CN115601700A

CN115601700A - Method, device and medium for monitoring use of data center network device

Info

Publication number: CN115601700A
Application number: CN202211285869.6A
Authority: CN
Inventors: 郝虹; 高岩
Original assignee: Shandong New Generation Information Industry Technology Research Institute Co Ltd
Current assignee: Shandong New Generation Information Industry Technology Research Institute Co Ltd
Priority date: 2022-10-20
Filing date: 2022-10-20
Publication date: 2023-01-13

Abstract

The embodiment of the specification discloses a method, equipment and a medium for monitoring the use of data center network equipment, which relate to the technical field of artificial intelligence, and the method comprises the following steps: acquiring an interface use image acquired by the inspection robot at a specified inspection task point, acquiring a specified template image corresponding to a preset specified inspection task point, wherein the specified template image is an image under a normal use condition; inputting the interface use image and the specified template image into a pre-trained interface use detection model, and outputting a detection annotation file with the same size as the interface use image and the specified template image, wherein the detection annotation file comprises the category identification of the pixel points; the method comprises the steps of obtaining the category identification of each pixel point in a detection label file, determining the interface use state of the network equipment at the appointed routing inspection task point according to the category identification of each pixel point, and monitoring the use of the data center network equipment based on the interface use state, wherein the interface use state comprises an abnormal use state and a normal use state.

Description

Method, device and medium for monitoring use of data center network device

Technical Field

The present disclosure relates to the field of artificial intelligence technologies, and in particular, to a method, a device, and a medium for monitoring usage of a data center network device.

Background

With the development of internet technology, data centers are becoming larger and larger. The network data transmission amount of the large-scale data center is large, the number of laid network lines is large, and the inspection difficulty of network communication equipment is increased. If the abnormal use behavior of the network equipment exists, unnecessary loss can be caused if the abnormal use behavior of the network equipment is not found in time, and serious consequences can be caused, such as illegal private network connection, stealing of data center network resources, malicious network disconnection of a server caused by network cable unplugging, even stealing of communication data and other abnormal use behaviors. Under the general condition, arrange the full-time staff to carry out the manual work and patrol and examine, on the one hand can't guarantee incessant patrol and examine, on the other hand to numerous network circuit of data center or intensive net twine on a certain switch, router, even appear a private connection optic fibre, people's eye also is lost easily, causes the missed measure. For similar tasks, the inspection robot has obvious advantages.

The robot patrols and examines mainly through high definition camera shooting equipment image, monitors equipment. Due to the fact that the equipment is used abnormally, especially vision change caused by the interface abnormal use is small, and clear trace characteristics do not exist, the existing detection accuracy and recall ratio are not high, and the problem that the interface abnormal use cannot be timely and accurately monitored.

Disclosure of Invention

One or more embodiments of the present specification provide a method, a device, and a medium for monitoring usage of a data center network device, so as to solve the following technical problems: due to the fact that the equipment is used abnormally, especially vision change caused by the interface abnormal use is small, and clear trace characteristics do not exist, the existing detection accuracy and recall ratio are not high, and the problem that the interface abnormal use cannot be timely and accurately monitored.

One or more embodiments of the present disclosure adopt the following technical solutions:

one or more embodiments of the present disclosure provide a usage monitoring method for a data center network device, which is applied to an inspection robot, and the method includes: acquiring an interface use image acquired by an inspection robot at a specified inspection task point, and acquiring a preset specified template image corresponding to the specified inspection task point, wherein the image size of the interface use image is the same as that of the specified template image, and the specified template image is an image of the specified inspection task point under the condition that network equipment is normally used; inputting the interface use image and the designated template image into a pre-trained interface use detection model, and outputting a detection annotation file with the same size as the interface use image and the designated template image, wherein the detection annotation file comprises the category identification of each pixel point; the method comprises the steps of obtaining category identification of each pixel point in the detection labeling file, determining interface use state of network equipment at the appointed routing inspection task point according to the category identification of each pixel point, and monitoring use of the network equipment of the data center based on the interface use state, wherein the interface use state comprises an abnormal use state and a normal use state.

Further, before inputting the interface usage image and the specified template image into the interface usage detection model trained in advance, the method further includes: constructing an initial model using a neural network structure fusing multi-scale features, wherein the initial model comprises a feature extraction network using convolution and pooling operations and an upsampling network using upsampling and feature map fusion.

Further, before inputting the interface usage image and the specified template image into the interface usage detection model trained in advance, the method further includes: generating a simulation data set, specifically comprising: acquiring an interface element image corresponding to an interface, wherein the interface element image comprises a plurality of specified interface images, and each specified interface image comprises an interface idle state image or an interface occupied state image; acquiring a plurality of template images, wherein the template images are images of a plurality of network devices in normal use; covering the interface element images in each template image according to a random covering rule, and generating an initial interface abnormal use image corresponding to each template image; generating a designated annotation file with the same size as the initial interface abnormal use image according to the initial interface abnormal use image, wherein the designated annotation file comprises a violation interface identifier and a normal interface identifier; carrying out color space transformation and geometric space transformation on the abnormal use image corresponding to each template image to generate an interface abnormal use image corresponding to each template image; forming an image pair by the interface abnormal use image corresponding to each template image and the template image corresponding to the interface abnormal use image; and generating a simulation data set according to the image pair and the specified annotation file.

Further, after generating the simulation data set, the method further comprises: training the initial model by using the data in the simulation data set to obtain an interface use detection model meeting the requirements, specifically comprising: inputting the image pairs in the simulation data set into the initial model, and outputting a presumed annotation file; and adjusting parameters of the initial model according to the presumed annotation file and the appointed annotation file in the simulation data set to obtain an interface use detection model meeting the requirements.

Further, according to the category identification of each pixel point, determining the interface use state of the network device at the specified routing inspection task point, specifically comprising: judging whether the class identification of each pixel point in the detection label file has a first class identification, and if the first class identification exists, judging that the interface use state at the appointed routing inspection task point is abnormal use; the class identification comprises a first class identification and a second class identification, the first class identification is used for indicating that the interface state corresponding to the pixel point is an abnormal use state, and the second class identification is used for indicating that the interface state corresponding to the pixel point is a normal use state; according to the category identification, dividing the pixel points in the detection labeling file into a first category pixel point and a second category pixel point, wherein the category identification of each pixel point in the first category pixel point is a first category identification, and the category identification of each pixel point in the second category pixel point is a second category identification; taking an area formed by the first type of pixel points as an interface abnormal use area, and determining area position coordinate data of the interface abnormal use area; and according to the area position coordinate data, visually displaying the interface abnormal use area in the interface use image and the specified template image.

Further, before acquiring a preset specified template image corresponding to the specified inspection task point, the method further includes: presetting a plurality of inspection task points; at the inspection task point, acquiring a template image through an inspection robot, and recording the robot posture state of the inspection robot; establishing a corresponding relation between the template image and the pose state of the robot as a first corresponding relation; and setting a place identifier for each inspection task point, and establishing a corresponding relation between the inspection task point and the template image as a second corresponding relation according to the place identifier of each inspection task point and the template image acquired by each inspection task.

Further, the interface that acquires inspection robot and appoint to patrol and examine the collection of task point department uses the image, specifically includes: determining a designated place identification corresponding to the designated inspection task point; according to the appointed place identification, an appointed template image corresponding to the appointed routing inspection task point is determined in the second corresponding relation; determining the position and posture state of the designated robot corresponding to the designated template image in the first corresponding relation based on the designated template image corresponding to the designated inspection task point; and using the posture state of the appointed robot, and acquiring an interface use image at the appointed routing inspection task point.

Further, before inputting the interface usage image and the specified template image into a pre-trained interface usage detection model, the method further includes: and according to a preset image alignment mode, performing alignment transformation on the interface use image and the specified template image to obtain an interface use image after the alignment transformation, so that the interface use image after the alignment transformation and the specified template image are input into a pre-trained interface use detection model.

One or more embodiments of the present specification provide a usage monitoring device for a data center network device, including:

at least one processor; and (c) a second step of,

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores instructions executable by the at least one processor to enable the at least one processor to:

acquiring an interface use image acquired by an inspection robot at a specified inspection task point, and acquiring a preset specified template image corresponding to the specified inspection task point, wherein the image size of the interface use image is the same as that of the specified template image, and the specified template image is an image of the network equipment at the specified inspection task point under the normal use condition; inputting the interface use image and the designated template image into a pre-trained interface use detection model, and outputting a detection annotation file with the same size as the interface use image and the designated template image, wherein the detection annotation file comprises the category identification of each pixel point; the method comprises the steps of obtaining category identification of each pixel point in the detection labeling file, determining interface use state of network equipment at the appointed routing inspection task point according to the category identification of each pixel point, and monitoring use of the network equipment of the data center based on the interface use state, wherein the interface use state comprises an abnormal use state and a normal use state.

One or more embodiments of the present specification provide a non-transitory computer storage medium storing computer-executable instructions configured to:

acquiring an interface use image acquired by an inspection robot at a specified inspection task point, and acquiring a preset specified template image corresponding to the specified inspection task point, wherein the image size of the interface use image is the same as that of the specified template image, and the specified template image is an image of the specified inspection task point under the condition that network equipment is normally used; inputting the interface use image and the designated template image into a pre-trained interface use detection model, and outputting a detection annotation file with the same size as the interface use image and the designated template image, wherein the detection annotation file comprises the category identification of each pixel point; the method comprises the steps of obtaining category identification of each pixel point in the detection labeling file, determining interface use state of network equipment at the appointed routing inspection task point according to the category identification of each pixel point, and monitoring use of the network equipment of the data center based on the interface use state, wherein the interface use state comprises an abnormal use state and a normal use state.

The embodiment of the specification adopts at least one technical scheme which can achieve the following beneficial effects: through the technical scheme, the detection annotation file of the category identification of each pixel point is generated according to the actually acquired interface use image and the template image at the position, and compared with the method of only using the actual image, the obtained result has higher accuracy and better accords with the equipment characteristics of the current inspection task point. According to the detection marking file, the interface use state at the task point is determined, the interface use difference is presented in a type identification mode, and the small visual change and the undefined trace characteristics caused by illegal use are quantitatively displayed, so that the detection accuracy of illegal use of the interface is improved, and the illegal use problem of the interface of the data center network equipment can be timely found.

Drawings

In order to more clearly illustrate the embodiments of the present specification or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only some embodiments described in the present specification, and for those skilled in the art, other drawings can be obtained according to the drawings without any creative effort. In the drawings:

fig. 1 is a schematic flowchart of a usage monitoring method for a data center network device according to an embodiment of the present disclosure;

fig. 2 is a schematic structural diagram of a usage monitoring device of a data center network device according to an embodiment of the present disclosure.

Detailed Description

In order to make those skilled in the art better understand the technical solutions in the present specification, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only a part of the embodiments of the present specification, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present specification without any creative effort shall fall within the protection scope of the present specification.

With the development of internet technology, data centers are becoming larger and larger. The network data transmission amount of the large-scale data center is large, the number of laid network lines is large, and the inspection difficulty of network communication equipment is increased. If the abnormal use behavior of the network equipment exists, unnecessary loss can be caused if the abnormal use behavior of the network equipment is not found in time, and serious consequences can be caused, such as illegal private network connection, stealing of data center network resources, malicious network disconnection of a server caused by network cable unplugging, even stealing of communication data and other abnormal use behaviors. Under the general condition, arranging the full-time staff to carry out manual inspection, on one hand, the uninterrupted inspection cannot be guaranteed, and on the other hand, even if a private optical fiber is connected to numerous network lines of a data center or dense network lines on a certain switch and a router, human eyes are easily lost and missed, so that the inspection is missed. For similar tasks, the inspection robot has obvious advantages.

The robot patrols and examines mainly through high definition camera shooting equipment image, monitors equipment. Due to the fact that the equipment is used abnormally, particularly vision change caused by the abnormal use of the interface is small, and clear trace characteristics do not exist, the existing detection accuracy and recall ratio are not high, and the problem that the abnormal use of the interface cannot be timely and accurately monitored.

The embodiment of the specification provides a use monitoring method of data center network equipment, which is applied to an inspection robot. The execution subject in the embodiments of the present specification may be a server, or may be any device having a data processing capability. Fig. 1 is a schematic flow chart of a usage monitoring method for data center network devices provided in an embodiment of the present specification, as shown in fig. 1, the method mainly includes the following steps:

and S101, acquiring an interface use image acquired by the inspection robot at the specified inspection task point, and acquiring a preset specified template image corresponding to the specified inspection task point.

Wherein the image size of the interface use image is the same as the image size of the specified template image.

Before acquiring the preset appointed template image corresponding to the appointed routing inspection task point, the method further comprises the following steps: presetting a plurality of inspection task points; at the inspection task point, acquiring a template image through an inspection robot, and recording the robot posture state of the inspection robot; establishing a corresponding relation between the template image and the pose state of the robot as a first corresponding relation; and setting a place identifier for each inspection task point, and establishing a corresponding relation between the inspection task point and the template image as a second corresponding relation according to the place identifier of each inspection task point and the template image acquired by each inspection task.

In an embodiment of the specification, the inspection robot is initialized and deployed, a plurality of inspection task points are set in advance for a data center, the inspection task points are recorded in an inspection system, the inspection robot is supported by the inspection system to execute an inspection task, and the inspection robot is controlled to reach an inspection task point execution interface to use a monitoring task. In addition, because the number of network devices and the composition of device lines in each inspection task point are different, in order to ensure the integrity of the collected interface use images, the posture state of the robot of the inspection robot, that is, the shooting angle of the inspection robot, needs to be set at each inspection task point. And at each inspection task point, acquiring a template image corresponding to the inspection task point through the inspection robot, wherein the template image refers to an image under the normal use condition of the network equipment. And recording the pose state of the robot for collecting the template image, and establishing the corresponding relation between the template image and the pose state of the robot. In addition, in order to facilitate the control of the inspection robot, a location identifier needs to be set for each inspection task point, and a corresponding relation between the location identifier and the template image corresponding to each inspection task point is established according to the location identifier and the template image corresponding to each inspection task point.

In an embodiment of the present specification, since the number of the network devices and the composition of the device lines in each routing inspection task point are different, that is, the routing inspection complexity of each routing inspection task point is different, the number of the template images required by each routing inspection task point is determined according to the routing inspection complexity of the routing inspection task point. For example, the number of template images required by each inspection task point can be determined according to the number of network devices at each inspection task point, wherein the number of template images comprises one or more; when the number of template images required by the inspection task point is one, acquiring one template image through the inspection robot at the inspection task point based on the position and posture state of the designated robot; when the number of the template images required by the inspection task point is multiple, the inspection robot acquires the multiple template images at the inspection task point based on the position and posture state of the designated robot.

The interface that acquires inspection robot and patrols and examines the collection of task point department at appointed uses the image, specifically includes: determining a designated place identification corresponding to the designated inspection task point; according to the appointed place mark, an appointed template image corresponding to the appointed routing inspection task point is determined in the second corresponding relation; determining the position and posture state of the designated robot corresponding to the designated template image in the first corresponding relation based on the designated template image corresponding to the designated inspection task point; and using the posture state of the appointed robot, and acquiring an interface use image at the appointed routing inspection task point.

In an embodiment of the present description, when it is required to monitor device usage of network devices in a data center, an inspection robot is controlled to reach an appointed inspection task point, and an appointed location identifier corresponding to the appointed inspection task point is determined. And according to the appointed place identification, determining the appointed template image corresponding to the appointed routing inspection task point in the corresponding relation between the place identification and the template image corresponding to each routing inspection task point. And then, according to the specified template image corresponding to the specified inspection task point, determining the position and posture state of the specified robot corresponding to the specified template image in the corresponding relation between the template image and the position and posture state of the robot. And controlling the inspection robot to use the image of the designated robot position and posture state acquisition interface at the designated inspection task point.

And S102, inputting the interface use image and the specified template image into a pre-trained interface use detection model, and outputting a detection annotation file with the same size as the interface use image and the specified template image.

The detection labeling file may be a mask map, and the mask map includes the category identifier of each pixel point.

Before inputting the interface usage image and the specified template image into the interface usage detection model trained in advance, the method further includes: and according to a preset image alignment mode, performing alignment transformation on the interface use image and the specified template image to obtain an interface use image after the alignment transformation, so that the interface use image after the alignment transformation and the specified template image are conveniently input into a pre-trained interface use detection model.

In one embodiment of the present specification, an alignment tool or an alignment method for assisting image alignment is preset, and the alignment tool or the alignment method is deployed at an initialization stage of the inspection robot. After the inspection robot acquires the interface use image, the interface use image and the template image corresponding to the appointed inspection task point are aligned by using an image alignment tool or an image alignment method, the aligned and transformed interface use image and the appointed template image are input into a pre-trained interface use detection model, and a detection annotation file with the same size as the interface use image and the appointed template image is output.

Before inputting the interface usage image and the specified template image into the interface usage detection model trained in advance, the method further includes: an initial model is constructed using a neural network structure that fuses multi-scale features, wherein the initial model includes a feature extraction network using convolution and pooling operations and an upsampling network using upsampling and feature map fusion.

In an embodiment of the present specification, since the illegal interface portion occupies fewer pixels, only one percent, or even less pixels, compared to the entire image, in this embodiment, when an initial model is constructed, a neural network structure is adopted, and the neural network structure fuses multi-scale features, the first half is a feature extraction network mainly formed by convolution and pooling operations, and the second half is an upsampling network formed by upsampling and feature map fusing. And in the up-sampling network part, every time the up-sampling is carried out, the up-sampling network part is fused with the channel number corresponding to the feature extraction network part in the same scale, and then the up-sampling is continued until the feature graph scale is consistent with the size of the template. The output of the last layer of the model is called as an illegal interface probability thermodynamic diagram, the illegal interface probability thermodynamic diagram consists of two channels, the value of each pixel point in each channel is set to be between 0 and 1, the channel 1 is set to represent the probability that the illegal interface condition does not occur, and the channel 2 represents the probability that the illegal interface condition occurs. And comparing the values of the same pixel point in two channels in the violation interface probability thermodynamic diagram, taking the larger value as the class identifier of the pixel point, and so on to obtain the class identifier of each pixel point, and generating the finally output label file according to the class identifier of each pixel point.

Before inputting the interface usage image and the specified template image into the interface usage detection model trained in advance, the method further includes: generating a simulation data set, specifically comprising: acquiring an interface element image corresponding to an interface, wherein the interface element image comprises a plurality of specified interface images, and each specified interface image comprises an interface idle state image or an interface occupation state image; acquiring a plurality of template images, wherein the template images are images of a plurality of network devices under normal use; covering the interface element image in each template image according to a random covering rule to generate an initial interface abnormal use image corresponding to each template image; generating a designated annotation file with the same size as the initial interface abnormal use image according to the initial interface abnormal use image, wherein the designated annotation file comprises a violation interface identifier and a normal interface identifier; carrying out color space transformation and geometric space transformation on the abnormal use image corresponding to each template image to generate an interface abnormal use image corresponding to each template image; forming an image pair by the interface abnormal use image corresponding to each template image and the template image corresponding to the interface abnormal use image; a simulated data set is generated from the pair of images and the specified annotation file.

In one embodiment of the present specification, after the model is built, model training needs to be performed on the model, where a training data set needs to be used. Because the illegal use condition of the network equipment is rare and the samples are rare, the method for efficiently generating the simulation data set is provided and is used for simulating the image collected by the real inspection scene and automatically generating the simulation training data set. The method of generating a simulated data set includes generating an initial interface anomaly use image and generating a visual change in the image.

The initial interface illegal use image generation needs an interface element image and a template image corresponding to the inspection task point, the interface element image comprises a plurality of interface images, the plurality of interface images comprise different types of interfaces and different states of each interface, the types of the interfaces comprise network interfaces, optical fiber interfaces and the like, and the states of each interface comprise an interface occupation state and an interface idle state. The template image is an image of a plurality of network devices collected by the inspection robot under normal use conditions. And covering the interface element image in each template image according to a random covering rule to generate an initial interface abnormal use image corresponding to each template image, wherein the random covering rule refers to that the interface type is random, the interface position is random and the interface number is random, and the interface type is covered in the template image.

In order to enhance the adaptability of the model in an actual task, the generated interface illegally uses an image and needs to perform image visual transformation, wherein the image visual transformation refers to the transformation of the image in a color space and a geometric space, the color space mainly transforms brightness, contrast and color tone, the conditions of illumination, camera shooting parameter change and the like in the actual task are simulated, the geometric space mainly simulates slight affine transformation and perspective transformation, and slight change of a camera shooting angle caused by robot attitude control errors in the actual condition is simulated.

In addition, the initial interface abnormal use image is generated, and simultaneously, a specified annotation file with the same size as the initial interface abnormal use image is generated, and the annotation file is expected to be output by the model. The specified labeling file comprises an illegal interface identifier and a normal interface identifier, wherein the illegal interface identifier can be set to be 1, the normal interface identifier can be set to be 0, the normal interface identifier can also be called a non-illegal interface identifier, and the illegal interface identifier can also be called an abnormal interface identifier. And forming a pair of images, namely an image pair, by using the interface abnormal use image corresponding to each template image and the template image corresponding to the interface abnormal use image, and generating a simulation data set according to the image pair and the specified annotation file.

That is, one template image corresponds to one or more generated interface illegal use images, and the training data input by the model in this embodiment is composed of or generated by the template image and the interface illegal use image generated corresponding thereto, that is, one training data refers to a pair of images and an annotation file, one of the pair of images is the template image, and the other is the interface illegal use image generated by the template.

After generating the simulated data set, the method further comprises: training the initial model by using the data in the simulation data set to obtain an interface use detection model meeting the requirements, and specifically comprising the following steps: inputting the image pair in the simulation data set into the initial model, and outputting a presumed annotation file; and adjusting parameters of the initial model according to the presumed label file and the specified label file in the simulation data set to obtain an interface use detection model meeting the requirements.

In an embodiment of the present specification, the initial model is trained using data in the simulation data set to obtain a satisfactory interface usage detection model, the image pair in the simulation data set is input into the initial model, the inference markup file is output, and the initial model is subjected to parameter adjustment according to the inference markup file and the specified markup file in the simulation data set to obtain a satisfactory interface usage detection model. Further, a learning target of the model is defined as a difference between the inferred markup document and the specified markup document, such as a sum of squares of differences, cross entropy, and the like, which can be used for the learning target.

Step S103, obtaining the category identification of each pixel point in the detection label file, determining the interface use state of the network equipment at the appointed routing inspection task point according to the category identification of each pixel point, and monitoring the use of the data center network equipment based on the interface use state.

Wherein, the interface use state comprises an abnormal use state and a normal use state.

According to the category identification of each pixel point, the interface use state of the network equipment at the appointed routing inspection task point is determined, and the method specifically comprises the following steps: judging whether the class identifier of each pixel point in the detection label file has a first class identifier, and if the first class identifier exists, judging that the interface use state of the specified inspection task point is abnormal use; the category identification comprises a first category identification and a second category identification, wherein the first category identification is used for indicating that the interface state corresponding to the pixel point is an abnormal use state, and the second category identification is used for indicating that the interface state corresponding to the pixel point is a normal use state; dividing the pixel points in the detection labeling file into a first type of pixel points and a second type of pixel points according to the type identification, wherein the type identification of each pixel point in the first type of pixel points is a first type identification, and the type identification of each pixel point in the second type of pixel points is a first type identification; taking the area formed by the first type of pixel points as an interface abnormal use area, and determining area position coordinate data of the interface abnormal use area; and according to the area position coordinate data, visually displaying the interface abnormal use area on the interface use image and the specified template image.

In an embodiment of the present specification, a category identifier corresponding to each pixel point in a detection markup file is obtained according to the output detection markup file. It should be noted here that the detection markup file may be a mask map, and the size of the mask map is the same as the size of the input interface use image and the image of the specified template image. Judging whether a first class identifier exists in the class identifier of each pixel point in the detection label file, if so, judging that the interface use state of the specified inspection task point is an abnormal use state, wherein the class identifier comprises the first class identifier and a second class identifier, the first class identifier is used for indicating that the interface state corresponding to the pixel point is the abnormal use state, namely, the first class identifier can be called as an illegal interface identifier, and the illegal interface identifier arranged in the constructed analog data set is 1, so that the first class identifier is 1. The second class identifier is used to indicate that the interface state corresponding to the pixel point is a normal use state, and similarly, if the normal interface identifier set in the constructed analog data set is 0, the second class identifier is 0 here. If the type identification of the pixel point in the detection label file is 1, judging that the interface use state of the network equipment at the routing inspection task point is in an abnormal use state. And screening first-type pixel points in the detection labeling file, namely a large-block communication area with the category identifier of 1, taking an area formed by the first-type pixel points as an interface abnormal use area, and determining area position coordinate data of the interface abnormal use area. Because the size of the detection labeling file is the same as the size of the interface use image and the designated template image, the interface abnormal use area can be corresponded to the interface use image and the designated template image according to the area position coordinate data of the interface abnormal use area in the detection labeling file for visual display, and the use of the data center network equipment can be monitored by further auditing and confirming according to the display content.

Through the technical scheme, the detection annotation file of the category identification of each pixel point is generated according to the actually acquired interface use image and the template image at the position, and compared with the method of only using the actual image, the obtained result has higher accuracy and better accords with the equipment characteristics of the current inspection task point. According to the detection marking file, the interface use state at the task point is determined, the interface use difference is presented in a type identification mode, and the small visual change and the undefined trace characteristics caused by illegal use are quantitatively displayed, so that the detection accuracy of illegal use of the interface is improved, and the illegal use problem of the interface of the data center network equipment can be timely found.

An embodiment of the present specification further provides a usage monitoring device for a data center network device, as shown in fig. 2, the device includes: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to cause the at least one processor to:

acquiring an interface use image acquired by an inspection robot at a specified inspection task point, and acquiring a preset specified template image corresponding to the specified inspection task point, wherein the image size of the interface use image is the same as that of the specified template image, and the specified template image is an image under the normal use condition of network equipment at the specified inspection task point; inputting the interface use image and the appointed template image into a pre-trained interface use detection model, and outputting a detection annotation file with the same size as the interface use image and the appointed template image, wherein the detection annotation file comprises the category identification of each pixel point; the method comprises the steps of obtaining category identification of each pixel point in the detection labeling file, determining the interface use state of the network equipment at the appointed routing inspection task point according to the category identification of each pixel point, and monitoring the use of the data center network equipment based on the interface use state, wherein the interface use state comprises an abnormal use state and a normal use state.

Embodiments of the present description also provide a non-volatile computer storage medium storing computer-executable instructions configured to:

acquiring an interface use image acquired by an inspection robot at a specified inspection task point, and acquiring a preset specified template image corresponding to the specified inspection task point, wherein the image size of the interface use image is the same as that of the specified template image, and the specified template image is an image under the normal use condition of network equipment at the specified inspection task point; inputting the interface use image and the specified template image into a pre-trained interface use detection model, and outputting a detection annotation file with the same size as the interface use image and the specified template image, wherein the detection annotation file comprises a category identifier of each pixel point; the method comprises the steps of obtaining the category identification of each pixel point in the detection marking file, determining the interface use state of the network equipment at the appointed routing inspection task point according to the category identification of each pixel point, and monitoring the use of the data center network equipment based on the interface use state, wherein the interface use state comprises an abnormal use state and a normal use state.

All the embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the embodiments of the apparatus, the device, and the nonvolatile computer storage medium, since they are substantially similar to the embodiments of the method, the description is simple, and for the relevant points, reference may be made to the partial description of the embodiments of the method.

The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

The devices and the media provided in the embodiments of the present description correspond to the methods one to one, and therefore, the devices and the media also have beneficial technical effects similar to the corresponding methods.

As will be appreciated by one skilled in the art, embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The description has been presented with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the description. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising a … …" does not exclude the presence of another identical element in a process, method, article, or apparatus that comprises the element.

The above description is merely one or more embodiments of the present disclosure and is not intended to limit the present disclosure. Various modifications and alterations to one or more embodiments of the present description will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement or the like made within the spirit and principle of one or more embodiments of the present specification should be included in the scope of the claims of the present specification.

Claims

1. A usage monitoring method of data center network equipment is applied to an inspection robot, and comprises the following steps:

acquiring an interface use image acquired by an inspection robot at a specified inspection task point, and acquiring a preset specified template image corresponding to the specified inspection task point, wherein the image size of the interface use image is the same as that of the specified template image, and the specified template image is an image of the network equipment at the specified inspection task point under the normal use condition;

inputting the interface use image and the designated template image into a pre-trained interface use detection model, and outputting a detection annotation file with the same size as the interface use image and the designated template image, wherein the detection annotation file comprises a category identifier of each pixel point;

the method comprises the steps of obtaining category identification of each pixel point in the detection labeling file, determining interface use state of network equipment at the appointed routing inspection task point according to the category identification of each pixel point, and monitoring use of the network equipment of the data center based on the interface use state, wherein the interface use state comprises an abnormal use state and a normal use state.

2. The usage monitoring method of a data center network device according to claim 1, wherein before the interface usage image and the specified template image are input into a pre-trained interface usage detection model, the method further comprises:

constructing an initial model using a neural network structure that fuses multi-scale features, wherein the initial model comprises a feature extraction network using convolution and pooling operations and an upsampling network using upsampling and feature map fusion.

3. The usage monitoring method of a data center network device according to claim 2, wherein before the interface usage image and the specified template image are input into a pre-trained interface usage detection model, the method further comprises:

generating a simulation data set, specifically comprising:

acquiring an interface element image corresponding to an interface, wherein the interface element image comprises a plurality of designated interface images, and each designated interface image comprises an interface idle state image or an interface occupied state image;

acquiring a plurality of template images, wherein the template images are images of a plurality of network devices under normal use;

covering the interface element images in each template image according to a random covering rule to generate an initial interface abnormal use image corresponding to each template image;

generating a designated annotation file with the same size as the initial interface abnormal use image according to the initial interface abnormal use image, wherein the designated annotation file comprises a violation interface identifier and a normal interface identifier;

carrying out color space transformation and geometric space transformation on the abnormal use image corresponding to each template image to generate an interface abnormal use image corresponding to each template image;

forming an image pair by the interface abnormal use image corresponding to each template image and the template image corresponding to the interface abnormal use image;

and generating a simulation data set according to the image pair and the specified annotation file.

4. The usage monitoring method of a data center network device of claim 3, wherein after generating the simulation data set, the method further comprises:

training the initial model by using the data in the simulation data set to obtain an interface use detection model meeting the requirements, specifically comprising:

inputting the image pairs in the simulation data set into the initial model, and outputting a presumed annotation file;

and adjusting parameters of the initial model according to the presumed annotation file and the appointed annotation file in the simulation data set to obtain an interface use detection model meeting the requirements.

5. The method according to claim 1, wherein the determining the interface usage status of the network device at the designated inspection task point according to the category identifier of each pixel point specifically includes:

judging whether the class identification of each pixel point in the detection label file has a first class identification, and if the first class identification exists, judging that the interface use state at the appointed routing inspection task point is abnormal use;

the class identification comprises a first class identification and a second class identification, the first class identification is used for indicating that the interface state corresponding to the pixel point is an abnormal use state, and the second class identification is used for indicating that the interface state corresponding to the pixel point is a normal use state;

according to the category identification, dividing the pixel points in the detection labeling file into a first category pixel point and a second category pixel point, wherein the category identification of each pixel point in the first category pixel point is a first category identification, and the category identification of each pixel point in the second category pixel point is a second category identification;

taking an area formed by the first type of pixel points as an interface abnormal use area, and determining area position coordinate data of the interface abnormal use area;

and according to the area position coordinate data, visually displaying the interface abnormal use area in the interface use image and the specified template image.

6. The method for monitoring the use of the data center network equipment according to claim 1, wherein before acquiring the preset specified template image corresponding to the specified inspection task point, the method further comprises:

presetting a plurality of inspection task points;

at the inspection task point, acquiring a template image through an inspection robot, and recording the robot posture state of the inspection robot;

establishing a corresponding relation between the template image and the pose state of the robot as a first corresponding relation;

and setting a place identifier for each inspection task point, and establishing a corresponding relation between the inspection task point and the template image as a second corresponding relation according to the place identifier of each inspection task point and the template image acquired by each inspection task.

7. The method for monitoring the use of the data center network equipment according to claim 6, wherein the step of acquiring the interface use image collected by the inspection robot at the designated inspection task point specifically comprises the following steps:

determining a designated place identification corresponding to the designated inspection task point;

according to the appointed place identification, an appointed template image corresponding to the appointed routing inspection task point is determined in the second corresponding relation;

determining the position and posture state of the designated robot corresponding to the designated template image in the first corresponding relation based on the designated template image corresponding to the designated inspection task point;

and using the posture state of the appointed robot, and acquiring an interface use image at the appointed routing inspection task point.

8. The usage monitoring method of a data center network device according to claim 1, wherein before the interface usage image and the specified template image are input into a pre-trained interface usage detection model, the method further comprises:

and according to a preset image alignment mode, performing alignment transformation on the interface use image and the specified template image to obtain an interface use image after the alignment transformation, so that the interface use image after the alignment transformation and the specified template image are input into a pre-trained interface use detection model.

9. A usage monitoring device for a data center network device, the device comprising:

at least one processor; and the number of the first and second groups,

acquiring an interface use image acquired by an inspection robot at a specified inspection task point, and acquiring a preset specified template image corresponding to the specified inspection task point, wherein the image size of the interface use image is the same as that of the specified template image, and the specified template image is an image of the specified inspection task point under the condition that network equipment is normally used;

inputting the interface use image and the designated template image into a pre-trained interface use detection model, and outputting a detection annotation file with the same size as the interface use image and the designated template image, wherein the detection annotation file comprises the category identification of each pixel point;

and acquiring the category identification of each pixel point in the detection labeling file, and determining the interface use state of the network equipment at the appointed routing inspection task point according to the category identification of each pixel point so as to monitor the use of the data center network equipment based on the interface use state, wherein the interface use state comprises an abnormal use state and a normal use state.

10. A non-transitory computer storage medium storing computer-executable instructions configured to: