High-density server testing methods, devices, electronic equipment and storage media

By identifying target nodes in high-density servers and frequently switching between power-on and power-off states, combined with numerical calculations and deep learning, the testing of high-density servers was optimized, solving the problem of electromagnetic interference and achieving more comprehensive testing and higher accuracy.

CN119248589BActive Publication Date: 2026-11-17INSPUR SUZHOU INTELLIGENT TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411365032.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-27
Publication Date
2026-11-17
Estimated Expiration
2044-09-27

AI Technical Summary

Technical Problem

Existing high-density server testing methods fail to fully consider electromagnetic interference between electronic devices, resulting in poor test comprehensiveness and accuracy.

Method used

By obtaining the overall power consumption limit of the high-density server under test, the target number of server nodes is determined, and these nodes are controlled to switch between power-on and power-off states at a preset frequency in each round of testing to obtain test data. Combined with numerical calculation and deep learning technology, the testing process is optimized to reduce electromagnetic interference.

Benefits of technology

It improves the comprehensiveness and accuracy of high-density server testing, reduces the risk of damage, and enhances testing efficiency and automation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119248589B_ABST
    Figure CN119248589B_ABST
Patent Text Reader

Abstract

This invention provides a high-density server testing method, apparatus, electronic device, and storage medium. The method includes: obtaining a target number based on the overall power consumption limit of the high-density server under test; determining each target number of server nodes in the high-density server under test as the target server nodes for each round of testing; and determining the server nodes in the high-density server under test other than the target server nodes for each round of testing as the test server nodes for each round of testing; in each round of testing, controlling the test server nodes for each round of testing to switch between a power-on state and a power-off state at a preset frequency within a first time period of a first preset duration, and acquiring the test data within the first time period as the test result for each round of testing. This invention provides a high-density server testing method, apparatus, electronic device, and storage medium, which can improve the comprehensiveness and accuracy of high-density server testing.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a high-density server testing method, apparatus, electronic device, and storage medium. Background Technology

[0002] High-density servers, designed for cloud computing, data centers, and internet applications, integrate more processors and I / O expansion capabilities within a smaller physical space. This significantly reduces space costs and improves computing performance, better meeting user needs.

[0003] Compared to regular servers, high-density servers have a larger internal space density and significantly higher power consumption, requiring a significantly higher power supply. Therefore, testing the reliability and stability of high-density servers is crucial for ensuring their normal operation.

[0004] However, traditional high-density server testing methods in related technologies are usually static stress tests, which do not take into account electromagnetic interference between electronic devices in high-density servers. This results in poor comprehensiveness and accuracy in testing high-density servers based on traditional methods. Therefore, how to conduct more comprehensive and accurate testing of high-density servers is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] This invention provides a high-density server testing method, apparatus, electronic device, and storage medium to address the shortcomings of existing high-density server testing methods that lack comprehensiveness, and to achieve a more comprehensive test of high-density servers.

[0006] This invention provides a high-density server testing method, comprising the following steps.

[0007] Obtain the overall power consumption limit of the high-density server under test, and obtain the target quantity based on the overall power consumption limit; Each of the target number of server nodes in the high-density server under test is determined as the target server node for each round of testing, and the server nodes in the high-density server under test other than the target server nodes for each round of testing are determined as the test server nodes for each round of testing. In each round of testing, within a first time period of a first preset duration, the test server node corresponding to each round of testing is controlled to switch between power-on and power-off states at a preset frequency, and the test data within the first time period is obtained as the test result of each round of testing. The test data includes log information generated by the high-density server under test and / or log information generated by the testing tool. Based on the test results of each round of testing, the test results of the high-density server under test are obtained.

[0008] According to a high-density server testing method provided by the present invention, in each round of testing, before controlling the test server node corresponding to each round of testing to switch between power-on and power-off states at a preset frequency within a first time period of a first preset duration, and acquiring the test data within the first time period as the test result of each round of testing, the method further includes: In each round of testing, stress tests are performed on the target test nodes corresponding to each round of testing within a second time period of a second preset duration, and the test data within the second time period is obtained as the test result of each round of testing. The end time of the second time period is the start time of the first time period.

[0009] According to a high-density server testing method provided by the present invention, the test data includes: the power consumption value of the target server node corresponding to each round of testing and the total power consumption value of the high-density server under test; The process of obtaining the test results of the high-density server under test based on the test results of each round of testing includes: Based on the test results of each round of testing, a recommended server node activation scheme is obtained for the high-density server under test, which is used as the test result of the high-density server under test. The recommended server node activation scheme includes the server nodes recommended to be activated in the high-density server under test.

[0010] According to a high-density server testing method provided by the present invention, the step of obtaining the recommended activation scheme for the server node corresponding to the high-density server under test based on the test results of each round of testing includes: Based on the total power consumption value of the high-density server under test and the power consumption value of the target server node corresponding to each round of testing obtained in each round of testing, the evaluation index value corresponding to each round of testing is obtained. Compare the evaluation index values ​​corresponding to each round of the test, and determine the round of the test with the smallest corresponding evaluation index value as the target round of the test; The target server node corresponding to the target round test is determined to be the server node recommended to be enabled in the recommended scheme for the server node corresponding to the high-density server under test.

[0011] According to a high-density server testing method provided by the present invention, the step of determining each target number of server nodes in the high-density server under test as the target server node corresponding to each round of testing, and determining the server nodes in the high-density server under test other than the target server nodes corresponding to each round of testing as the test server nodes corresponding to each round of testing, includes: Select the target number of server nodes among the server nodes in the high-density server to be tested as the target server nodes for the first round of testing, and determine the server nodes in the high-density server to be tested other than the target server nodes for the first round of testing as the test servers for the first round of testing. For the i-th round of testing, if the target number is less than the total number of server nodes in the high-density server under test, a combination of server nodes with the target number is selected again from among the server nodes in the high-density server under test. If it is determined that the combination of server nodes with the target number selected again is different from the combination of target server nodes corresponding to each round of testing before the i-th round, the combination of server nodes with the target number selected again is determined as the i-th round of testing. i The target server node corresponding to the round of testing, except for the first high-density server under test. i Server nodes other than the target server node corresponding to the round of testing are determined to be the first... i For the test server node corresponding to the next test, i is increased by 1, and the process returns to the step of selecting the target number of server nodes from each server node in the high-density server to be tested again, until it is determined that there is no combination of the target number of server nodes in the high-density server to be tested that is different from the combination of the target server nodes corresponding to each test before the i-th round of testing.

[0012] According to a high-density server testing method provided by the present invention, obtaining the target quantity based on the overall power consumption limit value includes: With all server nodes in the high-density server under test in a powered-off state and all fans in the high-density server under test at their maximum speed, the total power consumption of the high-density server under test is obtained as the first total power consumption value. When any server node in the high-density server under test is running at full load, all other server nodes in the high-density server under test are shut down, and the speed of each fan in the high-density server under test is at its maximum, the total power consumption value of the high-density server under test is obtained as the second total power consumption value. The target quantity is obtained based on the first total power consumption value, the second total power consumption value, and the total power consumption limit value.

[0013] According to a high-density server testing method provided by the present invention, obtaining the target quantity based on the first overall power consumption value, the second overall power consumption value, and the overall power consumption limit value includes: Calculate the difference between the second total power consumption value and the first total power consumption value as the first power consumption value; calculate the sum of the second total power consumption value and the first power consumption value as the second power consumption value; calculate the sum of the second total power consumption value and twice the first power consumption value as the third power consumption value; calculate the sum of the second total power consumption value and three times the first power consumption value as the fourth power consumption value; calculate the sum of the second total power consumption value and four times the first power consumption value as the fifth power consumption value. Based on the relationship between the first total power consumption value and at least two of the second total power consumption value, the third power consumption value, the fourth power consumption value, and the fifth power consumption value, the number of server nodes that can be enabled in the high-density server under test is obtained as the target number.

[0014] The present invention also provides a high-density server testing device, comprising the following modules: The data acquisition module is used to acquire the overall power consumption limit value of the high-density server under test, and to acquire the target quantity based on the overall power consumption limit value; The node determination module is used to determine each of the target number of server nodes in the high-density server under test as the target server node corresponding to each round of testing, and to determine the server nodes in the high-density server under test other than the target server node corresponding to each round of testing as the test server node corresponding to each round of testing. The node testing module is used to control the test server node corresponding to each round of testing to switch between power-on and power-off states at a preset frequency within a first time period of a first preset duration in each round of testing, and to acquire the test data within the first time period as the test result of each round of testing. The test data includes log information generated by the high-density server under test and / or log information generated by the testing tool. The result output module is used to obtain the test results of the high-density server under test based on the test results of each round of testing.

[0015] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the high-density server testing method described above.

[0016] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the high-density server testing method as described above.

[0017] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the high-density server testing method as described above.

[0018] The high-density server testing method, apparatus, electronic device, and storage medium provided by this invention obtain the number of usable server nodes in the high-density server under test based on the overall power consumption limit of the server under test, and then determine each target number of server nodes in the high-density server under test as the target server nodes for each round of testing. Other server nodes in the high-density server under test besides the target server nodes for each round of testing are determined as test server nodes for each round of testing. Then, in each round of testing, stress testing is performed on the target test nodes for each round of testing within a second time period of a second preset duration, and the test data within the second time period is obtained as the test result for each round of testing. Finally, within a first time period of a first preset duration, the test server nodes for each round of testing are controlled to switch between power-on and power-off states at a preset frequency, and the first time period is obtained. The test data within the segment also serves as the test result for each round of testing. Based on the test results of each round, the test results of the high-density server under test are obtained. While avoiding the risk of damage to the high-density server under test, by controlling the test server node corresponding to each round of testing to switch between power-on and power-off states at a preset frequency, the overall current of the high-density server under test fluctuates frequently. This makes the electromagnetic interference between the server nodes and different electronic devices in the high-density server under test more severe, thus better testing the reliability and stability of the high-density server under test under high-intensity, interactive electromagnetic interference. It can test the operation of the high-density server nodes under high-intensity, interactive electromagnetic interference, improve the comprehensiveness and accuracy of high-density server testing, and the test method is simple, highly automated, and improves the testing efficiency of high-density servers. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a structural diagram of a high-density server in related technologies.

[0021] Figure 2 This is one of the flowcharts of the high-density server testing method provided by the present invention.

[0022] Figure 3 This is the second flowchart of the high-density server testing method provided by the present invention.

[0023] Figure 4This is a schematic diagram of the structure of the high-density server testing device provided by the present invention.

[0024] Figure 5 This is a schematic diagram showing the connection between the high-density server testing device provided by the present invention and the high-density server under test.

[0025] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0027] In the description of the invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0028] In the description of this application, the terms "first," "second," etc., are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class, without limiting the number of objects; for example, a first object can be one or more. Furthermore, in the description of this application, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects have an "or" relationship.

[0029] It should be noted that a server is a type of computer. Compared to a regular computer, a server runs faster, can handle higher loads, but is more expensive.

[0030] Servers can generally be categorized into four types: tower servers, rack servers, blade servers, and high-density servers. High-density servers, designed for cloud computing, data centers, and internet applications, integrate more processors and I / O expansion capabilities within a smaller physical space. This significantly reduces customer space costs and substantially improves computing performance, while also offering flexible scalability to meet specific industry needs.

[0031] Unlike ordinary servers that use independent power supplies and fans, high-density servers share power supplies and fans among multiple server nodes in the same chassis, which greatly improves the efficiency of power supply and heat dissipation systems and makes the overall weight of high-density servers lighter and the manufacturing cost lower.

[0032] However, with the introduction of new-generation Central Processing Units (CPUs) and DDR5 memory on next-generation server platforms, the number of cores integrated on a single CPU is increasing, and the power consumption of the CPU is also rising. For high-density servers, the significant increase in CPU power consumption poses a severe challenge to their power supply capabilities.

[0033] Figure 1 This is a structural diagram of a high-density server in related technologies. Taking a high-density server with four server nodes as an example... Figure 1 As shown, the high-density server also includes components such as a management board, a system backplane, and a fan backplane. Each server node and the management board, system backplane, and fan backplane are connected to the system backplane.

[0034] If one of the server nodes in the aforementioned high-density server is equipped with two CPUs and nine DDR5 memory modules, with each CPU consuming approximately 350W and each DDR5 memory module consuming 12W, then the power consumption of one server node in the aforementioned high-density server would reach 796W (350W×2+12W×8), and the power consumption of four server nodes in the aforementioned high-density server would reach approximately 3200W.

[0035] Typically, high-density servers are equipped with high-performance fans on the fan backplane, with a power consumption of around 400W. The power consumption of the management board and system backplane is around 100W. The total power consumption of a high-density server including four server nodes is over 3700W.

[0036] High-density servers typically use a dual power supply mode; therefore, the power supply for these high-density servers needs to be at least 2000W.

[0037] In related technologies, the output voltage of power supplies is typically 12V. When the power supply reaches 2000W, the output current exceeds 300A. This high output current leads to increased electromagnetic interference in the power supply lines, affecting the normal operation and signal transmission of surrounding electronic equipment. This electromagnetic interference becomes particularly severe when the power supply's output current fluctuates frequently.

[0038] Furthermore, the heat density inside a high-density server chassis is higher, and the electromagnetic environment is more complex. The increase in temperature inside the chassis will further aggravate the electromagnetic interference of electronic equipment.

[0039] The interaction between different electronic devices in a high-density server can also generate uncertain and unpredictable electromagnetic interference, which can be transmitted to other electronic devices, creating a chain reaction.

[0040] Therefore, ensuring the normal operation of high-density servers within the relatively limited chassis space is a significant challenge. Testing the reliability and stability of high-density servers is crucial for ensuring their proper functioning.

[0041] Traditional high-density server testing methods in related technologies are usually static stress tests. For example, traditional high-density server testing methods can control each server node in the high-density server to be under full load, thereby making the entire high-density server under full load, and then obtaining various performance data of the high-density server under full load, and obtaining the test results of the high-density server based on the above performance data; another example is that traditional high-density server testing methods can also control a specified server node in the high-density server to run a specified program and then cut off the power, and then obtain the running status of other server nodes in the high-density server, and then obtain the test results of the high-density server.

[0042] However, traditional high-density server testing methods do not consider electromagnetic interference between electronic devices within the server, resulting in poor comprehensiveness and accuracy in testing high-density servers using these methods. Therefore, how to conduct more comprehensive and accurate testing of high-density servers is a pressing technical problem that needs to be solved in this field.

[0043] To address this issue, this invention provides a high-density server testing method. This method requires only one control unit and a single command to install testing tools and deploy test scripts, enabling testing of multiple server nodes within a high-density server. Due to the high power consumption of high-density servers, it is difficult to ensure all server nodes operate at full load when the power supply's output voltage is limited. Therefore, this invention provides an optimized distribution selection based on actual testing conditions, achieving energy conservation and emission reduction. Considering maintenance needs such as system expansion or node plugging / unplugging in real-world scenarios, this method performs random stress tests on server nodes and frequent power-on / off tests on other server nodes, causing frequent fluctuations in the overall current of the high-density server and exacerbating electromagnetic interference between server nodes. This allows for better testing of the high-density server's reliability under such interactive stress conditions. The high-density server testing method is simple to operate, highly automated, reduces manual workload, and significantly improves testing efficiency.

[0044] The following is combined Figures 2-3 This invention describes a high-density server testing method.

[0045] Figure 2 This is one of the flowcharts illustrating the high-density server testing method provided by the present invention, such as... Figure 2 As shown, the method includes the following steps: Step 201: Obtain the overall power consumption limit value of the high-density server to be tested, and obtain the target quantity based on the overall power consumption limit value.

[0046] It should be noted that the execution subject of this embodiment of the invention is a high-density server testing device. This high-density server testing device can be configured in electronic devices such as computers and servers.

[0047] Specifically, the high-density server under test is the test object of the high-density server testing method provided by this invention. Based on the high-density server testing method provided by this invention, the reliability and stability of the server under test can be tested more comprehensively and accurately.

[0048] It is understood that the high-density server to be tested in the embodiments of the present invention can be determined according to actual needs. The embodiments of the present invention do not impose specific limitations on the high-density server to be tested.

[0049] The power consumption limit of the high-density server under test is used to limit the maximum power consumption of the high-density server under test when it is working. If the power consumption of the high-density server under test exceeds the power consumption limit, the high-density server under test is at risk of being damaged.

[0050] The power consumption limit of the high-density server under test can be determined based on factors such as the server's factory configuration information, usage scenario, energy efficiency standards, and actual needs.

[0051] In this embodiment of the invention, the overall power consumption limit value of the high-density server under test can be obtained in various ways. For example, in this embodiment of the invention, the overall power consumption limit value of the high-density server under test can be obtained based on user input; or, in this embodiment of the invention, the overall power consumption limit value of the high-density server under test can also be received from other electronic devices.

[0052] It should be noted that the high-density server under test in this embodiment of the invention includes multiple server nodes.

[0053] It should be noted that although the high-density server under test includes multiple server nodes, if all server nodes are enabled, the total power consumption of the high-density server under test may exceed the power consumption limit, which may damage the high-density server under test.

[0054] Therefore, after obtaining the overall power consumption limit of the high-density server under test, the number of server nodes that can be enabled in the high-density server under test can be obtained through numerical calculation, mathematical statistics and deep learning technology, which can be used as the target number M, thereby reducing the risk of damage to the high-density server under test.

[0055] As an optional embodiment, obtaining the target quantity based on the overall power consumption limit includes: obtaining the overall power consumption value of the high-density server under test as the first overall power consumption value when all server nodes in the high-density server under test are in a powered-off state and the speed of each fan in the high-density server under test is at its maximum value.

[0056] Specifically, in this embodiment of the invention, control commands can be used to control all server nodes in the high-density server under test to be in a powered-off state and to control the speed of each fan in the high-density server under test to be at its maximum value.

[0057] If all server nodes in the high-density server under test are in a powered-off state and the speed of each fan in the high-density server under test is at its maximum, the total power consumption value of the high-density server under test can be obtained by data query and used as the first total power consumption value.

[0058] Under the condition that any server node in the high-density server under test is running at full load, all other server nodes in the high-density server under test are shut down, and the speed of each fan in the high-density server under test is at its maximum, the total power consumption value of the high-density server under test is obtained as the second total power consumption value.

[0059] Specifically, in this embodiment of the invention, control commands can be used to control any one of the server nodes in the high-density server under test to be in a full-load operating state, control all server nodes in the high-density server under test other than the aforementioned server nodes to be kept in a powered-off state, and control the speed of each fan in the high-density server under test to be kept at its maximum value.

[0060] If it is determined that any one of the server nodes in the high-density server under test is running at full load, all other server nodes in the high-density server under test are kept powered off, and the fan speeds of the high-density server under test are kept at their maximum values, the total power consumption of the high-density server under test can be obtained by data querying and used as the second total power consumption value.

[0061] It is understandable that the difference between the second total power consumption value and the first total power consumption value is the power consumption value of a server phase in the high-density server under test under full load operation.

[0062] The target quantity is obtained based on the first total power consumption value, the second total power consumption value, and the total power consumption limit value.

[0063] Specifically, after obtaining the first and second total power consumption values, the number of server nodes that can be enabled in the high-density server under test can be obtained based on the first and second total power consumption values ​​and the total power consumption limit value of the high-density server under test, through numerical calculation, mathematical statistics and deep learning technology, and used as the target number.

[0064] As an optional embodiment, obtaining the target quantity based on the first overall power consumption value, the second overall power consumption value, and the overall power consumption limit value includes: calculating the difference between the second overall power consumption value and the first overall power consumption value as the first power consumption value; calculating the sum of the second overall power consumption value and the first power consumption value as the second power consumption value; calculating the sum of twice the second overall power consumption value and the first power consumption value as the third power consumption value; calculating the sum of three times the second overall power consumption value and the first power consumption value as the fourth power consumption value; and calculating the sum of four times the second overall power consumption value and the first power consumption value as the fifth power consumption value.

[0065] Based on the relationship between the first overall power consumption value and at least two of the second, third, fourth, and fifth overall power consumption values, the number of server nodes that can be enabled in the high-density server under test is obtained as the target number.

[0066] Specifically, if the total power consumption limit of the high-density server under test is greater than the second total power consumption value but less than the second power consumption value, it can be determined that the number of server nodes that can be enabled in the high-density server under test is 1, that is, the target number is 1.

[0067] If the total power consumption limit of the high-density server under test is greater than the second power consumption limit but less than the third power consumption limit, it can be determined that the number of server nodes that can be enabled in the high-density server under test is 2, that is, the target number is 2.

[0068] If the total power consumption limit of the high-density server under test is greater than the third power consumption value but less than the fourth power consumption value, it can be determined that the number of server nodes that can be enabled in the high-density server under test is 3, that is, the target number is 3.

[0069] If the total power consumption limit of the high-density server under test is greater than the fourth power consumption value but less than the fifth power consumption value, it can be determined that the number of server nodes that can be enabled in the high-density server under test is 4, that is, the target number is 4.

[0070] The relationship between the target number and the power consumption limit, first power consumption value, and second power consumption value of the high-density server under test can be expressed by the following formula: in, Indicates the target quantity. This represents the first overall power consumption value. This indicates the second overall power consumption value; This indicates the overall power consumption limit of the high-density server under test; Indicates the first power consumption value. .

[0071] Accordingly, the second power consumption value = Third power consumption value = Fourth power consumption value = Fifth power consumption value = .

[0072] exist In the case of target quantity =1; in In the case of target quantity It is 2; in In the case of target quantity It is 3; in In the case of target quantity The value is 1.

[0073] This invention provides a first overall power consumption value for a high-density server under test, obtained when all server nodes are powered off and all fans are running at their maximum speed. A second overall power consumption value is obtained when any server node is running at full load, all other server nodes are powered off, and all fans are running at their maximum speed. Based on the first, second, and power consumption limits, and through numerical calculations and conditional judgments, the number of usable server nodes in the high-density server under test can be determined more accurately and efficiently as a target number. This improves the security of testing the high-density server and provides a more accurate data foundation for testing.

[0074] Step 202: Determine the number of server nodes in each target of the high-density server under test as the target server nodes for each round of testing, and determine the server nodes in the high-density server under test other than the target server nodes for each round of testing as the test server nodes for each round of testing.

[0075] It should be noted that after obtaining the target number, the number of rounds of testing the high-density server to be tested can be calculated numerically based on the target number and the number of server nodes in the high-density server to be tested.

[0076] Specifically, obtain the target quantity Afterwards, each of the high-density servers to be tested can be... In each server phase, the target server node for each round of testing is determined, and the server nodes in the high-density server under test other than the target server node for each round of testing are determined as the test server nodes for each round of testing.

[0077] For example, in the target quantity When =1, each server stage in the high-density server under test can be identified as the target server node for each round of testing, and the server nodes in the high-density server under test other than the target server node for each round of testing can be identified as the test server nodes for each round of testing.

[0078] For example, in the target quantity When the value is 2, every two server phases in the high-density server under test can be identified as the target server node for each round of testing, and the server nodes in the high-density server under test other than the target server node for each round of testing can be identified as the test server node for each round of testing.

[0079] For example, in the target quantity When the value is 3, every 3 server phases in the high-density server under test can be identified as the target server node for each round of testing, and the server nodes in the high-density server under test other than the target server node for each round of testing can be identified as the test server node for each round of testing.

[0080] As an optional embodiment, each target number of server nodes in the high-density server under test is determined as the target server node for each round of testing, and the server nodes in the high-density server under test other than the target server nodes for each round of testing are determined as the test server nodes for each round of testing. This includes: selecting a target number of server nodes from among the server nodes in the high-density server under test as the target server nodes for the first round of testing, and determining the server nodes in the high-density server under test other than the target server nodes for the first round of testing as the test servers for the first round of testing.

[0081] For the i-th round of testing, if the target number is less than the total number of server nodes in the high-density server under test, a combination of server nodes with the target number is selected again from among the server nodes in the high-density server under test. If it is determined that the newly selected combination of server nodes with the target number is different from the combination of target server nodes corresponding to each round of testing before the i-th round, the newly selected combination of server nodes with the target number is determined as the i-th round of testing. i The target server nodes corresponding to the round of testing will be the high-density servers under test, excluding the first one. i Server nodes other than the target server node in the round of testing are determined to be the first. i The test server node corresponding to this test. i Increase by 1, then return to the step of selecting the target number of server node combinations from each server node in the high-density server under test, until it is determined that there is no combination of the target number of server nodes in the high-density server under test that is different from the combination of target server nodes corresponding to each test before the i-th round of testing.

[0082] To facilitate understanding of the embodiments of the present invention, the following examples illustrate the specific process of determining the target server node and test server node for each round of testing in the embodiments of the present invention.

[0083] The number of server nodes in the high-density server under test is 4, and the target number is... When =1, the first server node in the high-density server under test is designated as the target server node for the first round of testing; the second, third, and fourth server nodes in the high-density server under test are designated as the test server nodes for the first round of testing. The second server node in the high-density server under test is designated as the target server node for the second round of testing; the third server node in the high-density server under test is designated as the target server node for the third round of testing; the fourth server node in the high-density server under test is designated as the target server node for the fourth round of testing; the first, second, and third server nodes in the high-density server under test are designated as the test server nodes for the fourth round of testing.

[0084] The number of server nodes in the high-density server under test is 4, and the target number is... With a value of 2, the first and second server nodes in the high-density server under test are designated as the target server nodes for the first round of testing, and the third and fourth server nodes are designated as the test server nodes for the first round of testing. The first and third server nodes in the high-density server under test are designated as the target server nodes for the second round of testing, and the second and fourth server nodes are designated as the test server nodes for the second round of testing. The first and fourth server nodes in the high-density server under test are designated as the target server nodes for the third round of testing, and the second and third server nodes are designated as the test service nodes for the third round of testing. The test nodes are determined as follows: The 2nd and 3rd server nodes in the high-density server under test are designated as the target server nodes for the 4th round of testing; the 1st and 3rd server nodes in the high-density server under test are designated as the test server nodes for the 4th round of testing; the 2nd and 4th server nodes in the high-density server under test are designated as the target server nodes for the 5th round of testing; the 1st and 3rd server nodes in the high-density server under test are designated as the test server nodes for the 5th round of testing; the 3rd and 4th server nodes in the high-density server under test are designated as the target server nodes for the 6th round of testing; and the 1st and 2nd server nodes in the high-density server under test are designated as the test server nodes for the 6th round of testing.

[0085] The number of server nodes in the high-density server under test is 4, and the target number is... When the number of nodes is 3, the 1st, 2nd, and 3rd server nodes in the high-density server under test are designated as the target server nodes for the first round of testing, and the 4th server node in the high-density server under test is designated as the test server node for the first round of testing. The 1st, 2nd, and 4th server nodes in the high-density server under test are designated as the target server nodes for the second round of testing, and the 3rd server node in the high-density server under test is designated as the test server node for the second round of testing. The 1st, 3rd, and 4th server nodes in the high-density server under test are designated as the target server nodes for the third round of testing, and the 2nd server node in the high-density server under test is designated as the test server node for the third round of testing. The 2nd, 3rd, and 4th server nodes in the high-density server under test are designated as the target server nodes for the fourth round of testing, and the 1st server node in the high-density server under test is designated as the test server node for the fourth round of testing.

[0086] The number of server nodes in the high-density server under test is 4, and the target number is... When the value is 4, the 1st, 2nd, 3rd, and 4th server nodes in the high-density server under test are determined as the target server nodes for the first round of testing. At this time, the test server nodes for the first round of testing are empty.

[0087] Step 203: In each round of testing, within the first time period of the first preset duration, control the test server node corresponding to each round of testing to switch between the power-on state and the power-off state at a preset frequency, and obtain the test data within the first time period as the test result of each round of testing. The test data includes log information generated by the high-density server under test and / or log information generated by the test tool.

[0088] Specifically, for the first i Round testing can control the first time period within the first preset duration. i During each round of testing, the test server nodes switch between power-on and power-off states at a preset frequency. Simultaneously, within the first time period mentioned above, log information generated by the high-density server under test and / or log information generated by the testing tools are obtained through data querying. This information is also used as the basis for the second round of testing. i The test results of the round test.

[0089] It should be noted that, in the target quantity Given the same number of server nodes as the high-density server under test, the target number can be... Update to 1, and return to steps 202 and 203 to obtain the target quantity. The test results corresponding to each round of testing when the value is 1, and thus the target number. Update to 2, and return to execute steps 202 and 203 to obtain the target quantity. The test results for each round of testing, assuming a value of 2, are repeated until the target number is obtained. for The test results corresponding to each round of testing.

[0090] It should be noted that the first preset duration and preset frequency in the embodiments of the present invention can be predefined based on prior knowledge and / or actual conditions. The specific values ​​of the first preset duration and the preset frequency are not limited in the embodiments of the present invention.

[0091] Optionally, the value of the first preset duration can be between 20 minutes and 40 minutes. For example, the value of the second preset duration can be 20 minutes, 30 minutes or 40 minutes.

[0092] Preferably, the first preset duration can be 30 minutes.

[0093] Optionally, the preset frequency can be between once every 40 seconds and once every 80 seconds. For example, the preset frequency can be once every 40 seconds, once every 60 seconds, or once every 80 seconds.

[0094] Preferably, the preset frequency can be once every 60 seconds.

[0095] It should be noted that in each round of testing, the fan speed of the high-density server under test is controlled by the system's pre-configured automatic speed adjustment measurement.

[0096] As an optional embodiment, in each round of testing, before controlling the test server node corresponding to each round of testing to switch between power-on and power-off states at a preset frequency during a first time period of a first preset duration, and acquiring the test data during the first time period as the test result of each round of testing, the method further includes: in each round of testing, performing stress testing on the target test node corresponding to each round of testing during a second time period of a second preset duration, and acquiring the test data during the second time period as the test result of each round of testing, wherein the end time of the second time period is the start time of the first time period.

[0097] Specifically, for the first i Round of testing, in the first i In the round of testing, the second time period of the second preset duration can be used to test the first round. i While stress testing is being performed on the target server nodes corresponding to the first test round, log information generated by the high-density server under test and / or log information generated by the testing tools during the second time period is obtained through data querying, and used as the basis for the second test round. i The test results of the round test.

[0098] In the i In round testing, the end time of the second time period can be determined as the start time of the first time period, and the first time period of the first preset duration can be controlled. i During each round of testing, the test server nodes switch between power-on and power-off states at a preset frequency. Simultaneously, within the first time period mentioned above, log information generated by the high-density server under test and / or log information generated by the testing tools are obtained through data querying. This information is also used as the basis for the second round of testing. i The test results of the round test.

[0099] It should be noted that, in the target quantity When the number of server nodes in the high-density server under test is the same, in the first round of testing, stress testing can be performed on the target server node corresponding to the first round of testing during the second time period of the second preset duration. At the same time, log information generated by the high-density server under test and / or log information generated by the testing tool during the second time period can be obtained by data querying and used as the test result of the first round of testing.

[0100] After obtaining the test results of the first round of testing, the target quantity can be... Update to 1, and return to steps 202 and 203 to obtain the target quantity. The test results corresponding to each round of testing when the value is 1, and thus the target number. Update to 2, and return to execute steps 202 and 203 to obtain the target quantity. The test results for each round of testing, assuming a value of 2, are repeated until the target number is obtained. for The test results corresponding to each round of testing.

[0101] It should be noted that the second preset duration in the embodiments of the present invention can be predefined based on prior knowledge and / or actual circumstances. The specific value of the second preset duration is not limited in the embodiments of the present invention.

[0102] Optionally, the value of the second preset duration can be between 20 minutes and 40 minutes. For example, the value of the second preset duration can be 20 minutes, 30 minutes or 40 minutes.

[0103] Preferably, the second preset duration can be 30 minutes.

[0104] Step 204: Based on the test results of each round of testing, obtain the test results of the high-density server to be tested.

[0105] Specifically, after obtaining the test results of each round of testing, the test results of the high-density server under test can be obtained based on the test results of each round of testing through mathematical statistics, numerical calculation, conditional judgment and deep learning technology.

[0106] As an optional embodiment, the log information generated by the high-density server under test includes operating system log information and baseboard management controller log information; the testing tools include a central processing unit stress testing tool.

[0107] Based on the test results of each round of testing, obtain the test results of the high-density server under test, including: based on the test results of each round of testing, evaluate the operation of the high-density server under test in each round of testing.

[0108] Based on the evaluation results of the operation of the high-density server under test in each round of testing, the test results of the high-density server under test are obtained.

[0109] It should be noted that operating system log information is used to record various events, errors, warnings, and system resource usage during system operation. Operating system log information is usually stored in the form of event records, and each event includes a timestamp, event type, event description, and other relevant attributes.

[0110] Baseboard Management Controller (BMC) logs are crucial data that records the BMC's operational status and events. These logs include server hardware status information such as CPU temperature, fan speed, power supply status, and hard drive health. When hardware malfunctions, the BMC logs corresponding warnings or error messages.

[0111] It should be noted that the CPU stress testing tools in the embodiments of the present invention include, but are not limited to, StressAppTest (SAT) and Performance Tuning Utility (PTU).

[0112] PTU is a CPU stress testing tool used to evaluate CPU performance and stability by simulating various load scenarios to test the CPU's response and processing capabilities.

[0113] SAT is a feature-rich, multi-threaded, multi-process concurrency testing tool that allows users to simulate a large number of concurrent requests to check the stability and performance limits of a system. Written in C++, it is highly scalable and supports custom scripts and plugins, enabling stress testing of various server-side and client-side applications. StressAppTest employs a flexible concurrency model, allowing the creation and control of a large number of concurrent threads or processes to simulate real-world high-concurrency scenarios, such as large-scale user access or big data processing tasks. It also supports Python as a scripting language, making test logic richer and easier to write. Furthermore, StressAppTest provides a plugin interface, allowing users to extend it with new test behaviors and analysis modules.

[0114] In each round of testing of the high-density server under test, the test server node corresponding to each round of testing switches between power-on and power-off states at a preset frequency within a preset time period. At the same time, the log information generated by the high-density server under test and / or the log information generated by the testing tool during the above-mentioned time period is used to evaluate whether alarm events occur, and whether the temperature, power consumption, frequency and other indicators of electronic devices such as the CPU in the high-density server under test are within the normal range. The operating status of the high-density server under test in each round of testing is evaluated, and the operating status evaluation results of the high-density server under test in each round of testing are obtained.

[0115] After obtaining the evaluation results of the operation of the high-density server under test in each round of testing, the test results of the high-density server under test can be obtained through mathematical statistics, conditional judgment and deep learning techniques.

[0116] It should be noted that the test results of the high-density server under test in this embodiment of the invention may include indication information indicating whether the high-density server under test has passed the test. The test results of the high-density server under test in this embodiment of the invention may also include scores describing the stability and reliability of the high-density server under test.

[0117] This invention, in its embodiments, obtains the number of usable server nodes in the high-density server under test based on the overall power consumption limit of the server under test. This number is then used as the target number. Each target number of server nodes in the high-density server under test is then designated as the target server node for each round of testing. All other server nodes in the high-density server under test, excluding the target server nodes for each round, are designated as test server nodes for each round of testing. In each round of testing, stress tests are performed on the target test nodes within a second time period of a second preset duration, and the test data within this second time period is used as the test result for each round. Then, within a first time period of a first preset duration, the test server nodes for each round of testing are controlled to switch between power-on and power-off states at a preset frequency, and the test data within this first time period is also used as the result for each round of testing. Based on the test results of each round of testing, the test results of the high-density server under test are obtained. This method avoids the risk of damage to the high-density server under test. By controlling the test server nodes for each round to switch between power-on and power-off states at a preset frequency during each test, the overall current of the high-density server under test fluctuates frequently. This intensifies electromagnetic interference between the server nodes and different electronic devices within the high-density server under test, thus better testing the reliability and stability of the high-density server under test under high-intensity, interactive electromagnetic interference. It also improves the comprehensiveness and accuracy of high-density server testing, and the testing method is simple, highly automated, and improves the testing efficiency of high-density servers.

[0118] As an optional embodiment, the test data includes: the power consumption value of the target server node and the total power consumption value of the high-density server under test for each round of testing; based on the test results of each round of testing, the test results of the high-density server under test are obtained, including: based on the test results of each round of testing, obtaining the recommended server node activation scheme for the high-density server under test as the test results of the high-density server under test, wherein the recommended server node activation scheme includes the server nodes recommended to be activated in the high-density server under test.

[0119] Specifically, for the first i Round of testing, in the first i In the round of testing, the second time period of the second preset duration can be used to test the first round. i While stress testing is being conducted on the target server nodes corresponding to the first test round, the overall power consumption of the high-density server under test during the second time period is obtained through data querying. i The power consumption value of the target server node corresponding to the round of testing is used as the first... i The test results of the round test.

[0120] In the i During the round of testing, the first time period of the first preset duration can also be controlled. i During each test round, the test server nodes switch between power-on and power-off states at a preset frequency. Simultaneously, within the first time period, data queries are used to obtain the overall power consumption value of the high-density server under test and the [other data points] within that time period. i The power consumption value of the target server node corresponding to the round of testing is also used as the first... i The test results of the round test.

[0121] Between obtaining the test results of each round of testing, the recommended activation scheme for the server nodes corresponding to the high-density server under test can be obtained based on the test results of each round of testing through numerical calculation, mathematical statistics, conditional judgment and deep learning technology, and used as the test results of the high-density server under test.

[0122] It should be noted that the recommended server node activation scheme for the high-density server under test in this embodiment of the invention includes the server nodes recommended to be activated in the high-density server under test. Activating the server nodes recommended to be activated in the high-density server under test can ensure that the energy efficiency of the high-density server under test is maximized while ensuring that the total power consumption of the high-density server under test does not exceed the limit value of the total power consumption of the high-density server under test during operation.

[0123] It is understandable that the recommended number of server nodes to be enabled in the high-density server under test is the target number.

[0124] As an optional embodiment, based on the test results of each round of testing, the recommended scheme for enabling the server node corresponding to the high-density server under test is obtained, including: based on the total power consumption value of the high-density server under test and the power consumption value of the target server node corresponding to each round of testing, the evaluation index value corresponding to each round of testing is obtained.

[0125] Specifically, for the first i Round testing, in this embodiment of the invention, can be used Indicates the first i The power consumption of the high-density server under test was obtained during the first time period in the round of testing. Indicates the first i The total power consumption of the high-density server under test was obtained during the second time period in the round of testing; Indicates the first i The first time period was obtained in the round of testing. i The power consumption value of the target server node corresponding to the round test is used. Indicates the first i The second time period was obtained in the round of testing. iThe power consumption value of the target server node corresponding to the round test.

[0126] In determining the first i The power consumption of the high-density server under test was obtained during the second time period in the round of testing. or the i The power consumption of the high-density server under test was obtained during the first time period in the round of testing. If the power consumption exceeds the overall power consumption limit of the high-density server under test, it is determined that the high-density server under test has failed the test. i Round test.

[0127] In determining the first i The power consumption of the high-density server under test was obtained during the second time period in the round of testing. or the i The power consumption of the high-density server under test was obtained during the first time period in the round of testing. Assuming that none of the values ​​exceed the overall power consumption limit of the high-density server under test, calculate the first... i The power consumption of the high-density server under test was obtained during the first time period in the round of testing. With the i The first time period was obtained in the round of testing. i The power consumption value of the target server node corresponding to the round test The difference, as the first i The first evaluation index value corresponding to the round test ; Calculate the first i The power consumption of the high-density server under test was obtained during the second time period in the round of testing. With the i The second time period was obtained in the round of testing. i The power consumption value of the target server node corresponding to the round test The difference, as the first i The second evaluation index value corresponding to the round test ; Calculate the first i The power consumption of the high-density server under test was obtained during the first time period in the round of testing. With the i The power consumption of the high-density server under test was obtained during the second time period in the round of testing. sum, and th i The first time period was obtained in the round of testing. i The power consumption value of the target server node corresponding to the round test With the i The power consumption of the high-density server under test was obtained during the second time period in the round of testing. The difference between the sums is used as the first... i The third evaluation index value corresponding to the round test The specific calculation formula is as follows: It should be noted that, in the embodiments of the present invention, the first... i The evaluation index values ​​corresponding to each round of testing can include the first round. i The first evaluation index value corresponding to the round test Second evaluation index value and the third evaluation index value Any one of them.

[0128] It should be noted that, in the embodiments of the present invention, the first... i The evaluation index value corresponding to the round of testing can also be based on the first round. i The first evaluation index value corresponding to the round test Second evaluation index value and the third evaluation index value Calculated, for example, the first i The first evaluation index value corresponding to the round test Second evaluation index value and the third evaluation index value The sum of these is determined to be the first. i The evaluation index values ​​corresponding to the round of testing; or, the evaluation index values ​​corresponding to the round of testing can also be... i The first evaluation index value corresponding to the round test Second evaluation index value and the third evaluation index value The weighted average of the values ​​is determined as the first. i The evaluation index values ​​corresponding to the round of testing; among them, the first round of testing... i The first evaluation index value corresponding to the round test Second evaluation index value and the third evaluation index value The weight values ​​can be predefined based on prior knowledge and / or actual circumstances.

[0129] Compare the evaluation index values ​​corresponding to each round of testing, and determine the round of testing with the smallest corresponding evaluation index value as the target round of testing.

[0130] Understandably, the first i The first evaluation index value corresponding to the round test , for the first i In the first time period of the round of testing i The power consumption value of the target server node corresponding to the round of testing. i The second evaluation index value corresponding to the round test , for the first i In the second period of the round of testing, the first i The power consumption value of the target server node corresponding to the round of testing; the first round iThe third evaluation index value corresponding to the round test , for the first i In the first round of testing i The power consumption value of the target server node corresponding to the round of testing. Therefore, regardless of the round... i The first evaluation index value corresponding to the round test Second evaluation index value and the third evaluation index value The smaller the value of the above evaluation index, the better the energy efficiency of the high-density server under test.

[0131] The first embodiment of the present invention i The evaluation index value corresponding to the round of testing is based on the first round. i The first evaluation index value corresponding to the round test Second evaluation index value and the third evaluation index value Therefore, the first i The smaller the evaluation index value corresponding to the test round, the better the energy efficiency of the high-density server under test.

[0132] After obtaining the evaluation index value for each round of testing, the round of testing with the smallest corresponding evaluation index value can be determined as the target round of testing.

[0133] The target server node corresponding to the target round test is determined to be the server node corresponding to the high-density server to be tested, and the server node recommended to be enabled in the recommended scheme is used.

[0134] Specifically, after determining the target round of testing, the target server node corresponding to the target test round can be identified as the server node recommended to be enabled in the recommended scheme for the high-density server under test.

[0135] This invention obtains the power consumption values ​​of the target server node and the total power consumption value of the high-density server under test in each round of testing as the test results. Based on the test results of each round, it obtains the recommended server node activation scheme for the high-density server under test with the best energy efficiency. This allows it to recommend the server node activation scheme with the best energy efficiency to users based on the test, which can better meet the needs of energy conservation and emission reduction and has broad market application value.

[0136] It's important to note that Python is a free, open-source, easy-to-learn, and easy-to-maintain programming language. Due to its open-source nature, Python has been ported to many platforms, becoming the primary language for scripting and rapid application development on most platforms. This is why this invention uses Python for programming, ensuring that our scripts can run directly on both Linux and Windows without code modification, meeting the requirements of different testing platforms.

[0137] In this embodiment of the invention, test scripts can be written in Python, thereby constructing a test environment by executing the test scripts, and executing the specific steps of the high-density server testing method provided by this invention in the test environment.

[0138] To facilitate understanding of the high-density server testing method provided by this invention, an example is provided below to illustrate the method. In this example, the high-density server under test includes four server nodes.

[0139] Figure 3 This is the second flowchart illustrating the high-density server testing method provided by this invention. For example... Figure 3 As shown, after the test begins, the test tools (PTU and StressAppTest, etc.) can be installed by executing the test script.

[0140] With all server nodes in the high-density server under test in a powered-off state and all fans in the high-density server under test at their maximum speed, the total power consumption value of the high-density server under test is obtained as the first total power consumption value.

[0141] Under the condition that any server node in the high-density server under test is running at full load, all other server nodes in the high-density server under test are shut down, and the speed of each fan in the high-density server under test is at its maximum, the total power consumption value of the high-density server under test is obtained as the second total power consumption value.

[0142] Control each fan in the high-density server under test to resume automatic speed adjustment strategy and regulate its speed, and control all server nodes in the high-density server under test to shut down.

[0143] Obtain the overall power consumption limit of the high-density server under test, and generate the target quantity based on the overall power consumption limit, the first overall power consumption value, and the second overall power consumption value of the high-density server under test.

[0144] Based on the target number, each target number of server nodes in the high-density server under test is determined as the target server node for each round of testing, and the server nodes in the high-density server under test other than the target server node for each round of testing are determined as the test server nodes for each round of testing.

[0145] Proceed to the first i Round of testing, and by querying data, obtain the first round of testing. i After reviewing the test results, determine whether the test is complete.

[0146] If it is determined that the test has not ended i Increase by 1, return to the execution state of the next step. i The process involves conducting rounds of testing, obtaining the test results of the first round through data query, and then determining whether the test is over until the test is confirmed to be over.

[0147] The test ends after obtaining the test results of the high-density server under test based on the test results of each round of testing.

[0148] With a target quantity M=1, the number of target server nodes corresponding to each round of testing is 1.

[0149] The first round of testing: During the second time period of the second preset duration, stress testing was performed on the first node, and the power consumption value of the entire high-density server under test and the power consumption value of the first server node were obtained during the second time period. These were used as the test results of the first round of testing. Then, during the first time period of the first preset duration, the second, third, and fourth server nodes were controlled to switch between power-on and power-off states at a preset frequency, and the power consumption value of the entire high-density server under test and the power consumption value of the first server node were obtained during the first time period. These were also used as the test results of the first round of testing.

[0150] The second round of testing: During the second time period of the second preset duration, stress testing was performed on the second node, and the power consumption value of the high-density server under test and the power consumption value of the second server node were obtained during the second time period. These were used as the test results of the second round of testing. Then, during the first time period of the first preset duration, the first, third, and fourth server nodes were controlled to switch between power-on and power-off states at a preset frequency, and the power consumption value of the high-density server under test and the power consumption value of the second server node were obtained during the first time period. These were also used as the test results of the second round of testing.

[0151] The third round of testing: During the second time period of the second preset duration, stress testing was conducted on the third node, and the power consumption values ​​of the entire high-density server under test and the third server node were obtained during the second time period. These were used as the test results of the third round of testing. Then, during the first time period of the first preset duration, the first, second, and fourth server nodes were controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the third server node were obtained during the first time period. These were also used as the test results of the third round of testing.

[0152] Fourth round of testing: During the second time period of the second preset duration, stress test is performed on the fourth node, and the power consumption value of the high-density server under test and the power consumption value of the fourth server node are obtained during the second time period. This is used as the test result of the fourth round of testing. Then, during the first time period of the first preset duration, the first, second, and third server nodes are controlled to switch between power-on and power-off states at a preset frequency, and the power consumption value of the high-density server under test and the power consumption value of the fourth server node are obtained during the first time period. This is also used as the test result of the fourth round of testing.

[0153] With a target quantity M=2, the number of target server nodes corresponding to each round of testing is 2.

[0154] The first round of testing: During the second time period of the second preset duration, stress tests were performed on the first and second nodes, and the power consumption values ​​of the entire high-density server under test and the first and second server nodes were obtained during the second time period. These were used as the test results of the first round of testing. Then, during the first time period of the first preset duration, the third and fourth server nodes were controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the first and second server nodes were obtained during the first time period. These were also used as the test results of the first round of testing.

[0155] The second round of testing: During the second time period of the second preset duration, stress tests were conducted on the 1st and 3rd nodes, and the power consumption values ​​of the entire high-density server under test and the 1st and 3rd server nodes were obtained during the second time period. These were used as the test results of the second round of testing. Then, during the first time period of the first preset duration, the 2nd and 4th server nodes were controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the 1st and 3rd server nodes were obtained during the first time period. These were also used as the test results of the second round of testing.

[0156] The third round of testing: During the second time period of the second preset duration, stress tests were conducted on the 1st and 4th nodes, and the power consumption values ​​of the entire high-density server under test and the 1st and 4th server nodes were obtained during the second time period. These were used as the test results of the third round of testing. Then, during the first time period of the first preset duration, the 2nd and 3rd server nodes were controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the 1st and 4th server nodes were obtained during the first time period. These were also used as the test results of the third round of testing.

[0157] Fourth round of testing: During the second time period of the second preset duration, stress tests are performed on the second and third nodes, and the power consumption values ​​of the entire high-density server under test and the power consumption values ​​of the second and third server nodes are obtained during the second time period. These are used as the test results of the fourth round of testing. Then, during the first time period of the first preset duration, the first and fourth server nodes are controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the power consumption values ​​of the second and third server nodes are obtained during the first time period. These are also used as the test results of the fourth round of testing.

[0158] Fifth round of testing: During the second time period of the second preset duration, stress tests are performed on the 2nd and 4th nodes, and the power consumption values ​​of the entire high-density server under test and the 2nd and 4th server nodes are obtained during the second time period. These are used as the test results of the fifth round of testing. Then, during the first time period of the first preset duration, the 1st and 3rd server nodes are controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the 2nd and 4th server nodes are obtained during the first time period. These are also used as the test results of the fifth round of testing.

[0159] The sixth round of testing: During the second time period of the second preset duration, stress tests were conducted on the 3rd and 4th nodes, and the power consumption values ​​of the entire high-density server under test and the 3rd and 4th server nodes were obtained during the second time period. These were used as the test results of the sixth round of testing. Then, during the first time period of the first preset duration, the 1st and 2nd server nodes were controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the 3rd and 4th server nodes were obtained during the first time period. These were also used as the test results of the sixth round of testing.

[0160] With a target quantity M=3, the number of target server nodes corresponding to each round of testing is 3.

[0161] The first round of testing: During the second time period of the second preset duration, stress tests were conducted on the 1st, 2nd, and 3rd nodes, and the power consumption values ​​of the entire high-density server under test and the 1st, 2nd, and 3rd server nodes were obtained during the second time period. These were used as the test results of the first round of testing. Then, during the first time period of the first preset duration, the 4th server node was controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the 1st, 2nd, and 3rd server nodes were obtained during the first time period. These were also used as the test results of the first round of testing.

[0162] The second round of testing: During the second time period of the second preset duration, stress tests were conducted on the 1st, 2nd, and 4th nodes, and the power consumption values ​​of the entire high-density server under test and the 1st, 2nd, and 4th server nodes were obtained during the second time period. These were used as the test results of the second round of testing. Then, during the first time period of the first preset duration, the 3rd server node was controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the 1st, 2nd, and 4th server nodes were obtained during the first time period. These were also used as the test results of the second round of testing.

[0163] The third round of testing: During the second time period of the second preset duration, stress tests were conducted on the 1st, 3rd, and 4th nodes, and the power consumption values ​​of the entire high-density server under test and the 1st, 3rd, and 4th server nodes were obtained during the second time period. These results were used as the test results of the third round of testing. Then, during the first time period of the first preset duration, the 2nd server node was controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the 1st, 3rd, and 4th server nodes were obtained during the first time period. These results were also used as the test results of the third round of testing.

[0164] Fourth round of testing: During the second time period of the second preset duration, stress tests are performed on the 2nd, 3rd, and 4th nodes, and the power consumption values ​​of the entire high-density server under test and the 2nd, 3rd, and 4th server nodes are obtained during the second time period. These are used as the test results of the fourth round of testing. Then, during the first time period of the first preset duration, the 1st server node is controlled to switch between power-on and power-off states at a preset frequency, and the power consumption values ​​of the entire high-density server under test and the 2nd, 3rd, and 4th server nodes are obtained during the first time period. These are also used as the test results of the fourth round of testing.

[0165] With a target quantity M=4, the number of target server nodes corresponding to each round of testing is 4.

[0166] The first round of testing: During the second time period of the second preset duration, stress tests were conducted on nodes 1, 2, 3, and 4. The power consumption of the entire high-density server under test and the power consumption of the first, 2, 3, and 4 server nodes were obtained during the second time period. These results were used as the test results for the first round of testing, which could then be used to determine the target number of... Update to 1, and return to steps 202 and 203 to obtain the target quantity. The test results corresponding to each round of testing when the value is 1, and thus the target number. Update to 2, and return to execute steps 202 and 203 to obtain the target quantity. The test results for each round of testing, assuming a value of 2, are repeated until the target number is obtained. for The test results corresponding to each round of testing.

[0167] The high-density server testing method provided by this invention uses a Python testing environment and only requires one control machine to test high-density servers with multiple server nodes. It is simple to operate, highly automated, reduces the need for manpower and material resources, greatly improves testing efficiency, and can also recommend an energy-efficient server node activation scheme for users, thereby achieving the requirements of energy conservation and emission reduction.

[0168] Figure 4 This is a structural schematic diagram of the high-density server testing device provided by the present invention. The following is in conjunction with… Figure 4 The high-density server testing apparatus provided by this invention is described below. The high-density server testing apparatus described below can be referred to in correspondence with the high-density server testing method provided by this invention described above. For example... Figure 4 As shown, the device includes: a data acquisition module 401, a node determination module 402, a node testing module 403, and a result output module 404.

[0169] The data acquisition module 401 is used to acquire the overall power consumption limit value of the high-density server under test, and to acquire the target quantity based on the overall power consumption limit value.

[0170] The node determination module 402 is used to determine each target number of server nodes in the high-density server under test as the target server node corresponding to each round of testing, and to determine the server nodes in the high-density server under test other than the target server node corresponding to each round of testing as the test server node corresponding to each round of testing.

[0171] The node testing module 403 is used to control the test server node corresponding to each round of testing to switch between power-on and power-off states at a preset frequency within a first time period of a first preset duration in each round of testing, and to obtain the test data within the first time period as the test result of each round of testing. The test data includes log information generated by the high-density server under test and / or log information generated by the testing tool.

[0172] The result output module 404 is used to obtain the test results of the high-density server under test based on the test results of each round of testing.

[0173] Specifically, the data acquisition module 401, the node determination module 402, the node testing module 403, and the result output module 404 are electrically connected.

[0174] Figure 5 This is a schematic diagram showing the connection between the high-density server testing device provided by this invention and the high-density server under test. The connection relationship between the high-density server testing device and the high-density server under test in this embodiment of the invention is as follows: Figure 5 As shown.

[0175] The high-density server testing device in this embodiment of the invention obtains the number of usable server nodes in the high-density server under test based on the overall power consumption limit value of the server under test, and sets this as the target number. Then, it determines each target number of server nodes in the high-density server under test as the target server nodes for each round of testing, and determines all server nodes in the high-density server under test other than the target server nodes for each round of testing as the test server nodes for each round of testing. In each round of testing, stress tests are performed on the target test nodes within a second time period of a second preset duration, and the test data within the second time period is obtained as the test result for each round of testing. Then, within a first time period of a first preset duration, the test server nodes for each round of testing are controlled to switch between power-on and power-off states at a preset frequency, and the test data within the first time period is obtained. According to the test results of each round of testing, the test results of the high-density server under test are obtained based on the test results of each round of testing. While avoiding the risk of damage to the high-density server under test, by controlling the test server node corresponding to each round of testing to switch between power-on and power-off states at a preset frequency, the overall current of the high-density server under test fluctuates frequently. This makes the electromagnetic interference between the server nodes and different electronic devices in the high-density server under test more severe, so as to better test the reliability and stability of the high-density server under test under high-intensity, interactive electromagnetic interference. It can test the operation of the high-density server node under test under high-intensity, interactive electromagnetic interference, improve the comprehensiveness and accuracy of the high-density server test, and the test method is simple, highly automated, and can improve the testing efficiency of high-density servers.

[0176] Figure 6 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 6 As shown, the electronic device may include: a processor 610, a communications interface 620, a memory 630, and a communications bus 640, wherein the processor 610, the communications interface 620, and the memory 630 communicate with each other through the communications bus 640. The processor 610 can call logical instructions in the memory 630 to execute a high-density server testing method. This method includes: obtaining the overall power consumption limit of the high-density server under test; obtaining a target number based on the overall power consumption limit; determining each target number of server nodes in the high-density server under test as the target server nodes for each round of testing; determining all server nodes in the high-density server under test other than the target server nodes for each round of testing as the test server nodes for each round of testing; in each round of testing, controlling the test server nodes for each round of testing to switch between power-on and power-off states at a preset frequency within a first time period of a first preset duration, and obtaining test data within the first time period as the test result for each round of testing. The test data includes log information generated by the high-density server under test and / or log information generated by the testing tool; and obtaining the test result of the high-density server under test based on the test result of each round of testing.

[0177] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0178] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the high-density server testing method provided by the above methods. The method includes: obtaining the overall power consumption limit value of the high-density server under test; obtaining a target number based on the overall power consumption limit value; determining each target number of server nodes in the high-density server under test as the target server nodes corresponding to each round of testing; determining the server nodes in the high-density server under test other than the target server nodes corresponding to each round of testing as the test server nodes corresponding to each round of testing; in each round of testing, controlling the test server nodes corresponding to each round of testing to switch between a power-on state and a power-off state at a preset frequency within a first time period of a first preset duration, and obtaining the test data within the first time period as the test result of each round of testing. The test data includes log information generated by the high-density server under test and / or log information generated by the testing tool; and obtaining the test result of the high-density server under test based on the test result of each round of testing.

[0179] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program is implemented to perform the high-density server testing method provided by the above methods. The method includes: obtaining the overall power consumption limit value of the high-density server under test; obtaining a target number based on the overall power consumption limit value; determining each target number of server nodes in the high-density server under test as the target server nodes corresponding to each round of testing; determining the server nodes in the high-density server under test other than the target server nodes corresponding to each round of testing as the test server nodes corresponding to each round of testing; in each round of testing, controlling the test server nodes corresponding to each round of testing to switch between a power-on state and a power-off state at a preset frequency within a first time period of a first preset duration, and obtaining the test data within the first time period as the test result of each round of testing. The test data includes log information generated by the high-density server under test and / or log information generated by the testing tool; and obtaining the test result of the high-density server under test based on the test result of each round of testing.

[0180] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0181] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0182] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for testing high-density servers, characterized in that, include: Obtain the overall power consumption limit of the high-density server under test, and obtain the target quantity based on the overall power consumption limit; Each of the target number of server nodes in the high-density server under test is determined as the target server node for each round of testing, and the server nodes in the high-density server under test other than the target server nodes for each round of testing are determined as the test server nodes for each round of testing. In each round of testing, stress tests are performed on the target server node corresponding to each round of testing during a second time period of a second preset duration, and the test data during the second time period is obtained as the test result of each round of testing. In each round of testing, the test server node corresponding to each round of testing is controlled to switch between power-on and power-off states at a preset frequency during a first time period of a first preset duration, and the test data during the first time period is obtained, which is also used as the test result of each round of testing. The end time of the second time period is the start time of the first time period. The test data includes log information generated by the high-density server under test and / or log information generated by the testing tool. Based on the test results of each round of testing, the test results of the high-density server under test are obtained; The step of determining each target number of server nodes in the high-density server under test as the target server node for each round of testing, and determining the server nodes in the high-density server under test other than the target server nodes for each round of testing as the test server nodes for each round of testing, includes: Select the target number of server nodes among the server nodes in the high-density server to be tested as the target server nodes for the first round of testing, and determine the server nodes in the high-density server to be tested other than the target server nodes for the first round of testing as the test servers for the first round of testing. For the i-th round of testing, if the target number is less than the total number of server nodes in the high-density server under test, a combination of server nodes with the target number is selected again from among the server nodes in the high-density server under test. If it is determined that the combination of server nodes with the target number selected again is different from the combination of target server nodes corresponding to each round of testing before the i-th round of testing, the combination of server nodes with the target number selected again is determined as the target server node corresponding to the i-th round of testing. All server nodes in the high-density server under test other than the target server node corresponding to the i-th round of testing are determined as the test server nodes corresponding to the i-th round of testing. i is incremented by 1, and the process of selecting the combination of server nodes with the target number selected again from among the server nodes in the high-density server under test is repeated until it is determined that there is no combination of server nodes with the target number that is different from the combination of target server nodes corresponding to each round of testing before the i-th round of testing in the high-density server under test; i is a natural number greater than 1.

2. The high-density server testing method according to claim 1, characterized in that, The test data includes: the power consumption value of the target server node corresponding to each round of testing and the total power consumption value of the high-density server under test; The process of obtaining the test results of the high-density server under test based on the test results of each round of testing includes: Based on the test results of each round of testing, a recommended server node activation scheme is obtained for the high-density server under test, which is used as the test result of the high-density server under test. The recommended server node activation scheme includes the server nodes recommended to be activated in the high-density server under test.

3. The high-density server testing method according to claim 2, characterized in that, The step of obtaining the recommended scheme for enabling the server node corresponding to the high-density server under test based on the test results of each round of testing includes: Based on the total power consumption value of the high-density server under test and the power consumption value of the target server node corresponding to each round of testing obtained in each round of testing, the evaluation index value corresponding to each round of testing is obtained. Compare the evaluation index values ​​corresponding to each round of the test, and determine the round of the test with the smallest corresponding evaluation index value as the target round of the test; The target server node corresponding to the target round test is determined to be the server node recommended to be enabled in the recommended scheme for the server node corresponding to the high-density server under test.

4. The high-density server testing method according to any one of claims 1 to 3, characterized in that, The step of obtaining the target quantity based on the overall power consumption limit value includes: With all server nodes in the high-density server under test in a powered-off state and all fans in the high-density server under test at their maximum speed, the total power consumption of the high-density server under test is obtained as the first total power consumption value. When any server node in the high-density server under test is running at full load, all other server nodes in the high-density server under test are shut down, and the speed of each fan in the high-density server under test is at its maximum, the total power consumption of the high-density server under test is obtained as the second total power consumption value. The target quantity is obtained based on the first total power consumption value, the second total power consumption value, and the total power consumption limit value.

5. The high-density server testing method according to claim 4, characterized in that, The step of obtaining the target quantity based on the first overall power consumption value, the second overall power consumption value, and the overall power consumption limit value includes: Calculate the difference between the second total power consumption value and the first total power consumption value as the first power consumption value; calculate the sum of the second total power consumption value and the first power consumption value as the second power consumption value; calculate the sum of the second total power consumption value and twice the first power consumption value as the third power consumption value; calculate the sum of the second total power consumption value and three times the first power consumption value as the fourth power consumption value; calculate the sum of the second total power consumption value and four times the first power consumption value as the fifth power consumption value. Based on the relationship between the first total power consumption value and at least two of the second total power consumption value, the third power consumption value, the fourth power consumption value, and the fifth power consumption value, the number of server nodes that can be enabled in the high-density server under test is obtained as the target number.

6. A high-density server testing device, characterized in that, include: The data acquisition module is used to acquire the overall power consumption limit value of the high-density server under test, and to acquire the target quantity based on the overall power consumption limit value; The node determination module is used to determine each of the target number of server nodes in the high-density server under test as the target server node corresponding to each round of testing, and to determine the server nodes in the high-density server under test other than the target server nodes corresponding to each round of testing as the test server nodes corresponding to each round of testing. The node testing module is used to perform stress tests on the target server node corresponding to each round of testing within a second time period of a second preset duration in each round of testing, and to obtain the test data within the second time period as the test result of each round of testing. In each round of testing, the module controls the test server node corresponding to each round of testing to switch between power-on and power-off states at a preset frequency within a first time period of a first preset duration, and to obtain the test data within the first time period as the test result of each round of testing. The end time of the second time period is the start time of the first time period. The test data includes log information generated by the high-density server under test and / or log information generated by the testing tool. The result output module is used to obtain the test results of the high-density server under test based on the test results of each round of testing; The node determination module is specifically used for: Select the target number of server nodes among the server nodes in the high-density server to be tested as the target server nodes for the first round of testing, and determine the server nodes in the high-density server to be tested other than the target server nodes for the first round of testing as the test servers for the first round of testing. For the i-th round of testing, if the target number is less than the total number of server nodes in the high-density server under test, a combination of server nodes with the target number is selected again from among the server nodes in the high-density server under test. If it is determined that the combination of server nodes with the target number selected again is different from the combination of target server nodes corresponding to each round of testing before the i-th round of testing, the combination of server nodes with the target number selected again is determined as the target server node corresponding to the i-th round of testing. All server nodes in the high-density server under test other than the target server node corresponding to the i-th round of testing are determined as the test server nodes corresponding to the i-th round of testing. i is incremented by 1, and the process of selecting the combination of server nodes with the target number selected again from among the server nodes in the high-density server under test is repeated until it is determined that there is no combination of server nodes with the target number that is different from the combination of target server nodes corresponding to each round of testing before the i-th round of testing in the high-density server under test; i is a natural number greater than 1.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the high-density server testing method as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the high-density server testing method as described in any one of claims 1 to 5.

Citation Information

Patent Citations

  • System for performing power supply cycle period test on multiple servers

    CN102479127A

  • Server pressure testing method and device

    CN116881124A