Anomaly detection apparatus and anomaly detection method

Abstract

The present application discloses an anomaly detection apparatus and an anomaly detection method, capable of implementing automation of anomaly detection of optical modules of physical servers. The apparatus in the present application may be arranged on a rack of a cloud data center, a hollow portion of the rack can accommodate a plurality of physical servers, and optical modules of the plurality of physical servers may face an opening of the hollow portion of the rack. The anomaly detection apparatus may comprise a moving component and a detector. The detector, with the support of the moving component, can freely move at the opening of the hollow portion of the rack, so as to face the optical modules of the plurality of physical servers in the rack. During movement, the detector can emit ultrasonic waves to the optical modules of the plurality of physical servers, the ultrasonic waves are reflected at the optical modules of the plurality of physical servers, and the reflected ultrasonic waves are received by the detector. On this basis, the detector can determine, on the basis of the reflected ultrasonic waves, whether the optical modules of the plurality of physical servers are abnormal.

Description

An anomaly detection device and anomaly detection method

[0001] This application claims priority to Chinese Patent Application No. 202411999281.6, filed with the State Intellectual Property Office of China on December 31, 2024, entitled "An Anomaly Detection Device and Anomaly Detection Method", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of cloud technology, and in particular to an anomaly detection device and an anomaly detection method. Background Technology

[0003] With the rapid development of cloud technology, more and more tenants are choosing cloud data centers provided by cloud vendors to complete their business. Cloud data centers typically contain a large number of physical servers that work together to efficiently complete tenant business tasks and meet their needs.

[0004] In related technologies, to enable rapid communication between multiple physical servers in a cloud data center, these physical servers are typically equipped with optical modules. Generally, the optical module of one physical server can convert the electrical signals to be transmitted within that physical server into optical signals, and transmit the optical signals to the optical module of another physical server. The optical module of the other physical server then converts the optical signals back into electrical signals, and the electrical signals are processed in the other physical server, thereby completing the communication between the two physical servers.

[0005] Since physical servers are typically deployed in cloud data centers, while optical modules are often exposed outside the server racks, they are susceptible to environmental contamination and other factors, leading to malfunctions. To determine if an optical module is malfunctioning, cloud providers often require maintenance personnel to inspect the optical modules of each physical server individually, resulting in excessively high labor costs for anomaly detection. Summary of the Invention

[0006] This application provides an anomaly detection device and anomaly detection method, which can automate the anomaly detection of optical modules of physical servers, thereby reducing the manual cost required for anomaly detection and improving the efficiency and accuracy of anomaly detection.

[0007] The first aspect of this application provides an anomaly detection device, which can be installed in a rack of a cloud data center. The hollow part of the rack is equipped with multiple physical servers, each of which includes an optical module. The multiple physical servers can communicate with each other through the optical modules of the multiple physical servers, and the optical modules of the multiple physical servers face the opening of the hollow part of the rack.

[0008] The device may include a moving part and a detector, with both ends of the moving part connected to a rack and the detector, respectively. When the detector needs to perform anomaly detection on the optical modules of multiple physical servers located in the hollow section of the rack, it can control the moving part to move the detector at the opening of the hollow section of the rack, thereby causing the detector to face the optical modules of the multiple physical servers.

[0009] After passing through the optical modules of these multiple physical servers, the detector can send ultrasonic waves to these optical modules. The ultrasonic waves are reflected at the optical modules of these physical servers, and the reflected ultrasonic waves are received by the detector. Therefore, the detector can determine whether there is any abnormality in the optical modules of these multiple physical servers based on the reflected ultrasonic waves.

[0010] As can be seen from the above device, since the anomaly detection device installed on the rack includes moving parts and detectors, it can move freely at the opening of the hollow part of the rack, thereby automatically performing anomaly detection on the optical modules of multiple physical servers in the rack. This eliminates the need for cloud data center maintenance personnel to perform anomaly detection on the optical modules of multiple physical servers in the rack one by one. This automates the anomaly detection of the optical modules of physical servers, thereby reducing the manual cost required for anomaly detection and improving the efficiency and accuracy of anomaly detection.

[0011] In one possible implementation, the moving component includes a first slide rail, a second slide rail, a first retractable connecting arm, and a second retractable connecting arm. The first slide rail is fixed to a first side of the frame, and the second slide rail is fixed to a second side of the frame, wherein the first side and the second side are perpendicular. One end of the first retractable connecting arm is connected to the first slide rail, and the other end of the first retractable connecting arm is connected to a detector. One end of the second retractable connecting arm is connected to the second slide rail, and the other end of the second retractable connecting arm is connected to the detector. In the aforementioned implementation, the moving component may include a first slide rail, a second slide rail, a first retractable connecting arm, and a second retractable connecting arm. The first slide rail is fixed to the first side of the frame, and the second slide rail is fixed to the second side of the frame. Since the first side and the second side are perpendicular, the first and second slide rails are also arranged perpendicularly. One end of the first retractable connecting arm is connected to the first slide rail, and the other end of the first retractable connecting arm is connected to the detector. One end of the second retractable connecting arm is connected to the second slide rail, and the other end of the second retractable connecting arm is connected to the detector. Therefore, the detector is supported by the first and second retractable connecting arms, causing it to suspend above the opening in the hollow section of the rack. Since the first retractable connecting arm can slide freely on the first slide rail and is freely extendable, and the second retractable connecting arm can slide freely on the second slide rail and is also freely extendable, the detector can move freely at the opening in the hollow section of the rack through the cooperation of the first and second slide rails, the first and second retractable connecting arms. This provides a basis for subsequent anomaly detection of the optical modules of the multiple physical servers within the hollow section.

[0012] In one possible implementation, the detector is specifically configured to: receive an optical module detection request sent by a cloud management platform, the optical module detection request indicating the position of optical modules of multiple physical servers in a rack, wherein the cloud management platform manages the cloud data center; and move at an opening based on the optical module detection request to move towards the optical modules of the multiple physical servers and send ultrasonic waves to the optical modules of the multiple physical servers. In the aforementioned implementation, the cloud management platform can send an optical module detection request to the detector, which can be used to indicate the optical module detection task and the position of the optical modules of the multiple physical servers in the rack. After receiving the optical module detection request, the detector can parse the optical module detection request to obtain the position of the optical modules of the multiple physical servers in the rack. Based on the position of the optical modules of the multiple physical servers in the rack, the detector can move sequentially by controlling a moving component, so that the moving component drives the detector to move sequentially at the opening of the hollow part of the rack, thereby causing the detector to move towards the optical modules of the multiple physical servers sequentially and send ultrasonic waves to the optical modules of the multiple physical servers, so as to successfully achieve anomaly detection of the optical modules of the multiple physical servers.

[0013] In one possible implementation, the detector is further configured to: send an optical module anomaly notification to the cloud management platform if it is determined that the optical module of the first physical server among multiple physical servers is contaminated; receive an optical module processing request sent by the cloud management platform, the optical module processing request indicating the position of the optical module of the first physical server in the rack; and move at an opening based on the optical module processing request to face the optical module of the first physical server and remove the contaminant from the optical module of the first physical server. In the aforementioned implementation, if the first physical server among the multiple physical servers is contaminated, the detector can send an optical module anomaly notification to the cloud management platform, indicating that the optical module of the first physical server is contaminated. After receiving the optical module anomaly notification, the cloud management platform can determine that the optical module of the first physical server is contaminated, and therefore can send an optical module processing request to the detector, the optical module processing request indicating the position of the optical module of the first physical server in the rack. After receiving the optical module processing request, the detector can parse the optical module processing request to obtain the position of the optical module of the first physical server in the rack. Based on the position of the optical module of the first physical server within the rack, the detector can control a moving component to move, causing the component to move the detector towards the opening in the hollow section of the rack. This allows the detector to clean the optical module of the first physical server, removing any dirt or contaminants. In this way, the detector can not only detect dirt on the optical module of the first physical server but also automatically remove it, thus automating anomaly detection and handling.

[0014] In one possible implementation, the detector is specifically used to: send focused ultrasonic waves to the optical module of the first physical server, the focused ultrasonic waves being used to control the detachment of dirt from the optical module of the first physical server; and collect the detached dirt. In the aforementioned implementation, the detector can send focused ultrasonic waves to the optical module of the first physical server, the focused ultrasonic waves vibrating the dirt on the optical module of the first physical server so that the dirt detaches from the optical module of the first physical server, and collect the detached dirt in a timely manner, thereby successfully achieving the decontamination of the optical module of the first physical server.

[0015] In one possible implementation, the multiple physical servers also include hard drives, and a detector is further configured to: receive hard drive detection requests sent by a cloud management platform, the hard drive detection requests indicating the positions of the hard drives of the multiple physical servers in the rack; move at an opening based on the hard drive detection requests to face the hard drives of the multiple physical servers; collect sounds generated by the hard drives of the multiple physical servers, and determine whether there are any abnormalities in the hard drives of the multiple physical servers based on the sounds generated by the hard drives of the multiple physical servers. In the aforementioned implementation, the cloud management platform can also send hard drive detection requests to the detector, which can be used to indicate the positions of the hard drives of the multiple physical servers in the rack. After receiving the hard drive detection request, the detector can parse the hard drive detection request to obtain the positions of the hard drives of the multiple physical servers in the rack. Based on the positions of the hard drives of the multiple physical servers in the rack, the detector can successively control the moving parts to move, so that the moving parts drive the detector to move successively at the opening of the hollow part of the rack, so that the detector faces the hard drives of the multiple physical servers successively, and collects the sounds generated by the hard drives of the multiple physical servers successively, to determine whether there are any abnormalities in the hard drives of the multiple physical servers based on the sounds generated by the hard drives of the multiple physical servers. In this way, the detector can successfully detect anomalies in the hard drives of these multiple physical servers.

[0016] In one possible implementation, the detector is further configured to: upon determining that the hard drive of the second physical server among multiple physical servers is making abnormal noise, send a hard drive anomaly notification to the cloud management platform. This notification indicates that the abnormal noise is caused by contact between the hard drive's read / write head and the disk. In this implementation, if the second physical server among the multiple physical servers is making abnormal noise, the detector can send a hard drive anomaly notification to the cloud management platform. Upon receiving this notification, the cloud management platform can confirm the abnormal noise and then alert the maintenance personnel to replace the hard drive of the second physical server with a new one, thus successfully handling the hard drive anomaly of the second physical server.

[0017] A second aspect of this application provides an anomaly detection method, implemented by an anomaly detection device installed in a rack of a cloud data center. The hollow portion of the rack is used to house multiple physical servers, each containing an optical module. Communication between the multiple physical servers is achieved through these optical modules, which face an opening in the hollow portion. The device includes a moving component and a detector. One end of the moving component is connected to the rack, and the other end is connected to the detector. The method includes: the detector moving at the opening with the support of the moving component; during the movement, the detector sending ultrasonic waves to the optical modules of the multiple physical servers, receiving the reflected ultrasonic waves, and determining whether there is an anomaly in the optical modules of the multiple physical servers based on the reflected ultrasonic waves, wherein the reflected ultrasonic waves are obtained by reflection of ultrasonic waves at the optical modules of the multiple physical servers.

[0018] In one possible implementation, the moving component includes a first slide rail, a second slide rail, a first retractable connecting arm, and a second retractable connecting arm; the first slide rail is fixed to a first side of the frame, and the second slide rail is fixed to a second side of the frame, wherein the first side and the second side are perpendicular; one end of the first retractable connecting arm is connected to the first slide rail, and the other end of the first retractable connecting arm is connected to a detector; one end of the second retractable connecting arm is connected to the second slide rail, and the other end of the second retractable connecting arm is connected to the detector.

[0019] In one possible implementation, the detector moving at the opening includes: the detector receiving an optical module detection request sent by a cloud management platform, the optical module detection request indicating the location of optical modules of multiple physical servers in the rack, wherein the cloud management platform is used to manage the cloud data center; the detector moving at the opening based on the optical module detection request toward the optical modules of the multiple physical servers and sending ultrasonic waves to the optical modules of the multiple physical servers.

[0020] In one possible implementation, the method further includes: when the detector determines that the optical module of the first physical server among a plurality of physical servers is contaminated, sending an optical module anomaly notification to the cloud management platform, the optical module anomaly notification indicating that the optical module of the first physical server is contaminated; the detector receiving an optical module processing request sent by the cloud management platform, the optical module processing request indicating the position of the optical module of the first physical server in the rack; and the detector moving at the opening based on the optical module processing request to face the optical module of the first physical server and removing the contaminant from the optical module of the first physical server.

[0021] In one possible implementation, the detector removes dirt from the optical module of the first physical server by: the detector sending focused ultrasonic waves to the optical module of the first physical server, the focused ultrasonic waves being used to control the dirt to detach from the optical module of the first physical server; and the detector collecting the detached dirt.

[0022] In one possible implementation, the multiple physical servers also include hard drives, and the method further includes: a detector receiving a hard drive detection request sent by a cloud management platform, the hard drive detection request indicating the position of the hard drives of the multiple physical servers in the rack; the detector moving at an opening based on the hard drive detection request to face the hard drives of the multiple physical servers; the detector collecting sounds generated by the hard drives of the multiple physical servers, and determining whether there are any abnormalities in the hard drives of the multiple physical servers based on the sounds generated by the hard drives of the multiple physical servers.

[0023] In one possible implementation, the method further includes: when the detector determines that the hard drive of the second physical server among multiple physical servers has abnormal noise, sending a hard drive abnormality notification to the cloud management platform, the hard drive abnormality notification is used to indicate that the hard drive of the second physical server has abnormal noise, which is caused by the contact between the read / write head and the disk of the hard drive of the second physical server.

[0024] A third aspect of this application provides a cloud service system, which includes a cloud data center. The cloud data center includes a rack and an anomaly detection device. The hollow portion of the rack is used to house multiple physical servers, each physical server containing an optical module. The multiple physical servers communicate with each other through their optical modules, with the optical modules of the multiple physical servers facing the opening of the hollow portion. The device includes a moving component and a detector. One end of the moving component is connected to the rack, and the other end of the moving component is connected to the detector. The moving component supports the detector in moving at the opening. During the movement, the detector sends ultrasonic waves to the optical modules of the multiple physical servers, receives the reflected ultrasonic waves, and determines whether there is an anomaly in the optical modules of the multiple physical servers based on the reflected ultrasonic waves. The reflected ultrasonic waves are obtained by the reflection of ultrasonic waves at the optical modules of the multiple physical servers.

[0025] In one possible implementation, the moving component includes a first slide rail, a second slide rail, a first retractable connecting arm, and a second retractable connecting arm; the first slide rail is fixed to a first side of the frame, and the second slide rail is fixed to a second side of the frame, wherein the first side and the second side are perpendicular; one end of the first retractable connecting arm is connected to the first slide rail, and the other end of the first retractable connecting arm is connected to a detector; one end of the second retractable connecting arm is connected to the second slide rail, and the other end of the second retractable connecting arm is connected to the detector.

[0026] In one possible implementation, the detector is specifically used to: receive an optical module detection request sent by a cloud management platform, the optical module detection request indicating the location of optical modules of multiple physical servers in the rack, wherein the cloud management platform is used to manage the cloud data center; move at the opening based on the optical module detection request to face the optical modules of the multiple physical servers, and send ultrasonic waves to the optical modules of the multiple physical servers.

[0027] In one possible implementation, the detector is further configured to: send an optical module anomaly notification to a cloud management platform if it is determined that the optical module of the first physical server among a plurality of physical servers is contaminated, the optical module anomaly notification indicating that the optical module of the first physical server is contaminated; receive an optical module processing request sent by the cloud management platform, the optical module processing request indicating the position of the optical module of the first physical server in the rack; and move at an opening based on the optical module processing request toward the optical module of the first physical server and remove the contaminant from the optical module of the first physical server.

[0028] In one possible implementation, the detector is specifically used to: send focused ultrasonic waves to the optical module of the first physical server, the focused ultrasonic waves being used to control the detachment of dirt from the optical module of the first physical server; and collect the detached dirt.

[0029] In one possible implementation, the multiple physical servers also include hard drives, and a detector is also used to: receive hard drive detection requests sent by the cloud management platform, the hard drive detection requests indicating the position of the hard drives of the multiple physical servers in the rack; move at the opening based on the hard drive detection requests to face the hard drives of the multiple physical servers; collect the sounds generated by the hard drives of the multiple physical servers, and determine whether there are any abnormalities in the hard drives of the multiple physical servers based on the sounds generated by the hard drives of the multiple physical servers.

[0030] In one possible implementation, the detector is also used to: send a hard drive anomaly notification to the cloud management platform when it is determined that the hard drive of the second physical server among multiple physical servers has an abnormal noise. The hard drive anomaly notification is used to indicate that the hard drive of the second physical server has an abnormal noise, which is caused by the contact between the read / write head and the disk of the hard drive of the second physical server.

[0031] A fourth aspect of this application provides an anomaly detection apparatus, which includes a memory and a processor; the memory stores code, and the processor is configured to execute the code. When the code is executed, the apparatus performs the method described in the second aspect or any possible implementation of the second aspect.

[0032] A fifth aspect of this application provides a computer storage medium storing one or more instructions that, when executed by one or more computers, cause the one or more computers to perform the method as described in the second aspect or any possible implementation of the second aspect.

[0033] A sixth aspect of this application provides a computer program product storing instructions that, when executed by a computer, cause the computer to perform the method as described in the second aspect or any possible implementation of the second aspect.

[0034] In this embodiment, the anomaly detection device can be installed on a rack in a cloud data center. The hollow portion of the rack can house multiple physical servers, and the optical modules of these physical servers face the opening in the hollow portion of the rack. The anomaly detection device may include a moving part and a detector. Supported by the moving part, the detector can move freely at the opening in the hollow portion of the rack to face the optical modules of the multiple physical servers. During movement, the detector can send ultrasonic waves to the optical modules of these physical servers. The ultrasonic waves are reflected at the optical modules of these physical servers, and the reflected ultrasonic waves are received by the detector. Based on this, the detector can determine whether there is an anomaly in the optical modules of these physical servers. In the aforementioned process, since the anomaly detection device installed on the rack includes moving parts and detectors, it can move freely at the opening of the hollow part of the rack, thereby automatically performing anomaly detection on the optical modules of multiple physical servers in the rack. This eliminates the need for cloud data center operations and maintenance personnel to perform anomaly detection on the optical modules of multiple physical servers in the rack one by one. This automates the anomaly detection of the optical modules of physical servers, thereby reducing the manual cost required for anomaly detection and improving the efficiency and accuracy of anomaly detection. Attached Figure Description

[0035] Figure 1 is a schematic diagram of a communication system provided in an embodiment of this application;

[0036] Figure 2 is a structural schematic diagram of a rack provided in an embodiment of this application;

[0037] Figure 3 is a schematic diagram of the rack and anomaly detection device provided in an embodiment of this application;

[0038] Figure 4 is another schematic diagram of the rack and anomaly detection device provided in an embodiment of this application;

[0039] Figure 5 is a schematic diagram of a detector provided in an embodiment of this application;

[0040] Figure 6 is a flowchart illustrating an anomaly detection method provided in an embodiment of this application.

[0041] Figure 7 is a schematic diagram of a coordinate system based on rack construction provided in an embodiment of this application;

[0042] Figure 8 is another structural schematic diagram of the rack provided in an embodiment of this application;

[0043] Figure 9 is a structural schematic diagram of an anomaly detection device provided in an embodiment of this application. Detailed Implementation

[0044] This application provides an anomaly detection device and anomaly detection method, which can automate the anomaly detection of optical modules of physical servers, thereby reducing the manual cost required for anomaly detection and improving the efficiency and accuracy of anomaly detection.

[0045] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0046] With the rapid development of cloud technology, more and more tenants are choosing cloud data centers provided by cloud vendors to complete their business. Cloud data centers typically contain a large number of physical servers that work together to efficiently complete tenant business tasks and meet their needs. For example, cloud vendors can provide tenants with multiple physical servers to fulfill their neural network model training requirements.

[0047] In related technologies, to achieve rapid communication between multiple physical servers in a cloud data center, these physical servers are typically equipped with optical modules, and these optical modules are interconnected via optical fibers. This allows the optical modules of these physical servers to transmit optical signals, thus enabling communication between them. When two physical servers need to communicate, the source server generates an electrical signal to be transmitted. Its optical module then converts this electrical signal into an optical signal and transmits it to the optical module of the destination server. The destination server's optical module then converts the optical signal back into an electrical signal, processes it, and completes the communication between the two physical servers.

[0048] Since physical servers are typically deployed in cloud data centers, while optical modules are often exposed outside the server racks, they are susceptible to environmental contaminants (such as dust, grease, and unidentified liquids), which can lead to malfunctions. To determine if an optical module is malfunctioning, cloud providers often require maintenance personnel to inspect the optical modules of each physical server individually. This results in excessively high labor costs for anomaly detection, and the efficiency and accuracy of anomaly detection are both low.

[0049] To address the aforementioned issues, this application provides an anomaly detection device, which is typically deployed in a cloud service system. Figure 1 is a schematic diagram of a cloud service system provided in this application embodiment. As shown in Figure 1, the cloud service system includes infrastructure that provides cloud services to tenants and a cloud management platform that manages this infrastructure. The cloud management platform and the infrastructure are described in detail below:

[0050] A cloud management platform can centrally manage the infrastructure of the entire cloud service system. For example, following a tenant's instructions, the platform selects multiple physical servers from the infrastructure to meet the tenant's needs and runs the tenant's applications on these servers to fulfill their business requirements. The platform can also be accessible to tenants outside the cloud service system and respond to their requests. For instance, it can provide various interfaces, such as login and resource acquisition interfaces, for tenant clients (e.g., the tenant's terminal device or browser on that device) to access. The login interface allows the platform to authenticate a tenant's client, granting access upon successful authentication. Similarly, the resource acquisition interface allows the tenant's client to send resource acquisition requests. Based on these requests, the platform can select a number of physical servers from the infrastructure and allocate them to the tenant to meet their business needs.

[0051] The infrastructure comprises multiple cloud data centers that provide cloud services to the tenant. Each cloud data center may contain multiple racks, each rack may contain multiple physical servers, and each physical server may contain physical resources of a certain specification. For example, any one of these physical servers may contain any one or any combination of at least one computing resource (e.g., a central processing unit (CPU) and a graphics processing unit (GPU), etc.), at least one storage resource (e.g., memory and hard disks), and at least one network resource (e.g., a network interface card and an optical module), without any limitation.

[0052] It is worth noting that, as shown in Figure 2 (Figure 2 is a structural schematic diagram of a rack provided in an embodiment of this application), the several physical servers allocated by the cloud management platform to the tenant can be located in the same rack or multiple racks in the cloud data center. Since these physical servers can cooperate with each other to complete the tenant's business, these physical servers often need to communicate with each other. For example, after the CPU of the source physical server (hereinafter referred to as the source physical server) generates the electrical signal to be transmitted, it can send it to the network card of the source physical server, and then to the optical module of the source physical server. The optical module of the source physical server can convert the electrical signal into an optical signal and send the optical signal to the optical module of another physical server (hereinafter referred to as the destination physical server) through optical fiber. The optical module of the destination physical server can convert the optical signal into an electrical signal and send it to the network card of the destination physical server, and then to the CPU of the destination physical server for processing, thereby completing the communication between the source physical server and the destination physical server. Of course, this communication process usually also involves the hard disk of the source physical server and the hard disk of the destination physical server, which will not be elaborated here.

[0053] Since the optical module plays a crucial role in this communication process, it is necessary to monitor the optical modules of each physical server for anomalies in real time to ensure the stable operation of the communication process. Based on this, this application provides an anomaly detection device for cloud data center racks. This device can be used to detect whether the optical modules of each physical server in the rack are abnormal. Specifically, as shown in Figure 3 (Figure 3 is a schematic diagram of a rack and anomaly detection device provided in this application embodiment), an anomaly detection device is installed on a rack in a cloud data center. The rack is shell-shaped, and its hollow portion can hold multiple physical servers. The optical modules of these multiple physical servers face the opening in the hollow portion of the rack (because the optical modules of these multiple physical servers need to be connected to optical fibers, and optical fibers of a certain length need to extend outwards from the rack, they need to face the opening in the hollow portion of the rack).

[0054] The anomaly detection device may include a moving part and a detector. Supported by the moving part, the detector can move freely through an opening in the hollow section of the rack to face the optical modules of multiple physical servers within the rack. During movement, the detector can send ultrasonic waves to the optical modules of these physical servers. The ultrasonic waves are reflected by the optical modules, and the reflected ultrasonic waves are received by the detector. Based on this, the detector can determine whether there is an anomaly in the optical modules of these physical servers.

[0055] Specifically, the moving parts of the anomaly detection device may include a first slide rail (a slide rail along the horizontal direction (or x-axis direction), a second slide rail (a slide rail along the vertical direction (or y-axis direction), a first telescopic connecting arm, and a second telescopic connecting arm. The first slide rail is fixed to a first side of the frame (one side of the frame along the horizontal direction), and the second slide rail is fixed to a second side of the frame (one side of the frame along the vertical direction). The first side and the second side of the frame are perpendicular, therefore the first and second slide rails are also vertically arranged. Furthermore, since the first and second sides are two of the four sides constituting the hollow portion of the frame that are connected together, the first slide rail deployed outside the first side and the second slide rail deployed outside the second side effectively form a sliding structure around the hollow portion of the frame.

[0056] One end of the first retractable connecting arm is connected to the first slide rail, and the first retractable connecting arm is arranged perpendicular to the first slide rail. Therefore, the main body of the first retractable connecting arm is suspended in the opening of the hollow part of the frame. One end of the second retractable connecting arm is connected to the second slide rail, and the second retractable connecting arm is arranged perpendicular to the second slide rail. Therefore, the main body of the second retractable connecting arm is also suspended in the opening of the hollow part of the frame. Since the other end of the first retractable connecting arm is connected to the detector, and the other end of the second retractable connecting arm is connected to the detector, the detector is supported by the first and second retractable connecting arms. Therefore, the detector is also suspended in the opening of the hollow part of the frame.

[0057] It is worth noting that, since the first retractable connecting arm can slide freely on the first slide rail and can be freely stretched, and the second retractable connecting arm can slide freely on the second slide rail and can be freely stretched, the detector can move freely at the opening of the hollow part of the frame with the cooperation of the first slide rail, the second slide rail, the first retractable connecting arm and the second retractable connecting arm.

[0058] It should be understood that the above embodiments are only illustrated by taking the deployment of one anomaly detection device in one rack as an example. In practical applications, multiple racks can also share one anomaly detection device, as shown in Figure 4 (Figure 4 is another schematic diagram of the rack and anomaly detection device provided in the embodiments of this application).

[0059] Specifically, as shown in Figure 5 (Figure 5 is a structural schematic diagram of a detector provided in an embodiment of this application), the detector may include a processor (e.g., CPU, etc.), a memory (e.g., RAM or hard disk, etc.), a wireless communication device (e.g., network card, etc.), a microphone, an ultrasonic transducer, and a dust collector, etc.

[0060] The processor can receive requests from the cloud management platform and, based on these requests, control the sliding and extending of the first and second retractable connecting arms. This allows the entire detector to move within the opening of the hollow section of the rack, aligning with components such as the optical module, hard drive, battery, and fan of a physical server within the hollow section. Furthermore, the processor can control an ultrasonic transducer to send ultrasonic waves to the optical module (or battery) of a physical server and receive the reflected ultrasonic waves, then determine whether the optical module (or battery) of the physical server is malfunctioning based on the reflected ultrasonic waves. Additionally, the processor can control a microphone to collect sound from the hard drive (or fan) of a physical server and determine whether the hard drive (or fan) of the physical server is malfunctioning based on the collected sound.

[0061] The memory can store the detection results obtained by the processor and information from requests issued by the cloud management platform, etc.

[0062] The wireless communication device can communicate with the cloud management platform. For example, the wireless communication device can receive requests sent from the cloud management platform and forward the requests to the processor to perform anomaly detection. Or, the wireless communication device can receive notifications sent from the processor and forward the notifications to the cloud management platform to inform it of the anomaly detection results.

[0063] The microphone can capture sound under the control of the processor and send the captured sound to the processor as a digital signal for processing.

[0064] The ultrasonic transducer can send ultrasonic waves under the control of the processor and receive the reflected ultrasonic waves, so as to send the reflected ultrasonic waves to the processor for processing.

[0065] The dirt collector can collect dirt, such as dust and oil stains, under the control of the processor.

[0066] Of course, the detector may also include other components with other functions, which will not be elaborated here, but can be added to the detector in actual applications.

[0067] Furthermore, for the multiple servers serving a tenant, these physical servers may deploy one or more cloud instances as needed by the tenant. These cloud instances can run the tenant's applications to meet the tenant's business requirements. It should be noted that these cloud instances can be presented in various forms. For example, they can be virtual machines (VMs) created on cloud resources by a cloud management platform using virtualization technology; they can also be containers (Docker) created on cloud resources by a cloud management platform using virtualization technology; and they can also be microVMs created on cloud resources by a cloud management platform using virtualization technology, and so on.

[0068] Furthermore, for multiple cloud data centers in the infrastructure, these multiple cloud data centers can be located at the same site or different sites. These sites can be presented in various forms, such as regions, availability zones, and so on.

[0069] Based on the aforementioned anomaly detection device, it can be installed on a rack in a cloud data center. The hollow section of this rack can house multiple physical servers, with their optical modules facing an opening in the hollow section. The anomaly detection device includes a moving part and a detector. Supported by the moving part, the detector can move freely at the opening in the hollow section of the rack, facing the optical modules of the multiple physical servers. During movement, the detector sends ultrasonic waves to the optical modules of these physical servers. These ultrasonic waves are reflected by the optical modules, and the reflected ultrasonic waves are received by the detector. Based on this, the detector can determine whether there is an anomaly in the optical modules of these physical servers. In the aforementioned process, since the anomaly detection device installed on the rack includes a moving part and a detector, it can move freely at the opening of the hollow part of the rack, thereby automatically detecting anomalies in the optical modules of multiple physical servers within the rack. This eliminates the need for cloud data center maintenance personnel to manually perform anomaly detection on the optical modules of multiple physical servers in the rack, thus automating the anomaly detection of physical server optical modules, reducing the manual costs required for anomaly detection, and improving the efficiency and accuracy of anomaly detection. To further understand the working process of the anomaly detection device, the following description is based on Figure 6. Figure 6 is a flowchart illustrating an anomaly detection method provided in an embodiment of this application. As shown in Figure 6, this method can be implemented using the anomaly detection device shown in Figures 3 to 5. This device can be installed on a rack in a cloud data center, where multiple physical servers are housed in the hollow part of the rack. These multiple physical servers contain optical modules, and communication between them can be achieved through these optical modules. The optical modules of these multiple physical servers face the opening of the hollow part of the rack. The device may include a moving part and a detector, with one end of the moving part connected to the rack and the other end connected to the detector. The method includes:

[0070] 601. The detector moves at the opening with the support of the moving parts.

[0071] In this embodiment, when the detector needs to perform anomaly detection on the optical modules of multiple physical servers in the hollow part of the rack, it can send a control signal to the moving part so that the moving part drives the detector to move at the opening of the hollow part of the rack, thereby causing the detector to move toward the optical modules of multiple physical servers located in the hollow part of the rack.

[0072] It should be noted that the detector can sequentially send multiple control signals to the moving component according to its anomaly detection needs. Each control signal causes the moving component to move the detector from the opening in the hollow section of the rack to a specific position. At this position, the detector can face the optical module of a physical server within the hollow section to perform anomaly detection. In this way, these multiple control signals can cause the moving component to continuously move the detector from the opening in the hollow section of the rack to multiple specific positions. At these positions, the detector can face the optical modules of multiple physical servers within the hollow section to perform anomaly detection.

[0073] 602. During the movement of the detector, it sends ultrasonic waves to the optical modules of multiple physical servers, receives the reflected ultrasonic waves, and determines whether there are any abnormalities in the optical modules of multiple physical servers based on the reflected ultrasonic waves. The reflected ultrasonic waves are obtained by the reflection of ultrasonic waves at the optical modules of multiple physical servers.

[0074] During its movement, the detector can accurately target the optical modules of multiple physical servers within the hollow section of the rack. The detector sends ultrasonic waves to these modules, which reflect the waves, and these reflected waves are then received by the detector. Based on the reflected waves, the detector can determine if any anomalies exist in the optical modules of these physical servers. Thus, the detector completes its anomaly detection for the optical modules of these physical servers.

[0075] It should be noted that, as the detector moves, it sequentially faces the optical modules of each of the multiple physical servers. When the detector faces the optical module of a particular physical server, it can remain at that position and send ultrasonic waves to that server's optical module. These ultrasonic waves are reflected by the optical module and received by the detector. Therefore, the detector can determine whether there is an anomaly in the optical module of that physical server based on the reflected ultrasonic waves, thus completing the anomaly detection for that physical server's optical module. Then, the detector can continue to move to the optical module of the next physical server and perform anomaly detection on that next physical server's optical module (this process can be referred to in the previous section on anomaly detection of the optical module of that physical server, and will not be repeated here), until all the anomaly detection of the optical modules of all the physical servers is completed.

[0076] Specifically, the detector can send ultrasonic waves toward and to the optical modules of these multiple physical servers in the following manner:

[0077] Suppose that multiple physical servers in the hollow section of the rack are assigned to tenants by the cloud management platform. In order to ensure normal communication between these multiple physical servers to meet the business needs of the tenants, the cloud management platform can initiate anomaly detection of the optical modules of these multiple physical servers. Therefore, the cloud management platform can send an optical module detection request to the detector. The optical module detection request can be used to indicate the optical module detection task and the position of the optical modules of these multiple physical servers in the rack (the hollow section). For example, the cloud management platform can construct a coordinate system based on the rack, and the position of the optical modules of these multiple physical servers in the rack is the coordinate of the optical modules of these multiple physical servers in the coordinate system, etc.

[0078] Upon receiving the optical module detection request, the detector can parse it to obtain the optical module scanning task and the positions of the optical modules of the multiple physical servers within the rack. Based on the optical module detection task, the detector can determine that the optical modules of the multiple physical servers need to be detected for anomalies in a certain order (this order can be determined by the arrangement of the multiple physical servers within the rack, such as from top to bottom, etc., without limitation here). Then, based on the positions of the optical modules of the multiple physical servers within the rack, the detector can successively send multiple control signals to a moving component, causing the moving component to drive the detector to move sequentially through the opening in the hollow section of the rack. This allows the detector to sequentially move towards the optical modules of the multiple physical servers and send ultrasonic waves to them sequentially, thus completing the anomaly detection of the optical modules of the multiple physical servers.

[0079] For example, as shown in Figure 7 (Figure 7 is a schematic diagram of a coordinate system based on rack construction provided in an embodiment of this application), suppose a rack contains physical servers 1 to n, and each physical server can contain 4 optical modules. The cloud management platform can use the lower left corner of the rack (the opening in the hollow part) as the origin of the coordinate system, with coordinates (0,0). Assuming physical server 1 is placed at the top of the rack and physical server n is placed at the bottom, the cloud management platform can determine the coordinates of each physical server in the coordinate system and the coordinates of the four optical modules of each physical server in the coordinate system. For example, the coordinates of the lower left corner of physical server n (the side facing the opening) are (1,1), the coordinates of the upper left corner of physical server n are (13,1), the coordinates of the lower right corner of physical server n are (1,9), the coordinates of the upper right corner of physical server n are (13,9), the coordinates of the first optical module of physical server n are (3,8), the coordinates of the second optical module are (5,8), the coordinates of the third optical module are (8,8), and the coordinates of the fourth optical module are (11,8).

[0080] The cloud management platform can then send an optical module detection request to the detectors deployed in the rack. This request includes the optical module detection tasks for physical servers 1 to n, as well as the location information for physical servers 1 to n, as shown in Table 1.

[0081] Table 1

[0082] Based on the optical module detection request, the detector can obtain the optical module detection task and determine, based on the optical module detection task, that the optical modules of these n physical servers need to be detected abnormally in a top-down order (i.e., from physical server 1 to physical server n, as shown in Figure 8 (Figure 8 is another structural schematic diagram of the rack provided in the embodiment of this application)). The detector can send control signal 1 to the moving part, causing the moving part to move the detector at the opening of the hollow part of the rack, thereby aligning it with the four optical modules of physical server 1 and performing anomaly detection on these four optical modules. The detector can further send control signal n to the moving part, causing the moving part to move the detector at the opening of the hollow part of the rack, aligning it with the four optical modules of physical server n in the order (3,8)→(5,8)→(8,8)→(11,8) (this order can be determined by the detector itself or specified by the cloud management platform in the location information; it is not limited here, and the order is not unique and can be adjusted according to actual needs; it is also not limited here). Ultrasonic waves are then sent to these four optical modules sequentially to determine whether there is an anomaly based on the received reflected ultrasonic waves. At this point, the detector has completed the anomaly detection of the optical modules of these n physical servers.

[0083] More specifically, the detector can also perform the following operations:

[0084] After sending ultrasonic waves to the optical modules of these multiple physical servers, the detector can determine whether the optical modules of these physical servers are contaminated based on the reflected ultrasonic waves. If the reflected ultrasonic waves obtained from the optical modules of a certain physical server contain abnormal wave packets, it indicates that the optical modules of that physical server are contaminated. If the first physical server among these multiple physical servers (which can be one or more of these physical servers) is contaminated (the contamination is a certain type of abnormality in the optical module; of course, optical modules can also have other types of abnormalities, which will not be discussed here), the detector can send an optical module abnormality notification to the cloud management platform. This optical module abnormality notification is used to indicate that the optical modules of the first physical server are contaminated.

[0085] After receiving the optical module anomaly notification, the cloud management platform can determine that the optical module of the first physical server is dirty. Therefore, the cloud management platform can send an optical module processing request to the detector. This optical module processing request is used to indicate the optical module processing task and the position of the optical module of the first physical server in the rack.

[0086] Upon receiving the optical module processing request, the detector can parse it to obtain the optical module processing task and the position of the optical module of the first physical server within the rack. Based on the optical module processing task, the detector can determine to remove contaminants from the optical module of the first physical server. Next, based on the position of the optical module of the first physical server within the rack, the detector can send a control signal to a moving component, causing the component to move the detector towards the opening in the hollow section of the rack, thus moving the detector toward the optical module of the first physical server and removing contaminants from it. For example, the detector can send focused ultrasonic waves to the optical module of the first physical server, which vibrate the contaminants, causing them to detach from the optical module. Since the detector is facing the optical module of the first physical server, it can promptly collect the detached contaminants, thereby completing the decontamination of the optical module of the first physical server.

[0087] Continuing with the example above, suppose the reflected ultrasonic wave from the third optical module of physical server n contains an abnormal wave packet. The detector can determine that the third optical module of physical server n is contaminated. Therefore, the detector can send an optical module anomaly notification to the cloud management platform, indicating that the third optical module of physical server n is contaminated. Then, the cloud management platform can send an optical module processing request to the detector. This request indicates the optical module processing task and the coordinates of the third optical module of physical server n in the rack, i.e., (8,8). Based on this optical module processing request, the detector can obtain the optical module processing task and determine that the third optical module of physical server n needs to be cleaned of contamination. Then, based on the coordinates (8,8), the detector can send a control signal n+1 to the moving part, causing the moving part to move the detector at the opening of the hollow part of the rack, thereby aligning it with the third optical module of physical server n and sending focused ultrasonic waves to vibrate the contaminant on the third optical module of physical server n, causing the contaminant to fall off and be absorbed.

[0088] More specifically, the detector can also perform the following operations:

[0089] To ensure normal communication between these multiple physical servers to meet the business needs of tenants, the cloud management platform can initiate anomaly detection of the hard drives of these multiple physical servers. Therefore, the cloud management platform can send a hard drive detection request to the detector. The hard drive detection request can be used to indicate the hard drive detection task and the position of the hard drives of these multiple physical servers in the rack (the hollow part). For example, the cloud management platform can construct a coordinate system based on the rack, and the position of the hard drives of these multiple physical servers in the rack is the coordinate of the hard drives of these multiple physical servers in the coordinate system, etc.

[0090] Upon receiving the hard drive detection request, the detector can parse it to obtain the hard drive scanning task and the positions of the hard drives of the multiple physical servers within the rack. Based on the hard drive detection task, the detector can determine that the hard drives of the multiple physical servers need to be detected for anomalies in a certain order (this order can be determined by the arrangement of the multiple physical servers in the rack, such as from top to bottom, etc., without limitation). Next, based on the positions of the hard drives of the multiple physical servers within the rack, the detector can successively send multiple control signals to the moving parts, causing the moving parts to drive the detector to move successively at the openings in the hollow section of the rack. This allows the detector to sequentially move towards the hard drives of the multiple physical servers and collect the sounds generated by the hard drives of the multiple physical servers, in order to determine whether there are any anomalies in the hard drives of the multiple physical servers.

[0091] Continuing with the example above, each physical server (from physical server 1 to physical server n) can contain 6 hard drives. The cloud management platform can determine the coordinates of each physical server in this coordinate system, as well as the coordinates of each physical server's 6 hard drives in this coordinate system. For example, the coordinates of the first hard drive of physical server n are (2,5), the second hard drive is (4,5), the third hard drive is (6,5), the fourth hard drive is (8,5), the fifth hard drive is (10,5), and the sixth hard drive is (12,5).

[0092] The cloud management platform can then send a hard disk detection request to the detector deployed on the rack. This request includes the hard disk detection tasks for physical servers 1 to n, as well as the location information for physical servers 1 to n, as shown in Table 2.

[0093] Table 2

[0094] Based on the hard drive detection request, the detector receives the hard drive detection task and determines that the hard drives of the n physical servers need to be detected for anomalies in a top-to-bottom order. The detector sends control signal n+2 to the moving component, causing it to move the detector through the opening in the hollow section of the rack, sequentially aligning it with the six hard drives of physical server 1 and performing anomaly detection on them. The detector can then send control signal 2n+1 to the moving component, causing it to move the detector through the opening in the hollow section of the rack, sequentially aligning it with the six hard drives of physical server n in the order (2,5)→(4,5)→(6,5)→(8,5)→(10,5)→(12,5), and collecting the sounds emitted by each of the six hard drives to determine if any anomalies exist. At this point, the detector has completed the anomaly detection of the hard drives of the n physical servers.

[0095] More specifically, the detector can also perform the following operations:

[0096] After collecting the sounds emitted by the hard drives of these multiple physical servers, the detector can determine whether there are any abnormal noises from the hard drives of these physical servers. If the sound emitted by the hard drive of a certain physical server is of an abnormal frequency, it indicates that the hard drive of that physical server is making abnormal noises (usually caused by the contact between the hard drive's read / write head and the disk). If the second physical server (which can be one or more of these physical servers) makes abnormal noises (the abnormal noises indicate some kind of hard drive abnormality; of course, the hard drive can also have other types of abnormalities, which will not be discussed here), the detector can send a hard drive abnormality notification to the cloud management platform. This hard drive abnormality notification is used to indicate that the hard drive of the second physical server is making abnormal noises.

[0097] After receiving the notification of the hard drive malfunction, the cloud management platform can determine that there is an abnormal noise from the hard drive of the second physical server. Therefore, the cloud management platform sends a reminder to the maintenance personnel so that they can replace the hard drive of the second physical server with a new one.

[0098] Continuing with the example above, suppose the sound emitted by the fourth hard drive of physical server n is of an abnormal frequency. The detector can determine that there is an abnormal noise from the fourth hard drive of physical server n. Therefore, the detector can send a hard drive anomaly notification to the cloud management platform. This notification indicates that there is an abnormal noise from the fourth hard drive of physical server n. Then, the cloud management platform can remind the operations and maintenance personnel to replace the fourth hard drive of physical server n with a new one.

[0099] It should be understood that in this embodiment, the anomaly detection performed by the detector on the optical modules of the multiple physical servers and the anomaly detection performed by the detector on the hard drives of the multiple logistics servers can be performed separately or simultaneously. For example, when the cloud management platform sends optical module detection requests and hard drive detection requests to the detector one after the other, these two anomaly detection processes are performed separately. When the cloud management platform sends optical module detection requests and hard drive detection requests to the detector simultaneously, these two anomaly detection processes are performed simultaneously. For example, if these two anomaly detection processes are performed simultaneously, the detector can target the 4 optical modules and 6 hard drives of physical server n in the order of (3,8)→(5,8)→(8,8)→(11,8)→(2,5)→(4,5)→(6,5)→(8,5)→(10,5)→(12,5) and perform anomaly detection on these 4 optical modules and 6 hard drives in sequence.

[0100] It should also be understood that in this embodiment, the detector can also perform anomaly detection on the batteries and fans of these multiple physical servers. It should be noted that the anomaly detection process performed by the detector on the batteries of these multiple physical servers can refer to the anomaly detection process performed by the detector on the optical modules of these multiple physical servers, and the anomaly detection process performed by the detector on the fans of these multiple physical servers can refer to the anomaly detection process performed by the detector on the hard drives of these multiple physical servers. It will not be described again here.

[0101] In this embodiment, the anomaly detection device can be installed on a rack in a cloud data center. The hollow portion of the rack can house multiple physical servers, and the optical modules of these physical servers face the opening in the hollow portion of the rack. The anomaly detection device may include a moving part and a detector. Supported by the moving part, the detector can move freely at the opening in the hollow portion of the rack to face the optical modules of the multiple physical servers. During movement, the detector can send ultrasonic waves to the optical modules of these physical servers. The ultrasonic waves are reflected at the optical modules of these physical servers, and the reflected ultrasonic waves are received by the detector. Based on this, the detector can determine whether there is an anomaly in the optical modules of these physical servers. In the aforementioned process, since the anomaly detection device installed on the rack includes moving parts and detectors, it can move freely at the opening of the hollow part of the rack, thereby automatically performing anomaly detection on the optical modules of multiple physical servers in the rack. This eliminates the need for cloud data center operations and maintenance personnel to perform anomaly detection on the optical modules of multiple physical servers in the rack one by one. This automates the anomaly detection of the optical modules of physical servers, thereby reducing the manual cost required for anomaly detection and improving the efficiency and accuracy of anomaly detection.

[0102] The above is a detailed description of the anomaly detection method provided in the embodiments of this application. The anomaly detection device will be further described below. Figure 9 is a structural schematic diagram of the anomaly detection device provided in the embodiments of this application. As shown in Figure 9, the anomaly detection device includes an Eevee bond and a detector 900. The detector 900 can be used to implement the anomaly detection function in the embodiment corresponding to Figure 6. Specifically, the detector 900 includes: a receiver 901, a transmitter 902, a processor 903, and a memory 904 (wherein the number of processors 903 in the detector 900 can be one or more; Figure 9 uses one processor as an example). The processor 903 may include an application processor 9031 and a communication processor 9032. In some embodiments of this application, the receiver 901, transmitter 902, processor 903, and memory 904 can be connected via a bus or other means.

[0103] Memory 904 may include read-only memory and random access memory, and provides instructions and data to processor 903. A portion of memory 904 may also include non-volatile random access memory (NVRAM). Memory 904 stores processor and operation instructions, executable modules, or data structures, or subsets thereof, or extended sets thereof, wherein the operation instructions may include various operation instructions for implementing various operations.

[0104] The processor 903 controls the operation of the detector. In specific applications, the various components of the detector are coupled together through a bus system, which may include not only the data bus but also power buses, control buses, and status signal buses. However, for clarity, all buses are referred to as the bus system in the diagram.

[0105] The methods disclosed in the embodiments of this application can be applied to or implemented by the processor 903. The processor 903 can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by the integrated logic circuits in the hardware of the processor 903 or by instructions in software form. The processor 903 can be a general-purpose processor, a digital signal processor (DSP), a microprocessor, or a microcontroller, and may further include an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The processor 903 can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 904, and processor 903 reads the information from memory 904 and, in conjunction with its hardware, completes the steps of the above method.

[0106] Receiver 901 can be used to receive input digital or character information, and to generate signal inputs related to detector settings and function control. Transmitter 902 can be used to output digital or character information through the first interface; transmitter 902 can also be used to send instructions to the disk group through the first interface to modify data in the disk group; transmitter 902 may also include a display device such as a display screen.

[0107] In one embodiment of this application, the processor 903 is used to implement the various steps in the anomaly detection method in the embodiment corresponding to FIG6.

[0108] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0109] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0110] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0111] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0112] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. An anomaly detection device, characterized in that, The device is installed in a rack in a cloud data center. The hollow part of the rack is used to house multiple physical servers. Each physical server contains an optical module. The multiple physical servers communicate with each other through the optical modules of the multiple physical servers. The optical modules of the multiple physical servers face the opening in the hollow part. The device includes a movable component and a detector, one end of the movable component is connected to the frame, and the other end of the movable component is connected to the detector; The movable component is used to support the movement of the detector at the opening; The detector is used to send ultrasonic waves to the optical modules of the plurality of physical servers during movement, receive the reflected ultrasonic waves, and determine whether there is an abnormality in the optical modules of the plurality of physical servers based on the reflected ultrasonic waves, wherein the reflected ultrasonic waves are obtained by the ultrasonic waves being reflected at the optical modules of the plurality of physical servers.

2. The apparatus according to claim 1, characterized in that, The moving component includes a first slide rail, a second slide rail, a first telescopic connecting arm, and a second telescopic connecting arm; The first slide rail is fixed to a first side of the frame, and the second slide rail is fixed to a second side of the frame, wherein the first side and the second side are perpendicular to each other; One end of the first retractable connecting arm is connected to the first slide rail, and the other end of the first retractable connecting arm is connected to the detector; One end of the second retractable connecting arm is connected to the second slide rail, and the other end of the second retractable connecting arm is connected to the detector.

3. The apparatus according to claim 1 or 2, characterized in that, The detector is specifically used for: The system receives an optical module detection request sent by a cloud management platform. The optical module detection request is used to indicate the location of the optical modules of the multiple physical servers in the rack. The cloud management platform is used to manage the cloud data center. Based on the optical module detection request, it moves at the opening toward the optical modules of the plurality of physical servers and sends ultrasonic waves to the optical modules of the plurality of physical servers.

4. The apparatus according to any one of claims 1 to 3, characterized in that, The detector is also used for: If it is determined that the optical module of the first physical server among the plurality of physical servers is dirty, an optical module abnormality notification is sent to the cloud management platform. The optical module abnormality notification is used to indicate that the optical module of the first physical server is dirty. The system receives an optical module processing request sent by the cloud management platform, the optical module processing request being used to indicate the position of the optical module of the first physical server in the rack. Based on the optical module processing request, it moves at the opening toward the optical module of the first physical server and removes the dirt from the optical module of the first physical server.

5. The apparatus according to claim 4, characterized in that, The detector is specifically used for: A focused ultrasonic wave is sent to the optical module of the first physical server, and the focused ultrasonic wave is used to control the dirt to fall off the optical module of the first physical server. The dirt and grime that has been collected.

6. The apparatus according to any one of claims 1 to 5, characterized in that, The plurality of physical servers also include hard drives, and the detector is further used for: Receive a hard disk detection request sent by the cloud management platform, the hard disk detection request being used to indicate the location of the hard disks of the plurality of physical servers in the rack; Based on the hard drive detection request, the device moves at the opening toward the hard drives of the plurality of physical servers; The system collects the sounds generated by the hard drives of the multiple physical servers and determines whether there are any abnormalities in the hard drives of the multiple physical servers based on the sounds generated by the hard drives.

7. The apparatus according to claim 6, characterized in that, The detector is also used for: If it is determined that the hard drive of the second physical server among the plurality of physical servers is making abnormal noise, a hard drive abnormality notification is sent to the cloud management platform. The hard drive abnormality notification is used to indicate that the hard drive of the second physical server is making the abnormal noise, which is caused by the contact between the read / write head and the disk of the hard drive of the second physical server.

8. An anomaly detection method, characterized in that, The method is implemented by an anomaly detection device, which is installed in a rack in a cloud data center. The hollow part of the rack is used to house multiple physical servers. Each physical server contains an optical module, and the multiple physical servers communicate with each other through the optical modules of the multiple physical servers. The optical modules of the multiple physical servers face the opening of the hollow part. The device includes a movable component and a detector, one end of the movable component is connected to the frame, and the other end of the movable component is connected to the detector. The method includes: The detector moves at the opening with the support of the moving component; During its movement, the detector sends ultrasonic waves to the optical modules of the multiple physical servers, receives the reflected ultrasonic waves, and determines whether there is an abnormality in the optical modules of the multiple physical servers based on the reflected ultrasonic waves. The reflected ultrasonic waves are obtained by the ultrasonic waves being reflected at the optical modules of the multiple physical servers.

9. The method according to claim 8, characterized in that, The moving component includes a first slide rail, a second slide rail, a first telescopic connecting arm, and a second telescopic connecting arm; The first slide rail is fixed to a first side of the frame, and the second slide rail is fixed to a second side of the frame, wherein the first side and the second side are perpendicular to each other; One end of the first retractable connecting arm is connected to the first slide rail, and the other end of the first retractable connecting arm is connected to the detector; One end of the second retractable connecting arm is connected to the second slide rail, and the other end of the second retractable connecting arm is connected to the detector.

10. The method according to claim 8 or 9, characterized in that, The movement of the detector at the opening includes: The detector receives an optical module detection request sent by the cloud management platform. The optical module detection request is used to indicate the position of the optical modules of the multiple physical servers in the rack. The cloud management platform is used to manage the cloud data center. The detector moves at the opening based on the optical module detection request to face the optical modules of the plurality of physical servers and sends ultrasonic waves to the optical modules of the plurality of physical servers.

11. The method according to any one of claims 8 to 10, characterized in that, The method further includes: If the detector determines that the optical module of the first physical server among the plurality of physical servers is dirty, it sends an optical module abnormality notification to the cloud management platform. The optical module abnormality notification is used to indicate that the optical module of the first physical server is dirty. The detector receives an optical module processing request sent by the cloud management platform. The optical module processing request is used to indicate the position of the optical module of the first physical server in the rack. The detector moves at the opening based on the optical module processing request to face the optical module of the first physical server and removes the dirt from the optical module of the first physical server.

12. The method according to claim 11, characterized in that, The detector removes the dirt from the optical module of the first physical server by: The detector sends focused ultrasonic waves to the optical module of the first physical server, and the focused ultrasonic waves are used to control the dirt to fall off the optical module of the first physical server. The detector collects the dirt that has fallen off.

13. The method according to any one of claims 8 to 12, characterized in that, The plurality of physical servers also include hard disks, and the method further includes: The detector receives a hard disk detection request sent by the cloud management platform, and the hard disk detection request is used to indicate the location of the hard disks of the multiple physical servers in the rack; The detector moves at the opening based on the hard drive detection request, toward the hard drives of the plurality of physical servers; The detector collects the sounds generated by the hard drives of the multiple physical servers and determines whether there are any abnormalities in the hard drives of the multiple physical servers based on the sounds generated by the hard drives of the multiple physical servers.

14. The method according to claim 13, characterized in that, The method further includes: If the detector determines that the hard drive of the second physical server among the plurality of physical servers is making abnormal noise, it sends a hard drive abnormality notification to the cloud management platform. The hard drive abnormality notification is used to indicate that the hard drive of the second physical server is making the abnormal noise, which is caused by the contact between the read / write head and the disk of the hard drive of the second physical server.

15. A cloud service system, characterized in that, The cloud service system includes a cloud data center, which includes racks and anomaly detection devices. The hollow portion of the rack is used to house multiple physical servers, each of which contains an optical module. The multiple physical servers communicate with each other through the optical modules of the multiple physical servers, and the optical modules of the multiple physical servers face the opening of the hollow portion. The device includes a movable component and a detector, one end of the movable component is connected to the frame, and the other end of the movable component is connected to the detector; The movable component is used to support the movement of the detector at the opening; The detector is used to send ultrasonic waves to the optical modules of the plurality of physical servers during movement, receive the reflected ultrasonic waves, and determine whether there is an abnormality in the optical modules of the plurality of physical servers based on the reflected ultrasonic waves, wherein the reflected ultrasonic waves are obtained by the ultrasonic waves being reflected at the optical modules of the plurality of physical servers.

16. An anomaly detection device, characterized in that, The device includes a memory and a processor; the memory stores code, and the processor is configured to execute the code, wherein when the code is executed, the device performs the method as described in any one of claims 8 to 14.

17. A computer storage medium, characterized in that, The computer storage medium stores one or more instructions that, when executed by one or more computers, cause the one or more computers to perform the method of any one of claims 8 to 14.

18. A computer program product, characterized in that, The computer program product stores instructions that, when executed by a computer, cause the computer to perform the method described in any one of claims 8 to 14.

Images