Cloud hard disk mounting and unloading method of Openstack cloud platform based on endogenous security architecture
By introducing adjudication mechanism and secondary thread mechanism into the endogenous security architecture of the Openstack cloud platform, the data consistency problem in cloud hard disk operations is solved, and the efficiency and security of cloud hard disk mount and uninstallation are improved.
Patent Information
- Application Number
- CN202410899724.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-05
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-07-05
AI Technical Summary
In the Openstack cloud platform with an endogenous security architecture, during the mount and uninstallation of cloud hard disk, the data written in the database is prone to chaos, duplication or conflict, affecting the normal operation of cloud hard disk.
By introducing a decision mechanism between the user-side proxy unit and the business-side proxy unit, we ensure the consistency of the execution results of multiple heterogeneous executors, use the nova-compute component to write the BDM table in the database and execute cloud hard disk operations through secondary threads , ensure data consistency and rotate abnormal isomer actuators through feedback control units.
It effectively avoids the confusion, duplication or conflict of data written in the database, ensures data consistency during cloud hard disk mount and uninstallation, and improves operational efficiency and security.
Smart Images

Figure CN119003541A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intrinsic security technology, and in particular to a cloud hard disk mounting and unmounting method of an Openstack cloud platform based on an intrinsic security architecture. Background Art
[0002] The intrinsic security architecture is an architecture with intrinsic security defense capabilities, which aims to transform security from "plug-in" to "intrinsic", and from only being able to defend against known threats to being able to deal with unknown threats. The core framework of the intrinsic security architecture is Dynamic Heterogeneous Redundancy (DHR), which contains multiple heterogeneous executors, each of which can independently calculate the received data.
[0003] With the development of cloud computing, cloud services are increasingly being applied to various industries. How to ensure that cloud platforms can mount and unmount cloud hard disks more quickly, efficiently and stably under the inherent security architecture has become the key to the normal operation and use of cloud hosts in cloud platforms. Openstack cloud platform is a widely used cloud computing management platform. Under the introduction of the inherent security architecture, each time a cloud hard disk is mounted (or unmounted), the corresponding request will be sent to multiple randomly selected heterogeneous executors at the same time. Each heterogeneous executor will independently execute the operation of writing the relevant processing data of the mounted (or unmounted) cloud hard disk to the database according to the received request. In actual situations, the execution progress of multiple randomly selected heterogeneous executors is usually inconsistent, and the execution results may be the same or different. As a result, when writing data to the database, multiple heterogeneous executors may write the same data or different data to the database one after another, resulting in confusion, duplication and even conflict of the data written in the database, affecting the normal mounting (or unmounting) process of the cloud hard disk. Summary of the invention
[0004] The present invention provides a cloud hard disk mounting and unmounting method for an Openstack cloud platform based on an inherent security architecture, so as to solve the defect in the prior art that when a cloud hard disk mounting operation is performed in an Openstack cloud platform with an inherent security architecture, the data written in the database is confused, repeated or conflicting, and the purpose of ensuring the consistency of the data written in the database when the cloud hard disk is mounted in the Openstack cloud platform with an inherent security architecture is achieved.
[0005] The present invention provides a cloud hard disk mounting method for an Openstack cloud platform based on an intrinsic security architecture. The intrinsic security architecture includes a user-side proxy unit, a business-side proxy unit, and m heterogeneous execution bodies. The method includes the following steps.
[0006] The user-side proxy unit receives the first Restful request for mounting a cloud hard disk from the UI; the first Restful request carries the first identifier of the target cloud host and the second identifier of the target cloud hard disk; the user-side proxy unit copies and distributes the first Restful request to n target heterogeneous executors among the m heterogeneous executors; m and n are positive integers, m≥n≥3; the target heterogeneous executor responds to the received first Restful request, and creates BDM information when it is determined that the target cloud host and the target cloud hard disk are both in an available state; the business-side proxy unit receives the RPC Call synchronization message carrying the BDM information, the first identifier and the second identifier from the target heterogeneous executor, and performs a RPC Call synchronization on all target heterogeneous executors. The nova-compute component determines whether the BDM information in the Call synchronization message is consistent; the nova-compute component monitors the rabbitmq message queue to obtain the first message carrying the BDM information, the first identifier and the second identifier sent to the rabbitmq message queue by the business-side agent unit when the first determination result is yes; the nova-compute component writes the BDM table obtained according to the BDM information, the first identifier and the second identifier into the database; when the BDM table is successfully written to the database, the nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor, and executes the operation of mounting the target cloud hard disk in the target cloud host.
[0007] According to a cloud hard disk mounting method of an Openstack cloud platform based on an intrinsic security architecture provided by the present invention, after the nova-compute component writes the BDM table obtained according to the BDM information, the first identifier and the second identifier into the database, it also includes: the nova-compute component creates a secondary thread when the BDM table is successfully written into the database; when the BDM table is successfully written into the database, the nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor, and executes the operation of mounting the target cloud hard disk in the target cloud host, including: the nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor through the current main thread, and executes the operation of mounting the target cloud hard disk in the target cloud host through the secondary thread.
[0008] According to a cloud hard disk mounting method of an Openstack cloud platform based on an intrinsic security architecture provided by the present invention, after creating a secondary thread, the method further includes: the nova-compute component sends a first state change request to change the state of a target cloud hard disk to the cinder-api component through the secondary thread; the cinder-api component sets the state of the target cloud hard disk to a reserved state in response to the received first state change request; the target cloud hard disk in the reserved state is prohibited from being mounted by other cloud hosts other than the target cloud host.
[0009] According to a cloud hard disk mounting method of an Openstack cloud platform based on an intrinsic security architecture provided by the present invention, after creating a secondary thread, the method further includes: the nova-compute component obtains the task lock of the target cloud host through the secondary thread; and when the operation of mounting the target cloud hard disk in the target cloud host executed by the secondary thread is completed, the task lock is released.
[0010] According to a cloud hard disk mounting method of an Openstack cloud platform based on an intrinsic security architecture provided by the present invention, after the nova-compute component obtains the task lock of the target cloud host through a secondary thread, the method also includes: the nova-compute component sends a mount information request to the cinder-api component; the nova-compute component updates the BDM table in the database according to the mount information returned by the cinder-api component in response to the mount information request; wherein the mount information includes: the drive letter information of the target cloud hard disk in the target cloud host.
[0011] According to a cloud hard disk mounting method of an Openstack cloud platform based on an intrinsic security architecture provided by the present invention, after the user-side agent unit copies and distributes the first Restful request to n target heterogeneous executors among m heterogeneous executors, the method further includes: the target heterogeneous executor responds to the received first Restful request, and when it is determined that the target cloud host or the target cloud hard disk is not in an available state, returns an operation conflict message to the user-side agent unit; the user-side agent unit sends an operation conflict response message corresponding to the operation conflict message to the UI for display.
[0012] According to a cloud hard disk mounting method of an Openstack cloud platform based on an intrinsic security architecture provided by the present invention, the intrinsic security architecture also includes: a feedback control unit; after determining whether the BDM information in the RPC Call synchronization message of all target heterogeneous executors is consistent, it also includes: when the first determination result is no, the business side agent unit obtains the identification information of the first abnormal heterogeneous executor in an abnormal state among the n target heterogeneous executors, and sends the identification information of the first abnormal heterogeneous executor to the feedback control unit; the feedback control unit rotates and cleans the first abnormal heterogeneous executor.
[0013] According to a cloud hard disk mounting method of an Openstack cloud platform based on an intrinsic security architecture provided by the present invention, after returning the third identifier of the BDM table to the target heterogeneous executor, the method further includes: a user-side agent unit receives a mounting response message returned by the target heterogeneous executor according to the third identifier received by the target heterogeneous executor; the user-side agent unit determines whether the mounting response messages returned by all target heterogeneous executors are consistent; a second determination result is obtained; if the second determination result is yes, the second determination result is sent to a UI for display; if the second determination result is no, identification information of a second abnormal heterogeneous executor in an abnormal state among all current target heterogeneous executors is obtained, and the identification information of the second abnormal heterogeneous executor is sent to a feedback control unit; and the feedback control unit rotates and cleans the second abnormal heterogeneous executor.
[0014] The present invention also provides a cloud hard disk unloading method of an Openstack cloud platform based on an intrinsic security architecture. The intrinsic security architecture includes a user-side proxy unit, a business-side proxy unit, and m heterogeneous executors. The method includes the following steps.
[0015] The user-side proxy unit receives a second Restful request from the UI for uninstalling the cloud hard disk; the second Restful request carries the first identifier of the target cloud host and the second identifier of the target cloud hard disk; the user-side proxy unit copies and distributes the second Restful request to k target heterogeneous executors among the m heterogeneous executors; m and k are positive integers, m≥k≥3; the target heterogeneous executor responds to the received second Restful request, and when determining that the target cloud host is in an available state, sends an RPC Call asynchronous message carrying the first identifier and the second identifier to the business-side proxy unit; the business-side proxy unit determines whether the RPC Call asynchronous messages of all target heterogeneous executors are consistent; the nova-compute component obtains the second message carrying the first identifier and the second identifier sent to the rabbitmq message queue by the business-side proxy unit when the third determination result is yes by monitoring the rabbitmq message queue; the nova-compute component executes the operation of uninstalling the target cloud hard disk in the target cloud host according to the first identifier and the second identifier.
[0016] According to a cloud hard disk unloading method of an Openstack cloud platform based on an intrinsic security architecture provided by the present invention, a nova-compute component executes an operation of unloading a target cloud hard disk in a target cloud host according to a first identifier and a second identifier, including: the nova-compute component obtains a task lock of the target cloud host according to the first identifier; the nova-compute component obtains a BDM table containing the first identifier and the second identifier from a database according to the first identifier and the second identifier; the nova-compute component unloads the target cloud hard disk from the target cloud host according to the mounting information of the target cloud hard disk in the BDM table; the nova-compute component deletes the BDM table in the database; the nova-compute component sends a second state change request to the cinder-api component; and the cinder-api component sets the state of the target cloud hard disk to an available state in response to the received second state change request.
[0017] According to a cloud hard disk unloading method of an Openstack cloud platform based on an intrinsic security architecture provided by the present invention, the intrinsic security architecture also includes: a feedback control unit; after the business side agent unit determines whether the RPCCall asynchronous messages of all target heterogeneous executors are consistent, it also includes: when the third determination result is no, the business side agent unit obtains the identification information of the third abnormal heterogeneous executor in an abnormal state among all current target heterogeneous executors, and sends the identification information of the third abnormal heterogeneous executor to the feedback control unit; the feedback control unit rotates and cleans the third abnormal heterogeneous executor.
[0018] According to a cloud hard disk unloading method of an Openstack cloud platform based on an intrinsic security architecture provided by the present invention, the target heterogeneous executor responds to the received second Restful request, and after determining that the target cloud host is in an available state, sends an RPC Call asynchronous message carrying a first identifier and a second identifier to a business side agent unit, the method further includes: the target heterogeneous executor returns an unloading response message to the user side agent unit; after the user side agent unit receives the unloading response message, it performs consistency determination on the unloading response messages returned by all target heterogeneous executors to obtain a fourth determination result; when the fourth determination result is yes, the fourth determination result is sent to a UI for display; when the fourth determination result is no, the identification information of a fourth abnormal heterogeneous executor in an abnormal state among all current target heterogeneous executors is obtained, and the identification information of the fourth abnormal heterogeneous executor is sent to a feedback control unit; the feedback control unit rotates and cleans the fourth abnormal heterogeneous executor.
[0019] The present invention provides a cloud hard disk mounting and unmounting method for an Openstack cloud platform based on an intrinsic security architecture. The method can determine whether the execution results of multiple target heterogeneous executors are consistent in a business side agent unit. When the determination result is that the execution results of multiple target heterogeneous executors are consistent, a cloud hard disk mounting operation or a cloud hard disk unmounting operation is performed, thereby avoiding the situation where multiple heterogeneous executors successively write the same data or write different data to a database, resulting in confusion, duplication, or even conflict in the data written in the database. The method achieves the purpose of ensuring the consistency of data written in the database when mounting or unmounting a cloud hard disk in an Openstack cloud platform that introduces an intrinsic security architecture. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0021] Figure 1 This is one of the flow diagrams of the cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture provided by the present invention.
[0022] Figure 2 It is a schematic diagram of the intrinsic security architecture provided by the present invention.
[0023] Figure 3 This is the second flow chart of the cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture provided by the present invention.
[0024] Figure 4 This is one of the flow charts of the cloud hard disk uninstallation method of the Openstack cloud platform based on the intrinsic security architecture provided by the present invention.
[0025] Figure 5 This is the second flow chart of the cloud hard disk uninstallation method of the Openstack cloud platform based on the intrinsic security architecture provided by the present invention.
[0026] Figure 6 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION
[0027] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0028] Combine the following Figure 1-Figure 5 The present invention describes a cloud hard disk mounting and unmounting method of an Openstack cloud platform based on an intrinsic security architecture.
[0029] Figure 1 This is one of the flow diagrams of the cloud hard disk mounting method of the Openstack cloud platform based on the inherent security architecture provided by the embodiment of the present invention. Figure 1 As shown, the method includes the following S110~170.
[0030] S110: The user-side proxy unit receives a first RESTful request for mounting a cloud hard disk from the UI; the first RESTful request carries a first identifier of a target cloud host and a second identifier of a target cloud hard disk.
[0031] Before introducing S110, the intrinsic security architecture provided by the embodiment of the present invention is first introduced.
[0032] like Figure 2 As shown, the content security architecture provided by the embodiment of the present invention includes a user-side agent unit 201 , m (m is a positive integer and m≥3) heterogeneous executors (nova-api) 202 , a service-side agent unit 203 and a feedback control unit 204 .
[0033] The user-side proxy unit 201 is used to receive a user request from a user interface (UI), select n target heterogeneous executors from m heterogeneous executors, and copy and distribute the received user request to each target heterogeneous executor among the n target heterogeneous executors.
[0034] The heterogeneous execution body 202 is used to complete parameter verification, calculation, and execution of the received user request, and then output the execution result to the business side agent module 203.
[0035] The business-side proxy module 203 is used to adjudicate the execution results received from the heterogeneous executor 202, and send the adjudication results to the business execution module (the business execution module in this application is the nova-compute component of the Openstack cloud platform), so that the business execution module performs corresponding operations according to the adjudication results.
[0036] The feedback control unit 204 is used to rotate and clean abnormal heterogeneous execution bodies, so as to timely discover and process abnormal heterogeneous execution bodies among multiple heterogeneous execution bodies, thereby ensuring the normal operation of the entire architecture.
[0037] Next, we will introduce S110. Figure 3 As shown, Figure 3 The user-side proxy unit in Figure 2 In the user-side proxy unit 201, the user can send a mount request to the UI to mount the cloud hard disk. The UI generates a first Restful request according to the mount request and sends the first Restful request to the user-side proxy unit.
[0038] The first Restful request carries a first identifier of a target cloud host to be mounted by the user and a second identifier of a target cloud hard disk to be mounted.
[0039] The first identifier of the target cloud host is, for example, a universally unique identifier (UUID) of the target cloud host.
[0040] The second identifier of the target cloud hard disk is, for example, the UUID of the target cloud hard disk.
[0041] S120: The user-side proxy unit copies and distributes the first Restful request to n target heterogeneous executors among the m heterogeneous executors; m and n are positive integers, m≥n≥3.
[0042] The user-side proxy unit selects n target heterogeneous executors from the m heterogeneous executors, and copies and distributes the first Restful request to each target heterogeneous executor in the n target heterogeneous executors.
[0043] When the user-side proxy unit selects n target heterogeneous executables from m heterogeneous executables, it may first randomly select n target heterogeneous executables from the heterogeneous executables in a normal state according to relevant parameters in each heterogeneous executable.
[0044] In the embodiment of the present invention, see Figure 2 , the heterogeneous executor can be a nova-api component (corresponding to Figure 2 in nova-api).
[0045] For example, Figure 3 As shown, n is 3, and the user-side proxy unit randomly selects 3 nova-api components in normal state from multiple nova-api components (corresponding to Figure 3In nova-api1, nova-api2, and nova-api3, the first Restful request is copied and distributed to the selected three nova-api components.
[0046] S130: The target heterogeneous executor creates BDM information in response to the received first Restful request when it is determined that the target cloud host and the target cloud hard disk are both in an available state.
[0047] BDM is the abbreviation of block_device_mapping, which means block device mapping.
[0048] After receiving the first Restful request, the target heterogeneous executor can query the status of the target cloud host corresponding to the first identifier in the database through the business side agent unit, and query the status of the target cloud hard disk corresponding to the second identifier in the database through the business side agent unit.
[0049] When the target cloud host is in the active state or shutdown state and there are no other tasks being executed, determine that the target cloud host is in the available state.
[0050] When the status of the target cloud hard disk is in the available state, it is determined that the status of the target cloud hard disk is in the available state.
[0051] When it is determined that the target cloud host and the target cloud hard disk are both in an available state, the target heterogeneous executor creates BDM information, sends an RPC Call synchronization message carrying the BDM information, the first identifier and the second identifier to the service side agent unit, and then executes S140.
[0052] Before determining that the target cloud host and the target cloud hard disk are both in an available state, the target heterogeneous executor can perform parameter processing on the parameters in the first Restful request. Parameter processing includes parameter verification and parameter calculation. Parameter verification includes verifying the legality and integrity of the parameters, and parameter calculation includes calculating one or more parameters to obtain new parameters.
[0053] For example, see Figure 3 After receiving the first Restful request, the target heterogeneous executor first processes the parameters. After that, when it is determined that the target cloud host and the target cloud hard disk are both available, it creates the BDM information ( Figure 3 (not shown) the RPC Call synchronization message carrying the BDM information, the first identifier and the second identifier is sent to the service side agent unit.
[0054] It should be noted that the message sent to the service side proxy unit here is a synchronous message, that is, after receiving the return message returned by the service side proxy unit, the target heterogeneous executor will not process other received messages.
[0055] In some embodiments, when it is determined that the target cloud host or the target cloud hard disk is not in an available state, the target heterogeneous executor returns an operation conflict message to the user-side agent unit; after receiving the operation conflict message, the user-side agent unit can adjudicate the operation conflict messages returned by all target heterogeneous executors to determine whether the operation conflict messages are consistent. When it is determined that the operation conflict messages are consistent, an operation conflict response message corresponding to the operation conflict message is sent to the UI for display to prompt the user that the current operation has a conflict and cannot continue to be executed; when the operation conflict message is inconsistent, a target heterogeneous executor in an abnormal state among the n target heterogeneous executors is obtained, and information of the target heterogeneous executor is sent to the feedback control unit so that the feedback control unit rotates and cleans the target heterogeneous executor.
[0056] S140: The service-side proxy unit receives an RPC Call synchronization message carrying BDM information, a first identifier, and a second identifier from a target heterogeneous executor, and determines whether the BDM information in the RPC Call synchronization messages of all target heterogeneous executors is consistent.
[0057] The business side agent unit determines whether the BDM information in the RPC Call synchronization messages of all target heterogeneous executors is consistent. When the first determination result is yes, that is, when the BDM information in the RPC Call synchronization messages of all target heterogeneous executors is consistent, a first message carrying the BDM information, the first identifier and the second identifier is sent to the rabbitmq message queue, and then S150 is executed.
[0058] For example, Figure 3 As shown, when the BDM information in the RPC Call synchronization messages of all target heterogeneous executors is consistent, the service-side proxy unit sends the first message to the rabbitmq message queue ( Figure 3 The corresponding one is rabbitmq).
[0059] When the first judgment result is no, that is, when the BDM information in the RPC Call synchronization message of all target heterogeneous executors is inconsistent, the business side agent unit obtains the identification information of the first abnormal heterogeneous executor in an abnormal state among the n target heterogeneous executors, and sends the identification information of the first abnormal heterogeneous executor to the feedback control unit, and the feedback control unit rotates and cleans the first abnormal heterogeneous executor.
[0060] S150: the nova-compute component monitors the rabbitmq message queue to obtain a first message carrying the BDM information, the first identifier, and the second identifier, which is sent by the service-side agent unit to the rabbitmq message queue when the first determination result is yes.
[0061] For example, Figure 3 As shown, the nova-compute component ( Figure 3 The corresponding one is nova-compute) can monitor the rabbitmq message queue, and when the rabbitmq message queue receives the first message, obtain the first message from the rabbitmq message queue.
[0062] S160: The nova-compute component writes the BDM table obtained according to the BDM information, the first identifier, and the second identifier into the database.
[0063] The nova-compute component may first generate an initial BDM data structure table according to the BDM information, and fill the first identifier and the second identifier into the initial BDM data structure table to obtain a BDM table.
[0064] Before writing the BDM table to the database, the nova-compute component can first write the operation record of mounting the target cloud disk to the database, making it easier for administrators to view operation logs and other information.
[0065] The BDM information in the first message obtained by the nova-compute component is consistent in n target heterogeneous executors. In this case, the nova-compute component will write the BDM table obtained based on the BDM information, the first identifier and the second identifier into the database. There will be no problems such as confusion, duplication or conflict in the written data, effectively ensuring the consistency of the data written to the database.
[0066] S170: When the BDM table is successfully written into the database, the nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor, and performs an operation of mounting the target cloud hard disk in the target cloud host.
[0067] When the BDM table is successfully written into the database, the nova-compute component may return a third identifier of the BDM table (eg, a UUID of the BDM table) to the target heterogeneous executor, so that the target heterogeneous executor returns a mount response to the user agent unit according to the received third identifier.
[0068] On the user-side proxy unit side, the user-side proxy unit receives the mount response message returned by the target heterogeneous executor according to the third identifier received by the target heterogeneous executor, and then determines whether the mount response messages returned by all target heterogeneous executors are consistent to obtain a second determination result.
[0069] When the second decision result is yes, the user-side agent unit sends the second decision result to the UI for display.
[0070] When the second determination result is no, the user-side proxy unit obtains identification information of a second abnormal heterogeneous executor in an abnormal state among all current target heterogeneous executors, and sends the identification information of the second abnormal heterogeneous executor to the feedback control unit.
[0071] The feedback control unit rotates and cleans the second abnormal heterogeneous executor.
[0072] For example, see Figure 3 , the nova-compute component returns the third identifier of the BDM table to each nova-api component. After receiving the third identifier, each nova-api component returns a mount response message (corresponding to Figure 3 After receiving the mount responses of all target heterogeneous executable bodies, the user-side proxy unit determines whether the mount response messages of all target heterogeneous executable bodies are consistent, and obtains a second determination result. The second determination result may include: the mount response messages of all target heterogeneous executable bodies are consistent, or the mount response messages of all target heterogeneous executable bodies are inconsistent. The user-side proxy unit may send the above second determination result to the UI for display, so that the user can understand the current cloud hard disk mount processing status.
[0073] In some embodiments, the two events of returning the third identifier of the BDM table to the target heterogeneous executive and performing the operation of mounting the target cloud hard disk in the target cloud host can be executed simultaneously on the same thread.
[0074] In some embodiments, in order to improve processing efficiency, after the nova-compute component writes the BDM table obtained according to the BDM information, the first identifier and the second identifier into the database, if the BDM table is successfully written into the database, the nova-compute component can create a secondary thread. In this case, the nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor through the current main thread, and executes the operation of mounting the target cloud hard disk in the target cloud host through the secondary thread.
[0075] For example, see Figure 3, the nova-compute component returns the UUID of the BDM table to the nova-api component, and creates a new thread (i.e., the secondary thread) to perform the operation of mounting the target cloud hard disk in the target cloud host.
[0076] In some embodiments, see Figure 3 , the nova-compute component can send a first state change request to the cinder-api component to change the state of the target cloud hard disk through the created secondary thread; the cinder-api component responds to the received first state change request and sets the state of the target cloud hard disk to the reserved state; the target cloud hard disk in the reserved state is prohibited from being mounted by other cloud hosts except the target cloud host, so as to ensure that the target cloud hard disk can be uniquely mounted on the target cloud host.
[0077] In some embodiments, the nova-compute component may acquire the task lock of the target cloud host through the secondary thread, and release the task lock when the operation of mounting the target cloud hard disk in the target cloud host executed through the secondary thread is completed.
[0078] Based on the setting of the above task lock, when the target cloud hard disk is mounted in the target cloud host, if the target cloud host also needs to mount other cloud hard disks, the task lock of the target cloud host will be occupied, resulting in a block in the secondary thread waiting for the release of the task lock, thereby achieving the purpose of concurrently mounting cloud hard disks on the same target cloud host on the user side.
[0079] In some embodiments, after the nova-compute component obtains the task lock of the target cloud host through the secondary thread, the nova-compute component can send a mount information request to the cinder-api component, and the nova-compute component updates the BDM table in the database according to the mount information returned by the cinder-api component in response to the mount information request; wherein the mount information includes: the drive letter information of the target cloud hard disk in the target cloud host.
[0080] For example, Figure 3 As shown in the figure, after obtaining the task lock of the target cloud host, the nova-compute component sends a request to the cinder-api component to update the target cloud disk status and obtain the backend mounting information, update the mounting information in the BDM, and complete operations such as mounting the cloud disk by calling the libvirt driver.
[0081] It should be noted that Figure 3For example, three heterogeneous executors obtain the message processing response forwarded by the business-side proxy module through the message synchronization mechanism, obtain the UUID in the BDM table, and query the database based on the UUID to obtain the BDM table. Since the three heterogeneous executors obtain the same UUID, when not attacked, the content of the BDM table queried by each heterogeneous executor is the same, ensuring data consistency when the three heterogeneous executors perform the same operation.
[0082] The cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture provided by the present invention can determine whether the execution results of multiple target heterogeneous executors are consistent in the business side agent unit. When the determination result is that the execution results of multiple target heterogeneous executors are consistent, the cloud hard disk mounting operation is performed, thereby avoiding the situation where multiple heterogeneous executors successively write the same data or write different data to the database, thereby causing confusion, duplication or even conflict in the data written in the database. The defect of confusion, duplication or conflict in the data written in the database when the cloud hard disk mounting operation is performed in the Openstack cloud platform introduced with the intrinsic security architecture in the prior art is solved, and the purpose of ensuring the consistency of the data written in the database when the cloud hard disk is mounted in the Openstack cloud platform introduced with the intrinsic security architecture is achieved.
[0083] The present invention also provides a cloud hard disk uninstallation method for an Openstack cloud platform with an inherent security architecture. The introduction of the inherent security architecture can refer to the corresponding description of each unit in the content security architecture 200 in the above embodiment, which will not be repeated here.
[0084] like Figure 4 As shown, the cloud hard disk uninstallation method of the Openstack cloud platform with an intrinsic security architecture includes the following S410~S460.
[0085] S410: The user-side proxy unit receives a second RESTful request for uninstalling the cloud hard disk from the UI; the second RESTful request carries the first identifier of the target cloud host and the second identifier of the target cloud hard disk.
[0086] The first identifier of the target cloud host is, for example, the UUID of the target cloud host, and the second identifier of the target cloud hard disk is, for example, the UUID of the target cloud hard disk.
[0087] S420: The user-side proxy unit copies and distributes the second Restful request to k target heterogeneous executors among the m heterogeneous executors; m and k are positive integers, m≥k≥3.
[0088] When the user-side proxy unit selects k target heterogeneous executors from m heterogeneous executors, it may first determine the heterogeneous executors in a normal state according to relevant parameters in each heterogeneous executor, and randomly select k target heterogeneous executors from all heterogeneous executors in a normal state.
[0089] For example, Figure 5 As shown, k is 3, and the user-side proxy unit randomly selects 3 nova-api components in normal state from multiple nova-api components (corresponding to Figure 5 In nova-api), the second Restful request is copied and distributed to the three selected nova-api components.
[0090] S430: In response to the received second Restful request, the target heterogeneous executor sends an RPC Call asynchronous message carrying the first identifier and the second identifier to the service-side proxy unit when determining that the target cloud host is in an available state.
[0091] After receiving the second Restful request, the target heterogeneous executor first checks whether the target cloud host meets the uninstallation requirement, such as that the target cloud host is in an available state.
[0092] The target cloud host is in an available state, including active state or shutdown state, and has no other tasks being executed.
[0093] S440: The service-side proxy unit determines whether the RPC Call asynchronous messages of all target heterogeneous executors are consistent.
[0094] The service-side proxy unit determines whether the RPC Call asynchronous messages of all target heterogeneous executors are consistent, and obtains a third determination result.
[0095] When the third determination result is yes, that is, when the RPC Call asynchronous messages of all target heterogeneous execution bodies are consistent, S450 is executed.
[0096] When the third judgment result is no, that is, when the RPC Call asynchronous messages of all target heterogeneous executors are inconsistent, the identification information of the third abnormal heterogeneous executor in the abnormal state among the k target heterogeneous executors is obtained, and the identification information of the third abnormal heterogeneous executor is sent to the feedback control unit, and the feedback control unit rotates and cleans the third abnormal heterogeneous executor.
[0097] S450: the nova-compute component monitors the rabbitmq message queue to obtain the second message carrying the first identifier and the second identifier that is sent to the rabbitmq message queue by the service-side agent unit when the third determination result is yes.
[0098] For example, Figure 5 As shown, the nova-compute component ( Figure 5 The corresponding one is nova-compute) can monitor the rabbitmq message queue, and when the rabbitmq message queue receives the second message, obtain the second message from the rabbitmq message queue.
[0099] S460: The nova-compute component performs an operation of uninstalling the target cloud hard disk in the target cloud host according to the first identifier and the second identifier.
[0100] For example, Figure 5 As shown in the figure, the nova-compute component can first obtain the task lock of the target cloud host and create an operation record of uninstalling the cloud hard disk in the database. The operation record includes identification information such as the first identifier of the target cloud host and the second identifier of the target cloud hard disk. Creating an operation record can facilitate task rollback and administrators to view information such as operation logs.
[0101] While creating the operation record of uninstalling the cloud hard disk, the nova-compute component can send a second state change request to the cinder-api component to request the cinder-api component to change the state of the target cloud hard disk. In response to the received second state change request, the cinder-api component sets the state of the target cloud hard disk to uninstalling.
[0102] The nova-compute component queries the status of the target cloud host, obtains the BDM table containing the first identifier and the second identifier in the database, obtains the mounting information (such as the drive letter information) of the target cloud hard disk from the BDM table, and then the libvirt driver uninstalls the target cloud hard disk from the target cloud host.
[0103] Finally, the nova-compute component deletes the BDM table containing the first identifier and the second identifier in the database, and calls the cinder-api component again to change the status of the target cloud hard disk to an available status, so that other cloud hosts can mount, delete, and perform other operations on the target cloud hard disk.
[0104] Finally, release the task lock to complete the process of uninstalling the target cloud hard disk.
[0105] Please continue to see Figure 5As shown, the nova-api component sends the RPCCall asynchronous message to the business-side agent unit to uninstall the cloud hard disk, and also returns an uninstall response message to the user-side agent unit. The user-side agent unit determines whether the uninstall response messages of all received target heterogeneous executable bodies are consistent to obtain a fourth determination result.
[0106] When the fourth determination result is yes, that is, when the uninstall response messages of all target heterogeneous executables are consistent, the fourth determination result is sent to the UI for display.
[0107] When the fourth determination result is no, that is, when the uninstall response messages of all target heterogeneous executors are inconsistent, the identification information of the fourth abnormal heterogeneous executor in the target heterogeneous executors that is in an abnormal state is obtained, and the identification information of the fourth abnormal heterogeneous executor is sent to the feedback control unit, and the feedback control unit rotates and cleans the fourth abnormal heterogeneous executor.
[0108] The cloud hard disk unloading method of the Openstack cloud platform based on the intrinsic security architecture provided by the present invention can determine whether the execution results of multiple target heterogeneous executors are consistent in the business side agent unit. When the determination result is that the execution results of multiple target heterogeneous executors are consistent, the cloud hard disk unloading operation is performed, so as to avoid the situation where multiple heterogeneous executors successively write the same data or write different data to the database, thereby causing confusion, duplication or even conflict in the data written in the database. The defect of confusion, duplication or conflict in the data written in the database when the cloud hard disk unloading operation is performed in the Openstack cloud platform with the intrinsic security architecture in the prior art is solved, and the purpose of ensuring the consistency of the data written in the database when the cloud hard disk is unloaded in the Openstack cloud platform with the intrinsic security architecture is achieved.
[0109] The embodiments of the present invention can achieve the following beneficial effects: 1. Compared with other intrinsic security architectures, by adding nova-api to the database to mount and unmount cloud hard disks, database operation records and block_device_mapping table creation must be judged by the business-side agent, which seriously affects the efficiency of cloud hard disk mounting and unmounting, and also increases the instability of system operation. By verifying and calculating the request parameters in the heterogeneous executor, and saving the original input written to the database to the memory first, and sending it as the request message content to the business-side agent for judgment, after one judgment by the business-side agent unit, the original heterogeneous executor can complete the proxy judgment of multiple database additions, deletions and modifications. The addition, deletion and modification of the database are completed through the business layer unit nova-compute, which greatly reduces the complexity of the cloud hard disk mounting process and improves the cloud hard disk mounting efficiency.
[0110] 2. Compared with the native OpenStack cloud platform (the OpenStack cloud platform that has not introduced an inherent security architecture), by converting multiple nova-APIs into multiple heterogeneous executors to simultaneously perform the tasks of mounting and unmounting cloud hard disks, the security of the cloud platform in mounting and unmounting cloud hard disks is enhanced through the dynamic, heterogeneous, and redundant characteristics.
[0111] 3. Since each business-side proxy unit decision consumes a lot of time and resources, two RPC Call message calls are changed to one RPC Call call. Therefore, in the entire cloud hard disk mounting process, only one RPC Call message synchronization call is required to complete the entire function, thereby simplifying the cloud hard disk mounting process and improving the efficiency of the cloud hard disk mounting process.
[0112] 4. Implement concurrent cloud hard disk mounting operations. In order to implement concurrent operations on multiple cloud hard disks for the same cloud host, a coroutine mechanism (i.e., creating a secondary thread) is added to the cloud hard disk mounting process in the nova-compute service to ensure that after writing block_device_mapping (BDM table) and operation logs and other data to the database, it is immediately returned to the nova-api service (the executor of the nova-api service, the target heterogeneous executor in the present invention); at the same time, a task lock mechanism is added to solve the conflict problem of mounting multiple cloud hard disks on the same cloud host at the same time.
[0113] Figure 6 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 6As shown, the electronic device may include: a processor (processor) 610 , a communication interface (Communications Interface) 620 , a memory (memory) 630 and a communication bus 640 , wherein the processor 610 , the communication interface 620 , and the memory 630 communicate with each other through the communication bus 640 . The processor 610 can call the logic instructions in the memory 630 to execute the cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture, and the method includes: the user-side agent unit receives the first Restful request for mounting the cloud hard disk from the UI; the first Restful request carries the first identifier of the target cloud host and the second identifier of the target cloud hard disk; the user-side agent unit copies and distributes the first Restful request to n target heterogeneous executors among the m heterogeneous executors; m and n are positive integers, m≥n≥3; the target heterogeneous executor responds to the received first Restful request, and creates BDM information when it is determined that the target cloud host and the target cloud hard disk are both in an available state; the business-side agent unit receives the RPC Call synchronization message carrying the BDM information, the first identifier and the second identifier from the target heterogeneous executor, and performs RPC Call synchronization on all target heterogeneous executors. The nova-compute component determines whether the BDM information in the Call synchronization message is consistent; the nova-compute component monitors the rabbitmq message queue to obtain the first message carrying the BDM information, the first identifier and the second identifier sent to the rabbitmq message queue by the business-side agent unit when the first determination result is yes; the nova-compute component writes the BDM table obtained according to the BDM information, the first identifier and the second identifier into the database; when the BDM table is successfully written to the database, the nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor, and executes the operation of mounting the target cloud hard disk in the target cloud host.
[0114] Alternatively, the processor 610 may call the logic instructions in the memory 630 to execute a cloud hard disk unloading method for the Openstack cloud platform based on an intrinsic security architecture, the method comprising: a user-side proxy unit receives a second Restful request for unloading a cloud hard disk from a UI; the second Restful request carries a first identifier of a target cloud host and a second identifier of a target cloud hard disk; the user-side proxy unit copies and distributes the second Restful request to k target heterogeneous executors among m heterogeneous executors; m and k are positive integers, m≥k≥3; the target heterogeneous executor responds to the received second Restful request, and when determining that the target cloud host is in an available state, sends an RPCCall asynchronous message carrying a first identifier and a second identifier to the business-side proxy unit; the business-side proxy unit determines whether the RPC Call asynchronous messages of all target heterogeneous executors are consistent; the nova-compute component obtains a second message carrying a first identifier and a second identifier sent to the rabbitmq message queue by the business-side proxy unit when the third determination result is yes by monitoring the rabbitmq message queue; the nova-compute component performs an operation of unloading the target cloud hard disk in the target cloud host according to the first identifier and the second identifier.
[0115] In addition, the logic instructions in the above-mentioned memory 630 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.
[0116] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture provided by the above-mentioned methods, and the method includes: a user-side agent unit receives a first Restful request for mounting a cloud hard disk from a UI; the first Restful request carries a first identifier of a target cloud host and a second identifier of a target cloud hard disk; the user-side agent unit copies and distributes the first Restful request to n target heterogeneous executors among m heterogeneous executors; m and n are positive integers, m≥n≥3; the target heterogeneous executor responds to the received first Restful request, and creates BDM information when it is determined that the target cloud host and the target cloud hard disk are both in an available state; the business-side agent unit receives an RPC Call synchronization message carrying BDM information, a first identifier and a second identifier from the target heterogeneous executor, and performs RPC on all target heterogeneous executors. The nova-compute component determines whether the BDM information in the Call synchronization message is consistent; the nova-compute component monitors the rabbitmq message queue to obtain the first message carrying the BDM information, the first identifier and the second identifier sent to the rabbitmq message queue by the business-side agent unit when the first determination result is yes; the nova-compute component writes the BDM table obtained according to the BDM information, the first identifier and the second identifier into the database; when the BDM table is successfully written to the database, the nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor, and executes the operation of mounting the target cloud hard disk in the target cloud host.
[0117] Alternatively, when the computer program is executed by the processor, the computer can execute the cloud hard disk uninstallation method of the Openstack cloud platform based on the intrinsic security architecture provided by the above methods, the method comprising: a user-side agent unit receives a second Restful request for uninstalling the cloud hard disk from the UI; the second Restful request carries the first identifier of the target cloud host and the second identifier of the target cloud hard disk; the user-side agent unit copies and distributes the second Restful request to k target heterogeneous executors among m heterogeneous executors; m and k are positive integers, m≥k≥3; the target heterogeneous executor responds to the received second Restful request, and when determining that the target cloud host is in an available state, sends an RPC Call asynchronous message carrying the first identifier and the second identifier to the business-side agent unit; the business-side agent unit determines whether the RPC Call asynchronous messages of all target heterogeneous executors are consistent; the nova-compute component obtains the second message carrying the first identifier and the second identifier sent to the rabbitmq message queue by the business-side agent unit when the third determination result is yes by monitoring the rabbitmq message queue; the nova-compute component executes the operation of uninstalling the target cloud hard disk in the target cloud host according to the first identifier and the second identifier.
[0118] On the other hand, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture provided by the above-mentioned methods, the method comprising: a user-side agent unit receives a first Restful request for mounting a cloud hard disk from a UI; the first Restful request carries a first identifier of a target cloud host and a second identifier of a target cloud hard disk; the user-side agent unit copies and distributes the first Restful request to n target heterogeneous executors among m heterogeneous executors; m and n are positive integers, m≥n≥3; the target heterogeneous executor responds to the received first Restful request, and creates BDM information when it is determined that the target cloud host and the target cloud hard disk are both in an available state; the business-side agent unit receives an RPC Call synchronization message carrying the BDM information, the first identifier and the second identifier from the target heterogeneous executor, and performs RPC on all target heterogeneous executors. The nova-compute component determines whether the BDM information in the Call synchronization message is consistent; the nova-compute component monitors the rabbitmq message queue to obtain the first message carrying the BDM information, the first identifier and the second identifier sent to the rabbitmq message queue by the business-side agent unit when the first determination result is yes; the nova-compute component writes the BDM table obtained according to the BDM information, the first identifier and the second identifier into the database; when the BDM table is successfully written to the database, the nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor, and executes the operation of mounting the target cloud hard disk in the target cloud host.
[0119] Alternatively, when the above-mentioned computer program is executed by the processor, it is implemented to execute the cloud hard disk unloading method of the Openstack cloud platform based on the intrinsic security architecture provided by the above-mentioned methods, and the method includes: the user-side agent unit receives a second Restful request for unloading the cloud hard disk from the UI; the second Restful request carries the first identifier of the target cloud host and the second identifier of the target cloud hard disk; the user-side agent unit copies and distributes the second Restful request to k target heterogeneous executors among the m heterogeneous executors; m and k are positive integers, m≥k≥3; the target heterogeneous executor responds to the received second Restful request, and when it is determined that the target cloud host is in an available state, sends an RPC Call asynchronous message carrying the first identifier and the second identifier to the business-side agent unit; the business-side agent unit determines whether the RPC Call asynchronous messages of all target heterogeneous executors are consistent; the nova-compute component obtains the second message carrying the first identifier and the second identifier sent to the rabbitmq message queue by the business-side agent unit when the third determination result is yes by monitoring the rabbitmq message queue; the nova-compute component performs an operation of unloading the target cloud hard disk in the target cloud host according to the first identifier and the second identifier.
[0120] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A cloud hard disk mounting method for an Openstack cloud platform based on an intrinsic security architecture, characterized in that: The endogenous security architecture includes a user-side proxy unit, a service-side proxy unit, and m heterogeneous execution bodies. The method includes: The user-side proxy unit receives a first RESTful request for mounting a cloud hard disk from the UI; the first RESTful request carries a first identifier of a target cloud host and a second identifier of a target cloud hard disk; The user-side proxy unit copies and distributes the first Restful request to n target heterogeneous executors among the m heterogeneous executors; m and n are positive integers, m≥n≥3; The target heterogeneous executor creates BDM information in response to the received first Restful request when it is determined that the target cloud host and the target cloud hard disk are both in an available state; The service-side proxy unit receives an RPC Call synchronization message carrying the BDM information, the first identifier, and the second identifier from the target heterogeneous executor, and determines whether the BDM information in the RPC Call synchronization messages of all the target heterogeneous executors is consistent; The nova-compute component monitors the rabbitmq message queue to obtain the first message carrying the BDM information, the first identifier and the second identifier, which is sent by the service-side agent unit to the rabbitmq message queue when the first determination result is yes; The nova-compute component writes a BDM table obtained according to the BDM information, the first identifier and the second identifier into a database; When the BDM table is successfully written into the database, the nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor, and performs an operation of mounting the target cloud hard disk in the target cloud host.
2. The cloud hard disk mounting method of the Openstack cloud platform based on the inherent security architecture according to claim 1 is characterized in that: After the nova-compute component writes the BDM table obtained according to the BDM information, the first identifier, and the second identifier into a database, the method further includes: The nova-compute component creates a secondary thread when the BDM table is successfully written into the database; When the BDM table is successfully written into the database, the nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor, and performs an operation of mounting the target cloud hard disk in the target cloud host, including: The nova-compute component returns the third identifier of the BDM table to the target heterogeneous executor through the current main thread, and executes the operation of mounting the target cloud hard disk in the target cloud host through the secondary thread.
3. The cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture according to claim 2 is characterized in that: After creating the secondary thread, the following steps are also included: The nova-compute component sends a first state change request to change the state of the target cloud hard disk to the cinder-api component through the secondary thread; The cinder-api component sets the state of the target cloud hard disk to a reserved state in response to the received first state change request; the target cloud hard disk in the reserved state is prohibited from being mounted by other cloud hosts other than the target cloud host.
4. The cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture according to claim 2 is characterized in that: After creating the secondary thread, the following steps are also included: The nova-compute component obtains the task lock of the target cloud host through the secondary thread; and releases the task lock when the operation of mounting the target cloud hard disk in the target cloud host executed by the secondary thread is completed.
5. The cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture according to claim 4 is characterized in that: After the nova-compute component obtains the task lock of the target cloud host through the secondary thread, the method further includes: The nova-compute component sends a mount information request to the cinder-api component; The nova-compute component updates the BDM table in the database according to the mount information returned by the cinder-api component in response to the mount information request; wherein the mount information includes: the drive letter information of the target cloud hard disk in the target cloud host.
6. The cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture according to claim 1 is characterized in that: After the user-side proxy unit copies and distributes the first Restful request to n target heterogeneous executors among the m heterogeneous executors, the method further includes: In response to the received first Restful request, the target heterogeneous executor returns an operation conflict message to the user-side proxy unit if it is determined that the target cloud host or the target cloud hard disk is not in the available state; The user-side agent unit sends an operation conflict response message corresponding to the operation conflict message to the UI for display.
7. The cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture according to claim 1 is characterized in that: The intrinsic security architecture further includes: a feedback control unit; after determining whether the BDM information in the RPC Call synchronization message of all the target heterogeneous execution bodies is consistent, it also includes: When the first ruling result is negative, the service-side agent unit obtains identification information of a first abnormal heterogeneous executor in an abnormal state among the n target heterogeneous executors, and sends the identification information of the first abnormal heterogeneous executor to the feedback control unit; The feedback control unit rotates and cleans the first abnormal heterogeneous executor.
8. The cloud hard disk mounting method of the Openstack cloud platform based on the intrinsic security architecture according to claim 7 is characterized in that: After returning the third identifier of the BDM table to the target heterogeneous executable body, the method further includes: The user-side proxy unit receives a mount response message returned by the target heterogeneous executable according to the third identifier received by the target heterogeneous executable; The user-side proxy unit determines whether the mount response messages returned by all the target heterogeneous execution bodies are consistent; and obtains a second determination result; If the second decision result is yes, sending the second decision result to the UI for display; If the second determination result is no, obtaining identification information of a second abnormal heterogeneous executor in an abnormal state among all current target heterogeneous executors, and sending the identification information of the second abnormal heterogeneous executor to the feedback control unit; The feedback control unit rotates and cleans the second abnormal heterogeneous executor.
9. A cloud hard disk uninstallation method for an Openstack cloud platform based on an intrinsic security architecture, characterized in that: The endogenous security architecture includes a user-side proxy unit, a service-side proxy unit, and m heterogeneous execution bodies. The method includes: The user-side proxy unit receives a second RESTful request for uninstalling the cloud hard disk from the UI; the second RESTful request carries a first identifier of the target cloud host and a second identifier of the target cloud hard disk; The user-side proxy unit copies and distributes the second Restful request to k target heterogeneous executors among the m heterogeneous executors; m and k are positive integers, m≥k≥3; In response to the received second Restful request, the target heterogeneous executor sends an RPC Call asynchronous message carrying the first identifier and the second identifier to the service-side proxy unit when determining that the target cloud host is in an available state; The service-side proxy unit determines whether the RPC Call asynchronous messages of all the target heterogeneous executors are consistent; The nova-compute component monitors the rabbitmq message queue to obtain the second message carrying the first identifier and the second identifier sent by the service-side agent unit to the rabbitmq message queue when the third determination result is yes; The nova-compute component performs an operation of uninstalling the target cloud hard disk in the target cloud host according to the first identifier and the second identifier.
10. The cloud hard disk uninstallation method of the Openstack cloud platform based on the intrinsic security architecture according to claim 9 is characterized in that: The nova-compute component performs an operation of uninstalling the target cloud hard disk in the target cloud host according to the first identifier and the second identifier, including: The nova-compute component obtains the task lock of the target cloud host according to the first identifier; The nova-compute component obtains a BDM table including the first identifier and the second identifier from a database according to the first identifier and the second identifier; The nova-compute component uninstalls the target cloud hard disk from the target cloud host according to the mounting information of the target cloud hard disk in the BDM table; The nova-compute component deletes the BDM table in the database; The nova-compute component sends a second state change request to the cinder-api component; The cinder-api component sets the state of the target cloud hard disk to be usable in response to the received second state change request.
11. The cloud hard disk uninstallation method of the Openstack cloud platform based on the intrinsic security architecture according to claim 9 is characterized in that: The inherent security architecture further includes: a feedback control unit; after the service-side agent unit determines whether the RPC Call asynchronous messages of all the target heterogeneous execution bodies are consistent, it also includes: When the third decision result is negative, the service-side agent unit obtains identification information of a third abnormal heterogeneous executor in an abnormal state among all the current target heterogeneous executors, and sends the identification information of the third abnormal heterogeneous executor to the feedback control unit; The feedback control unit rotates and cleans the third abnormal heterogeneous executor.
12. The cloud hard disk uninstallation method of the Openstack cloud platform based on the intrinsic security architecture according to claim 11 is characterized in that: The target heterogeneous executor responds to the received second Restful request and, if it is determined that the target cloud host is in an available state, sends an RPC Call asynchronous message carrying the first identifier and the second identifier to the service-side proxy unit, further comprising: The target heterogeneous executive returns an uninstallation response message to the user-side proxy unit; After receiving the uninstallation response message, the user-side proxy unit performs consistency determination on the uninstallation response messages returned by all the target heterogeneous execution bodies to obtain a fourth determination result; If the fourth decision result is yes, sending the fourth decision result to the UI for display; When the fourth determination result is no, obtaining identification information of a fourth abnormal heterogeneous executor in an abnormal state among all the current target heterogeneous executors, and sending the identification information of the fourth abnormal heterogeneous executor to the feedback control unit; The feedback control unit rotates and cleans the fourth abnormal heterogeneous execution body.
Citation Information
Patent Citations
Method for mounting and unloading cloud hard disk based on OpenStack cloud platform
CN112463248A
Program deadlock detection method and device, computer equipment and storage medium
CN112631790A
Task management scheduling method for distributed multi-cloud management platform
CN115617493A
Endogenous security cloud platform and construction method
CN116471116A
