A method for implementing a multi-cloud hybrid management platform based on mimicry defense

Abstract

The present invention relates to a method for implementing a multi-cloud hybrid management platform based on mimetic defense. A mimetic arbitration framework is introduced for heterogeneous distributed key-value storage libraries. The controller management component completes resource information synchronization through various client components, and transmits control signals to the cloud base, which then monitors the resources in the distributed key-value storage library. The cloud base sends input data streams to the distributed key-value storage library and monitors. When the controller management component monitors resource changes through the cloud base, it triggers a feedback callback of the cloud base, which performs a comparison. The controller operates the corresponding resources according to the operation information transmitted by the cloud base, and synchronously transmits the resource configuration that needs to be updated to the cloud base. After the cloud base confirms that the controller has successfully operated on the resources, it stores the resource configuration transmitted by the controller in the distributed key-value storage library. This solves the problem of the risk of targeted exploitation in the cloud platform management mechanism and strengthens the platform's inherent security.

Description

Technical Field

[0001] The present invention relates to a cyberspace security technology, and in particular to a method for implementing a multi-cloud hybrid management platform based on mimicry defense. Background Art

[0002] With the current focus on AI, large models, and high computing power, both domestically and internationally, cloud platforms are widely used as portals for computing resource scheduling. They provide management of cloud services, unified deployment of core computing power, unified management of application lifecycles, and control over personnel security access. This demonstrates the crucial role of cloud platforms in the intelligent era.

[0003] In the current cloud platform security architecture system, most mainstream cloud platforms adopt traditional security defense. Since traditional security defense methods are based on the accumulation of prior knowledge, they cannot effectively defend against the various unknown security threats in cyberspace.

[0004] Mimetic space defense technology is a new type of defense technology. By constructing a mimetic endogenous security defense environment based on endogenous security theory and dynamic heterogeneous redundant structure DHR design inside the target object, it controls differential mode vulnerabilities, backdoors, viruses and Trojans, as well as attackers who use them as attack resources, and controls malicious attacks within a given threshold range.

[0005] Because mimicry defense is dynamic, heterogeneous, diverse, and randomized, it can take defensive measures before an attack occurs, improving system security, increasing the difficulty for attackers, and effectively responding to unknown attacks. The dynamic and randomized nature of redundant executors enhances the system's inherent security and reduces the uncertainty of actions involving multiple parties that require consistency or coordination. The multi-executor adjudication process, through its principles and architecture, increases the difficulty of coordinated attacks under non-cooperative conditions.

[0006] In the cloud platform management mechanism, a registrable client component with a local cache and indexing mechanism is established during component initialization. This component acts as an adapter for specifying resources in the distributed key-value store and obtaining synchronization information. It is used to synchronously observe resource changes in the distributed key-value store, trigger user-registered processing callbacks, and provide an interface for updating resources in the distributed key-value store.

[0007] The cloud service component management mechanism only monitors resource changes within the component and synchronizes the configuration of the corresponding resource states to achieve final state consistency. This synchronization process only considers the synchronization of component information with the cloud service, without ensuring the security of the transmission process. This process poses a risk of targeted exploitation and is susceptible to tampering or destruction by attackers. Summary of the Invention

[0008] Aiming at the risk of targeted exploitation in the cloud platform management mechanism, a multi-cloud hybrid management platform implementation method based on mimicry defense is proposed.

[0009] The technical solution of the present invention is:

[0010] A method for implementing a multi-cloud hybrid management platform based on mimicry defense introduces a mimicry adjudication framework for heterogeneous distributed key-value storage repositories during the cloud base's monitoring of resources in the distributed key-value storage repositories. The method for implementing the multi-cloud hybrid management platform is as follows:

[0011] Step 1: The controller management component synchronizes resource information with each client component and transmits control signals to the cloud base, which then monitors the resources in the distributed key-value store.

[0012] Step 2: The cloud base sends the input data stream InputStream to the distributed key-value storage library and waits for the distributed key-value storage library to provide feedback on the output data stream OutputStream. The output data stream with a fixed format will indicate the resource changes of the distributed key-value storage library. Based on the above process, the cloud base can complete the monitoring of the distributed key-value storage library.

[0013] Step 3: When the controller management component monitors resource changes in the distributed key-value store through the output stream Outstream T received by the cloud base, it triggers a feedback callback of the cloud base;

[0014] Step 4: The cloud base records the feedback results of the changes, compares them internally, and then sends the confirmed operation information to the controller management component, requesting the controller to operate the corresponding resources;

[0015] Step 5: The controller operates the corresponding resources according to the operation information transmitted by the cloud base, and transmits the feedback information of the operation completion to the cloud base. At the same time, the controller will synchronously transmit the resource configuration that needs to be updated to the cloud base;

[0016] Step 6: After the cloud base confirms that the controller successfully operates on the resource, it stores the resource configuration passed by the controller in the distributed key-value storage library.

[0017] Further, the specific steps are as follows:

[0018] Step 1: The controller management component synchronizes resource information with each client component and transmits control signals to the cloud base, which then monitors the resources in the distributed key-value store.

[0019] Step 1-1: Build a heterogeneous software and hardware platform based on the characteristics of the cloud platform;

[0020] Step 1-2: Perform heterogeneous transformation on distributed key-value storage libraries based on software and hardware platforms with different architectures, and establish N distributed key-value storage libraries with the same functions but different architectures or underlying instruction sets. After the construction is completed, ensure that the resources in each distributed key-value storage library are consistent.

[0021] Step 2: The cloud base sends the input data stream InputStream to the distributed key-value storage library and waits for the distributed key-value storage library to provide feedback on the output data stream OutputStream. The output data stream with a fixed format will indicate the resource changes of the distributed key-value storage library. Based on the above process, the cloud base can complete the monitoring of the distributed key-value storage library.

[0022] Step 2-1: The cloud base monitoring request is sent to the distributed key-value storage library in the form of a data stream InputStream. The input agent in the mimic arbitration framework will intercept the data stream and distribute it to generate N monitoring request data streams with the same structure and content. At the same time, during the distribution process, the input agent will sequence the data streams respectively, corresponding to sequence numbers 1-N;

[0023] Step 2-2: The monitoring request data stream sequenced by the input agent will be sent by the input agent to the distributed key-value storage library of the corresponding sequence number. After receiving the corresponding requests, the distributed key-value storage libraries of different sequence numbers will generate feedback information and send it to the output agent in the mimic arbitration framework;

[0024] Step 2-3: The output agent makes a unified decision on each distributed key-value store using a voting model, with the majority of the results being considered the true result. Specifically, when the output data streams OutStream1, OutStream2, ..., OutStreamN arrive at the output agent, if more than 66% of the output data streams have the same data content (Outstream T), Outstream T is confirmed as the true result and the process proceeds normally to Step 2-4. The minority of data streams that are inconsistent with Outstream T are marked as abnormal results and discarded to Step 2-5.

[0025] Step 2-4: The output agent feeds the output data stream OutStream T obtained in step 2-3 as the actual result back to the cloud base, and then proceeds to step 3;

[0026] Step 2-5: The output agent reports the sequence number of the distributed key-value storage library with the abnormal result in the output stream Outstream to the feedback control scheduler in the mimic arbitration framework. The feedback control scheduler rotates or cleans the abnormal distributed key-value storage library.

[0027] Step 3: When the controller management component monitors resource changes in the distributed key-value store through the output stream Outstream T received by the cloud base, it triggers a feedback callback of the cloud base;

[0028] Step 4: The cloud base records the feedback results of the changes, compares them internally, and then sends the confirmed operation information to the controller management component, requesting the controller to operate the corresponding resources;

[0029] Step 5: The controller operates the corresponding resources according to the operation information transmitted by the cloud base, and transmits the feedback information of the operation completion to the cloud base. At the same time, the controller will synchronously transmit the resource configuration that needs to be updated to the cloud base;

[0030] Step 6: After the cloud base confirms that the controller has successfully operated on the resource, it stores the resource configuration passed by the controller in the distributed key-value storage library;

[0031] Step 6-1: At this time, the cloud base update request data stream InputStream will be intercepted and split by the input agent, generating N identical update request data streams, which are sequenced and correspond to sequence numbers 1-N.

[0032] Step 6-2: The update request data stream sequenced by the input agent will be passed by the input agent to the distributed key-value storage library of the corresponding sequence number. After receiving the corresponding request, the distributed key-value storage library will uniformly execute the resource configuration update storage.

[0033] Furthermore, in step 1-1, when applying the mimicking framework, it is necessary to first build a heterogeneous software and hardware platform and deploy the corresponding software and hardware environment based on the characteristics of the cloud platform, such as arm and x86.

[0034] Furthermore, in steps 2-5, rotation refers to scheduling a standby distributed key-value storage library to replace the existing abnormal distributed key-value storage library, and cleaning refers to initializing the abnormal distributed key-value storage library.

[0035] The beneficial effects of the present invention are:

[0036] The traditional defenses used by existing cloud security are based on prior knowledge, requiring a large virus library, vulnerability library, and signature library, and are unable to cover and defend against vulnerabilities that do not exist in the library. Figure 2 As shown in the figure, the cloud platform management mechanism using the technology in this article strengthens the intrinsic security of the platform. The heterogeneous characteristics make the platform security dynamic, diversified, and random, and it can also defend against unknown risks with a high degree of confidence. At the same time, the cleaning and recovery function of the mimetic component ensures that the cloud platform can recover seamlessly when an anomaly occurs. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1This is the overall structure diagram of the controller of the present invention;

[0038] Figure 2 A schematic diagram of a mimic arbitration framework introduced into the controller of the present invention;

[0039] Figure 3 This is a working sequence diagram of the mimicry arbitration framework of the present invention;

[0040] Figure 4 This is a flow chart of the cloud base monitoring the distributed key-value storage library of the present invention. DETAILED DESCRIPTION

[0041] The present invention is described in detail below with reference to 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.

[0042] The controller management component monitors the changes of resources in the distributed key-value storage library and synchronizes the configuration of the corresponding resource status to achieve the consistency of the final state. Its overall structure is as follows: Figure 1 .

[0043] After the client component synchronizes with resource changes, the controller implements the control logic for the corresponding resource based on the resource changes. For example, when the microservice framework component detects changes in network governance configuration, the virtualization service will operate the microservice framework to create or update the network policy.

[0044] The controller includes package management tools, resource clusters, permission control, applications, virtualization services, etc., which provide specific operations for corresponding resource synchronization.

[0045] The controller management component operates the distributed key-value storage library resources through the cloud base. The distributed key-value storage library resources serve as the storage middleware for key-value pairs, and store key information such as the cloud platform's status information and resource location.

[0046] When applying the mimetic framework, you need to first build a heterogeneous software and hardware platform and deploy the corresponding software and hardware environment based on the characteristics of the cloud platform, such as arm, x86, etc.

[0047] Heterogeneous transformation of distributed key-value storage libraries based on different hardware and software architectures. N distributed key-value storage libraries with the same functionality but different architectures or underlying instruction sets are established. After the construction is complete, ensure that the resources within each distributed key-value storage library are consistent.

[0048] In the process of cloud base monitoring the resources in the distributed key-value storage library, a mimic arbitration framework is introduced for the heterogeneous distributed key-value storage library. Figure 2 .

[0049] The specific steps of a multi-cloud hybrid management platform implementation method based on mimicry defense are as follows: Figure 2-4 :

[0050] Step 1: The controller management component completes resource information synchronization through various client components, and transmits control signals to the cloud base, which then monitors the resources in the distributed key-value storage library.

[0051] Step 1-1: Build a heterogeneous software and hardware platform based on the characteristics of the cloud platform, such as ARM, x86, etc.

[0052] Step 1-2: Perform heterogeneous transformation of the distributed key-value store based on different hardware and software architectures. Establish N distributed key-value stores with the same functionality but different architectures or underlying instruction sets. After the construction is complete, ensure that the resources within each distributed key-value store are consistent.

[0053] Step 2: The cloud base sends the input data stream InputStream to the distributed key-value storage library and waits for the distributed key-value storage library to give feedback to the output data stream OutputStream. The output data stream with a fixed format will indicate the resource changes of the distributed key-value storage library. According to the above process, the cloud base can complete the monitoring of the distributed key-value storage library.

[0054] Step 2-1: The cloud base listening request is sent to the distributed key-value storage library in the form of a data stream InputStream. The input agent in the mimic judgment framework will intercept the data stream and distribute the data stream to generate N listening request data streams with the same structure and content. At the same time, during the distribution process, the input agent will sequence the data streams respectively, corresponding to 1-N serial numbers.

[0055] Step 2-2: The monitoring request data stream marked by the input agent will be sent by the input agent to the distributed key-value storage library with the corresponding sequence number (such as data stream 1 corresponds to distributed key-value storage library 1, data stream N corresponds to distributed key-value storage library N). After receiving their respective corresponding requests, the distributed key-value storage libraries with different sequence numbers will generate feedback information and send it uniformly to the output agent in the mimicry arbitration framework.

[0056] Step 2-3: The output agent performs a unified decision on each distributed key-value store, using a voting model. The majority decision is considered the true result. Specifically, when the output data streams OutStream1, OutStream2, …, OutStreamN arrive at the output agent, if more than 66% of the output data streams have the same data content as Outstream T, Outstream T is confirmed as the true result and the process proceeds normally to Step 2-4. The minority of data streams that do not match Outstream T are marked as abnormal results and discarded to Step 2-5.

[0057] Step 2-4: The output agent feeds the output data stream OutStream T obtained in step 2-3 as the actual result back to the cloud base, and then proceeds to step 3.

[0058] Step 2-5: The output agent reports the sequence number of the distributed key-value storage library with the abnormal result in the output stream Outstream to the feedback control scheduler in the mimic arbitration framework. The feedback control scheduler rotates the abnormal distributed key-value storage library (scheduling the backup distributed key-value storage library to replace the existing abnormal distributed key-value storage library) or cleans (initializes the abnormal distributed key-value storage library).

[0059] Step 3: When the controller management component monitors the resource changes in the distributed key-value store through the output stream Outstream T received by the cloud base, it will trigger the feedback callback of the cloud base.

[0060] Step 4: The cloud base records the feedback results of the changes, compares them internally, and then sends the operation information after comparison confirmation to the controller management component, requesting the controller to operate the corresponding resources.

[0061] Step 5: The controller operates the corresponding resources according to the operation information transmitted by the cloud base, and transmits the feedback information of the operation completion to the cloud base. At the same time, the controller will synchronously transmit the resource configuration that needs to be updated to the cloud base.

[0062] Step 6: After the cloud base confirms that the controller successfully operates on the resource, it stores the resource configuration passed by the controller in the distributed key-value storage library.

[0063] Step 6-1: At this time, the cloud base update request data stream InputStream will be intercepted and diverted by the input agent, generating N identical update request data streams, which are sequenced and correspond to 1-N serial numbers.

[0064] Step 6-2: The update request data stream sequenced by the input agent will be passed by the input agent to the distributed key-value storage library of the corresponding sequence number. After receiving the corresponding request, the distributed key-value storage library will uniformly execute the resource configuration update storage.

[0065] The above-described embodiment merely represents one embodiment of the present invention. While the description is relatively specific and detailed, it should not be construed as limiting the scope of the patent. It should be noted that a person skilled in the art would be able to make various modifications and improvements without departing from the spirit of the present invention, and these modifications and improvements fall within the scope of protection of the present invention. Therefore, the scope of protection of the patent for this invention shall be determined by the appended claims.

Claims

1. A method for implementing a multi-cloud hybrid management platform based on mimicry defense, characterized in that: As the cloud base monitors resources in the distributed key-value store, a mimicking arbitration framework is introduced for the heterogeneous distributed key-value store. The multi-cloud hybrid management platform implements this approach as follows: Step 1: The controller management component synchronizes resource information with each client component and transmits control signals to the cloud base, which then monitors the resources in the distributed key-value store. Step 2: The cloud base sends the input data stream InputStream to the distributed key-value storage library and waits for the distributed key-value storage library to provide feedback on the output data stream OutputStream. The output data stream with a fixed format will indicate the resource changes of the distributed key-value storage library. According to the above process, the cloud base can complete the monitoring of the distributed key-value storage library. Step 2 specifically includes: Step 2-1: The cloud base monitoring request is sent to the distributed key-value storage library in the form of a data stream InputStream. The input agent in the mimic arbitration framework will intercept the data stream and distribute it to generate N monitoring request data streams with the same structure and content. At the same time, during the distribution process, the input agent will sequence the data streams respectively, corresponding to sequence numbers 1-N; Step 2-2: The monitoring request data stream sequenced by the input agent will be sent by the input agent to the distributed key-value storage library of the corresponding sequence number. After receiving the corresponding requests, the distributed key-value storage libraries of different sequence numbers will generate feedback information and send it to the output agent in the mimic arbitration framework; Step 2-3: The output agent makes a unified decision on each distributed key-value store using a voting model, with the majority of the results being considered the true result. Specifically, when the output data streams OutStream1, OutStream2, ..., OutStreamN arrive at the output agent, if more than 66% of the output data streams have the same data content Outstream T, then Outstream T is confirmed as the true result and the process proceeds normally to Step 2-4. The minority of data streams that are inconsistent with Outstream T are marked as abnormal results and discarded to Step 2-5. Step 2-4: The output agent feeds the output data stream OutStream T obtained in step 2-3 as the actual result back to the cloud base, and then proceeds to step 3; Step 2-5: The output agent reports the sequence number of the distributed key-value storage library with the abnormal result in the output stream Outstream to the feedback control scheduler in the mimic arbitration framework. The feedback control scheduler rotates or cleans the abnormal distributed key-value storage library. Step 3: When the controller management component monitors resource changes in the distributed key-value store through the output stream Outstream T received by the cloud base, it triggers a feedback callback of the cloud base; Step 4: The cloud base records the feedback results of the changes, compares them internally, and then sends the confirmed operation information to the controller management component, requesting the controller to operate the corresponding resources; Step 5: The controller operates the corresponding resources according to the operation information transmitted by the cloud base, and transmits the feedback information of the operation completion to the cloud base. At the same time, the controller will synchronously transmit the resource configuration that needs to be updated to the cloud base; Step 6: After the cloud base confirms that the controller successfully operates on the resource, it stores the resource configuration passed by the controller in the distributed key-value storage library.

2. The method for implementing a multi-cloud hybrid management platform based on mimicry defense according to claim 1 is characterized in that: Step 1 specifically includes: Step 1-1: Build a heterogeneous software and hardware platform based on the characteristics of the cloud platform; Step 1-2: Perform heterogeneous transformation on distributed key-value storage libraries based on software and hardware platforms with different architectures, and establish N distributed key-value storage libraries with the same functions but different architectures or underlying instruction sets. After the construction is completed, ensure that the resources in each distributed key-value storage library are consistent. Step 6 specifically includes: Step 6-1: At this time, the cloud base update request data stream InputStream will be intercepted and split by the input agent, generating N identical update request data streams, which are sequenced and correspond to sequence numbers 1-N. Step 6-2: The update request data stream sequenced by the input agent will be passed by the input agent to the distributed key-value storage library of the corresponding sequence number. After receiving the corresponding request, the distributed key-value storage library will uniformly execute the resource configuration update storage.

3. The method for implementing a multi-cloud hybrid management platform based on mimicry defense according to claim 2 is characterized in that: In step 1-1, when applying the mimicking framework, you need to first build a heterogeneous software and hardware platform and deploy the corresponding software and hardware environment based on the characteristics of the cloud platform.

4. The method for implementing a multi-cloud hybrid management platform based on mimicry defense according to claim 1 is characterized in that: In steps 2-5, rotation refers to scheduling a standby distributed key-value storage library to replace the existing abnormal distributed key-value storage library, and cleaning refers to initializing the abnormal distributed key-value storage library.

Images