Remote operation control system of data center

Abstract

The invention discloses a remote operation control system for a data center, relates to the technical field of remote operation control, and is used for solving the problems that the performance state of equipment is evaluated through a static or single load index, and an alarm system judges whether the equipment is normal or not through a fixed threshold value and cannot accurately calculate and trigger an alarm mechanism according to basic data. By performing weighted screening on the temperature data abnormal count and the energy consumption ratio of the equipment, the system can more accurately identify key equipment which possibly affects the overall load balance, so that a more accurate equipment screening mechanism is provided, the CPU utilization rate and the memory utilization rate are considered at the same time, and weighted calculation is performed on the CPU utilization rate and the memory utilization rate, so that the overall load balance can be more accurately identified. And the load balance condition of the equipment can be evaluated more accurately. The importance of the equipment is automatically judged according to the comparison of the load balance scores, a reasonable weight is calculated, the change of the load of the equipment and the fluctuation of the whole environment can be dynamically adapted, and a dynamic threshold value is set according to the weighted load balance score to judge whether to trigger an alarm or not.

Description

Technical Field

[0001] The present invention relates to the technical field of remote operation control, and more specifically, to a remote operation control system for a data center. Background Art

[0002] With the development of information technology and cloud computing, as the core infrastructure supporting large-scale computing and data storage, the efficient, stable, and secure operation of a data center is crucial for modern enterprises and Internet services. Therefore, how to ensure the sustainability of the data center, reduce operation and maintenance costs, improve work efficiency, and enhance data security while maintaining uninterrupted operation has become the core task in designing and building a remote control system for the data center. The remote operation control system provides a management platform for the data center by integrating monitoring technologies, automated operation and maintenance, remote access control, and other technologies.

[0003] The existing technologies have the following deficiencies:

[0004] Traditional methods solely rely on a certain piece of data to judge the importance of equipment, and cannot comprehensively consider the equipment status. By only evaluating the performance status of equipment through static or single load indicators, the performance fluctuations of the equipment at different time periods or under different loads will be ignored. The weight of the equipment may be set rigidly, lacking a mechanism for dynamic adjustment, which is likely to lead to inaccurate sorting of equipment importance. Existing alarm systems usually judge whether equipment is abnormal based on fixed thresholds, which may result in false alarms or missed alarms, and cannot avoid the existing problems of false alarms or missed alarms.

[0005] In view of the above problems, the present invention proposes a solution. Summary of the Invention

[0006] To overcome the above-mentioned defects of the existing technologies, an embodiment of the present invention provides a remote operation control system for a data center to solve the problems raised in the above background art.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A remote operation control system for a data center includes: an experimental screening module, a comprehensive scoring module, a weight calculation module, and an alarm selection module, and the modules are connected by signals:

[0009] Experimental screening module: Screening important equipment by weighted calculation of the abnormal count of temperature data and the energy consumption ratio of each device;

[0010] Comprehensive scoring module: Calculating the CPU utilization rate and memory utilization rate, and performing weighted calculation to obtain the load balance score of each important device;

[0011] Weight calculation module: Determine the importance of devices and calculate weights based on the comparison of the load balancing scores of various important devices;

[0012] Alarm selection module: Obtain the overall load balancing coefficient by weighting according to the weights, and design a threshold to determine whether an alarm is needed.

[0013] In a preferred embodiment, the experimental screening module is used to screen important devices by weighting the abnormal count of temperature data and the proportion of device energy consumption of each device:

[0014] Design an experimental period of 24 hours. During the experimental period, the remote operation control system runs normally, and record the abnormal count of temperature data and the data of the proportion of device energy consumption of each device during the experimental period;

[0015] Collect the real-time temperature data of each device, design a temperature threshold, count once when the threshold is exceeded, count each device in real time, and record the abnormal count of temperature data of each device within 24 hours, denoted as T i ;

[0016] Use linear normalization to map the temperature data of each device to the interval [0,1]. The formula is: Where: T normalized-i is the value after normalization of the temperature data of each device; T i is the temperature data of each device, T max is the maximum value of the temperature data of each device; T min is the minimum value of the temperature data of each device;

[0017] Collect the energy consumption data of each device, calculate the proportion of energy consumption of each device using the ratio of the energy consumption data of each device to the sum of the energy consumption data of all devices, and record the data of the proportion of device energy consumption of each device within 24 hours, denoted as E;

[0018] Use the abnormal count of temperature data and the proportion of device energy consumption to design weighting, and calculate the device screening score. The formula is expressed as: Screening score i = w 1 × T normalized-i + w 2 × E i , where: w 1 and w 2 are the weights of the abnormal count of temperature and the proportion of energy consumption;

[0019] Use the screening score for sorting, select important devices according to the sorting order, and select the top four devices in the screening score ranking.

[0020] In a preferred embodiment, the comprehensive scoring module is used to calculate the CPU utilization rate and the memory utilization rate, and perform weighting to obtain the load balancing scores of each important device:

[0021] Record the user time, system time, I / O wait time, and idle time of the CPU to calculate the CPU utilization rate, and use the user time, system time, I / O wait time, and idle time to calculate the CPU utilization rate. The calculation formula is expressed as: Where: The user time is the time consumed by the CPU to execute user code, excluding the execution of system-level operations and kernel code. The system time is the time consumed by the CPU to execute kernel code, mainly used for system-level operations. The I / O wait time is the time consumed by the CPU waiting for the completion of I / O operations. The idle time is the time when the CPU is completely idle and no tasks are being executed. The total time is equal to the sum of the user time, system time, I / O wait time, and idle time, and the total time = user time + system time + I / O wait time + idle time;

[0022] Obtain the total memory and the used memory of each device, and calculate the memory utilization rate. The memory utilization rate calculation formula is expressed as:

[0023] The load balancing score is obtained by weighting the CPU utilization rate and the memory utilization rate. The formula is expressed as: Load balancing score = α × CPU utilization rate + β × memory utilization rate, where: α and β are weighting coefficients, and α + β = 1. The CPU utilization rate: calculated by the aforementioned formula, and the memory utilization rate: calculated by the aforementioned formula.

[0024] In a preferred embodiment, the weight calculation module determines the importance of the device and calculates the weight based on comparing the load balancing scores of each important device:

[0025] After calculating the load balancing scores of each device, compare the magnitudes of the load balancing scores of each device, and assign weights to each device according to the load balancing scores by the preference ranking method to determine the weight values of the load balancing scores of each device.

[0026] In a preferred embodiment, the alarm selection module obtains the overall load balancing coefficient by weighting according to the weights, and designs a threshold to determine whether an alarm is required:

[0027] Perform weighted average calculation on the battery status of each functional part according to the weight values of the battery status distribution coefficients of each functional part to obtain the overall battery status. The formula is expressed as: OLBS = (score A × weight A ) + (score B × weight B ) + (scoreC × weight C + (rating D × weight D ), where: rating A 、 rating B 、 rating C 、 rating D are the load balance ratings of devices A, B, C, and D respectively, and weight A 、 weight B 、 weight C 、 weight D are the corresponding weight values of devices A, B, C, and D respectively. Among them: Devices A, B, C, and D correspond one-to-one to the four devices in this embodiment. After sorting each device according to the size of the load balance rating, they are respectively corresponding to devices A, B, C, and D in descending order;

[0028] And design a threshold for the overall load balance coefficient to determine whether to alarm. If OLBS is greater than the set alarm threshold, the system will trigger an alarm. If OLBS is less than the set alarm threshold, the system will not trigger an alarm.

[0029] The technical effects and advantages of a remote operation control system for a data center according to the present invention:

[0030] By performing weighted screening on the abnormal count of temperature data and energy consumption ratio of devices, the system can more accurately identify key devices that may affect the overall load balance, thereby providing a more accurate device screening mechanism. Considering the CPU utilization rate and memory utilization rate at the same time and performing weighted calculations on them can more precisely evaluate the load balance situation of devices. Automatically judging the importance of devices based on the comparison of load balance ratings and calculating reasonable weights can dynamically adapt to changes in device loads and fluctuations in the overall environment, and set a dynamic threshold based on the weighted load balance rating to determine whether to trigger an alarm, avoiding false alarms or missed alarms existing in traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 is a schematic structural diagram of a remote operation control system for a data center according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0032] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0033] Embodiment 1

[0034] The present invention discloses a remote operation control system for a data center, comprising: an experimental screening module, a comprehensive scoring module, a weight calculation module, and an alarm selection module, which are signal-connected to each other.

[0035] Experimental screening module: Screen important devices by weighting the abnormal count of temperature data of each device and the proportion of device energy consumption.

[0036] Comprehensive scoring module: Calculate the CPU utilization rate and memory utilization rate, and perform weighting to obtain the load balance score of each important device.

[0037] Weight calculation module: Judge the importance of devices based on the comparison of the load balance scores of each important device and calculate the weights.

[0038] Alarm selection module: Obtain the overall load balance coefficient by weighting according to the weights, and design a threshold to judge whether an alarm is needed.

[0039] The functions of each module are as follows:

[0040] In the experimental screening module, important devices are screened by weighting the abnormal count of temperature data of each device and the proportion of device energy consumption. The specific content includes:

[0041] Design an experimental period with a duration of 24 hours. During the experimental period, the remote operation control system runs normally. There are multiple devices in the remote operation control system, such as servers, remote controllers, storage devices, power management devices, acquisition devices, and network devices, etc. Record the abnormal count of temperature data and the proportion of device energy consumption data of each device during the experimental period.

[0042] Collect the real-time temperature data of each device, design a temperature threshold, and count once when the threshold is exceeded. Perform real-time counting on each device. Temperature anomaly is usually a warning signal of device failure. High-frequency temperature anomalies mean that the device may have a more serious failure risk. Therefore, use temperature as the judgment criterion for whether the device is abnormal, and record the abnormal count of temperature data of each device within 24 hours, denoted as T i 。

[0043] Use linear normalization to map the temperature data of each device to the interval [0,1]. The formula is: Where: T normalized-i is the normalized value of the temperature data of each device; T i is the temperature data of each device, T max is the maximum value of the temperature data of each device; T min is the minimum value of the temperature data of each device.

[0044] Collect the energy consumption data of each device, calculate the energy consumption percentage of each device by using the ratio of the energy consumption data of each device to the sum of the energy consumption data of all devices. Devices with a high energy consumption percentage usually play a more important resource role in the system and are more vulnerable to overload. Record the energy consumption percentage data of each device within 24 hours, denoted as E i 。

[0045] Design weights using the temperature data anomaly count and the energy consumption percentage of the device, and calculate the device screening score. The formula is expressed as: Screening score i =w 1 ×T normalized-i +w 2 ×E i ,where: w 1 and w 2 are the weights of the temperature anomaly count and the energy consumption percentage, which are usually adjusted according to the experimental objectives. The weight settings can be adjusted by those skilled in the art according to the actual situation. For example, if the fault warning relies more on temperature anomalies, set a larger weight w 1 ; otherwise, set a larger weight w 2 。

[0046] Sort using the screening score, select important devices according to the sorting order, and take the top four devices in the screening score sorting. The method of screening important devices can also be readjusted by professionals in the field, which will not be elaborated here.

[0047] In the comprehensive scoring module, calculate the CPU utilization rate and the memory utilization rate, and perform weighting to obtain the load balancing score of each important device. The specific content includes:

[0048] Record the user time, system time, I / O wait time, and idle time of the CPU, and calculate the CPU utilization rate using the user time, system time, I / O wait time, and idle time. The definitions of each CPU time type are as follows:

[0049] User time: The time consumed by the CPU to execute user code, excluding system-level operations and the execution of kernel code. For example, if a computationally intensive program is running, the CPU time consumed by the execution of the program will be counted as user time.

[0050] System time: The time consumed by the CPU to execute kernel code (such as system calls of the operating system), mainly used for system-level operations and has no direct relationship with user processes. For example, when a process requests an I / O operation, the operating system is responsible for scheduling and managing the I / O device, and the relevant CPU time is the system time.

[0051] I / O Wait Time: The time consumed by the CPU while waiting for I / O operations to complete, which is the time consumed by a process while waiting for responses from the hard disk, network, or other peripherals. For example, a program may need to read data from the disk. While waiting for the disk to read, the CPU is not performing actual computing work, and this period of time is counted as I / O wait time.

[0052] Idle Time: The time when the CPU is completely idle, i.e., the time when no tasks are being executed. This time indicates that the CPU is not occupied and is usually used to display the idle state of the system.

[0053] Total Time: The total running time of the system within a monitoring period, which is equal to the sum of user time, system time, I / O wait time, and idle time.

[0054] The calculation formula is expressed as: Where: Total Time = User Time + System Time + I / O Wait Time + Idle Time.

[0055] For example: User Time = 200 seconds, System Time = 50 seconds, I / O Wait Time = 30 seconds, Idle Time = 20 seconds, then Total Time = 200 + 50 + 30 + 20 = 300 seconds

[0056] Calculate CPU Utilization:

[0057]

[0058] Obtain the total memory and used memory of each device, and calculate the memory utilization rate. The total memory of a device (usually measured in bytes, kilobytes, megabytes, or gigabytes) is the size of the system's physical memory. The used memory refers to the amount of memory currently being used in the system. It can be obtained by viewing the operating system or device management tools. Generally, the used memory includes the memory occupied by running programs, caches, files, and applications. The calculation formula for memory utilization rate is expressed as:

[0059] The load balancing score is obtained by weighting the CPU utilization rate and the memory utilization rate. The formula is expressed as: Load Balancing Score = α × CPU Utilization Rate + β × Memory Utilization Rate, where: α and β are weighting coefficients. Generally, the weighting coefficients are related to the priority of the task, and α + β = 1. CPU Utilization Rate: Calculated by the aforementioned formula. Memory Utilization Rate: Calculated by the aforementioned formula.

[0060] In the alarm selection module, the importance of devices is judged and weights are calculated based on comparing the load balancing scores of each important device. The specific content includes:

[0061] After calculating the load balance scores of each device, compare the magnitudes of the load balance scores of each device. The larger the load balance score, the more important the device is proven to be. Assign weights to the load balance scores of each device respectively according to the preference ranking method to determine the load balance score weight values of each device.

[0062] Specifically, the weight assignment for the temperature scores of each monitoring device according to the preference ranking method is shown in Table 1 below:

[0063]

[0064] Table 1

[0065] It should be noted that in Table 1, devices A, B, C, and D correspond to the four devices in this embodiment one by one. After sorting each device according to the magnitude of the load balance score, they correspond to devices A, B, C, and D in descending order.

[0066] It should be noted that the preference ranking method is a network diagram method used in project management, especially in engineering project management. It is used to represent the dependency relationships and the sequence of activities among various activities in a project. Through a graphical way, it helps the project manager to conduct effective time management and resource allocation. The TTL indicator, as a weighted factor, calculates and evaluates the overall load balance coefficient reasonably by determining the relative importance of different devices in the remote operation control system, adding up the scores after comparing the relative importance of each device and assigning scores respectively. When calculating the weights of each device, the CPU utilization rate and memory utilization rate of each are considered, which in turn affects the final load balance coefficient.

[0067] In the alarm selection module, based on the weight to obtain the overall load balance coefficient, design a threshold to determine whether an alarm is needed. The specific content includes:

[0068] Perform a weighted average calculation on the load balance scores of each device according to the weight values of each device to obtain the overall load balance score. The formula is expressed as: OLBS = (score A × weight A ) + (score B × weight B ) + (score C × weight C ) + (score D × weight D ) where: score A , score B , score C , score D are the load balance scores of devices A, B, C, and D respectively, and weight a , weight B , weight C, Weight D They are the weight values corresponding to devices A, B, C, and D respectively.

[0069] In the present invention, by first evaluating the load balance scores of each device to determine the relative weights of the load balance scores of each device, and performing weighted averaging based on the relative weights, the overall load balance coefficient calculated in this way is more representative and can better grasp the overall load balance state of the remote operation control system.

[0070] And a threshold value is designed for the overall load balance coefficient to determine whether an alarm is needed. If OLBS is greater than the set alarm threshold, the system will trigger an alarm; if OLBS is less than the set alarm threshold, the system will not trigger an alarm.

[0071] It should be noted that the threshold value of the overall load balance coefficient can be set by those skilled in the art according to the actual situation, or the overall load balance coefficient can be calculated and evaluated based on historical data, which will not be elaborated here.

[0072] The above formulas are all dimensionless and take their numerical values for calculation. The formula is a formula obtained by collecting a large amount of data for software simulation to approximate the real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0073] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0074] Those of ordinary skill in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application of the technical solution and the invention constraints. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0075] In addition, in each embodiment of the present application, the functional modules can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0076] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claimed rights.

[0077] Finally, the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A data center remote operation control system, characterized in that: include: Experiment screening module, comprehensive scoring module, weight calculation module and alarm selection module, and signal connection between each module; Experimental screening module: Screen important equipment by counting abnormal temperature data of each device and weighting the proportion of equipment energy consumption; Comprehensive scoring module: calculates CPU utilization and memory utilization, and performs weighted calculations to obtain the load balancing score of each important device; Weight calculation module: Determine the importance of the device and calculate the weight based on the load balancing score of each important device; Alarm selection module: obtain the overall load balancing coefficient based on weights, and design the threshold to determine whether an alarm is needed.

2. A data center remote operation control system according to claim 1, characterized in that: Design a 24-hour experimental period, operate the remote control system normally during the experimental period, and record the abnormal temperature data count and equipment energy consumption ratio data of each device during the experimental period; Collect the real-time temperature data of each device, design a temperature threshold, count once when the threshold is exceeded, count each device in real time, and record the abnormal temperature data count of each device within 24 hours, recorded as T i ; Linear normalization is used to map the temperature data of each device to the interval [0,1], the formula is: Where: T normalized-i is the normalized value of the temperature data of each device; T i is the temperature data of each device, T max is the maximum value of the temperature data of each device; T min It is the minimum value of temperature data of each device; Collect the energy consumption data of each device, calculate the energy consumption proportion of each device by using the ratio of the energy consumption data of each device to the sum of the energy consumption data of all devices, and record the energy consumption proportion data of each device within 24 hours, recorded as E i ; The temperature data anomaly count and equipment energy consumption ratio are used to design weights and calculate the equipment screening score. The formula is: Screening score i =w1×T normalized-i +w2×E i , where: w1 and w2 are the weights of temperature anomaly count and energy consumption proportion; Use the screening scores to sort, select important devices according to the sorting order, and take the top four devices in the screening score ranking.

3. A data center remote operation control system according to claim 2, characterized in that: The CPU user time, system time, I / O waiting time, and idle time are recorded to calculate the CPU utilization, and the CPU utilization is calculated using the user time, system time, I / O waiting time, and idle time. The calculation formula is as follows: Among them: user time is the time consumed by the CPU to execute user code, excluding system-level operations and kernel code execution; system time is the time consumed by the CPU to execute kernel code, mainly used for system-level operations; I / O wait time is the time consumed by the CPU waiting for I / O operations to complete; idle time is the time when the CPU is completely idle and does not execute any tasks; total time is equal to the sum of user time, system time, I / O wait time and idle time, total time = user time + system time + I / O wait time + idle time; Get the total memory and used memory of each device and calculate the memory utilization. The memory utilization calculation formula is as follows: The load balancing score is obtained by weighting the CPU utilization and the memory utilization, and the formula is: Load balancing score = α × CPU utilization + β × memory utilization, where: α and β are weighting coefficients, and α + β = 1, CPU utilization: calculated by the above formula, memory utilization: calculated by the above formula.

4. A data center remote operation control system according to claim 3, It is characterized by: After calculating and obtaining the load balancing score of each device, the load balancing scores of each device are compared, and weights are assigned according to the load balancing scores of each device using the priority graph method to determine the weight value of the load balancing score of each device.

5. A data center remote operation control system according to claim 4, characterized in that: The battery status of each functional part is calculated by weighted average according to the weight value of the battery status distribution coefficient of each functional part to obtain the overall battery status. The formula is: OLBS = (score A ×Weight A )+(Rating B ×Weight B )+(Rating C ×Weight C +(Rating D ×Weight D ), where: Rating A ,score B ,score C ,score D The load balancing scores and weights of devices A, B, C, and D are A , weight B , weight C , weight D are weight values ​​corresponding to devices A, B, C, and D, respectively, wherein devices A, B, C, and D correspond to the four devices in this embodiment one by one, and after sorting the devices according to the size of the load balancing score, they correspond to devices A, B, C, and D in descending order; A threshold is designed for the overall load balancing factor to determine whether an alarm is needed. If OLBS is greater than the set alarm threshold, the system will trigger an alarm. If OLBS is less than the set alarm threshold, the system will not trigger an alarm.

Images