is also used to identify the association between the K clusters and the M clusters. A response result is randomly selected from the K responses and sent to the decision result output module to enter step 10; if M+K<3, K is recorded as the new M value and step 6 is repeated; Step 9: The cluster distribution module sends the specific identifier -1 to the decision result output module, and sends the N heterogeneous kubernates cluster numbers to the negative feedback module, and the negative feedback module reports the abnormal information to the cluster management module; Go to step 10; Step 10: The decision result output module determines whether the received content is a specific identifier-1 according to the received decision result, and if so, returns a preset error response to the user; If not, the received response result is returned to the user.

Images