A server abnormal heat dissipation detection system and server

Abstract

The application discloses a server heat dissipation abnormity detection system and a server, relates to the technical field of server hardware management, and comprises a plurality of light sensing cables used for judging the in-place state of a wind shielding component corresponding to each light sensing cable; wherein the wind shielding component is initially installed in the light indication range of a fan and provides light shielding for the light sensing segment of the light sensing cable; a signal connector is used for receiving the light state signals transmitted by the plurality of light sensing cables; a programmable logic device is used for receiving and saving the light state signals of the plurality of light sensing cables sent by the signal connector; and a baseboard management controller is used for accessing the programmable logic device through an I2C interface. By arranging the light sensing cables below the wind shielding component, the light shielding property of the wind shielding component and the photosensitivity of the light sensing cables are utilized to quickly locate the abnormity caused by the falling of the wind shielding component in the server heat dissipation system, the detection cost of the server heat dissipation abnormity is reduced, and the maintenance efficiency of the heat dissipation system is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of server hardware management, and particularly relates to a server heat dissipation anomaly detection system and a server. BACKGROUND

[0002] With the rapid development of business demands such as artificial intelligence and big data, the power consumption of servers and the device density inside the servers are increasing, and the heat dissipation requirements for the servers are also increasing. The wind-blocking foam is a material that can be easily cut and is not transparent to wind. In the server heat dissipation system, it is usually used as a wind-blocking component pasted on the side wall of the case and the like to prevent the hot air discharged by the fan from flowing back to the air inlet side, thereby reducing the heat dissipation circulation efficiency.

[0003] However, due to reasons such as aging of the glue or insufficient glue application, the wind-blocking foam may fall off. The fallen wind-blocking foam may damage the original smooth air duct and even become an obstacle to the discharge of hot air. The blocked hot air forms a vortex and a high-pressure area near the fan outlet. Due to the suction of the fan, these hot air is reabsorbed into the air inlet side of the fan, which causes the temperature inside the server to continuously rise, resulting in the problem of server heat dissipation anomaly. The related art can only determine that the server heat dissipation is abnormal by phenomena such as fan speed anomaly or server temperature over-temperature, which leads to low maintenance efficiency of the heat dissipation system. SUMMARY

[0004] The present application provides a server heat dissipation anomaly detection system and a server to solve at least one of the above technical problems.

[0005] In a first aspect, an embodiment of the present application provides a server heat dissipation anomaly detection system, comprising:

[0006] A plurality of light sensing cables are configured to determine the in-situ state of the wind-blocking component corresponding to each light sensing cable; wherein the wind-blocking component is initially installed in the light range of the indicator light of the fan and provides light shielding for the light sensing segment of the light sensing cable.

[0007] A signal connector is configured to receive the light state signals transmitted by the plurality of light sensing cables.

[0008] A programmable logic device is configured to receive and save the light state signals of the plurality of light sensing cables sent by the signal connector.

[0009] A baseboard management controller is configured to access the programmable logic device through an I2C interface and execute the following detection method:

[0010] According to the light state signals of each light sensing cable, the in-situ state of the wind-blocking component corresponding to each light sensing cable is determined.

[0011] When the in-place state is the first state representing that the wind shielding component has fallen off, if an error between the set rotation speed value and the measured rotation speed value of the target fan is greater than a preset error threshold, a first abnormality detection result representing that the wind shielding component has fallen into the target fan is outputted.

[0012] When the in-place state is the first state representing that the wind shielding component has fallen off, if an error between the set rotation speed value and the measured rotation speed value is not greater than the error threshold, and after the rotation speed of the target fan is controlled to be increased, a change value of the temperature of the corresponding region of the target fan after a first time length is less than a preset temperature drop threshold, a second abnormality detection result representing that the wind shielding component has fallen into the server case is outputted.

[0013] In some embodiments, the baseboard management controller is further configured to:

[0014] According to a preset first time interval, poll the light state signals of the light sensing cables until the light state signal of the target light sensing cable represents that the light sensing cable detects light; the target light sensing cable is one or more of the plurality of light sensing cables;

[0015] Repeat the acquisition of the light state signal of the target light sensing cable N times according to a preset second time interval; wherein N is a positive integer;

[0016] If the light state signals acquired N times are the same, it is determined that the in-place state of the wind shielding component corresponding to the target light sensing cable is the first state.

[0017] In some embodiments, the baseboard management controller is further configured to:

[0018] If the light state signal of the target light sensing cable represents that the target light sensing cable does not detect light, it is determined that the in-place state is a second state representing that the wind shielding component is located at an initial installation position.

[0019] In some embodiments, the server heat dissipation abnormality detection system further comprises:

[0020] The interface extender connected with the plurality of light sensing cables is configured to receive the light state signals transmitted by the plurality of light sensing cables, and when the light state signal represents that the target light sensing cable detects light, send an interrupt signal to the signal connector through the I2C interface, so that the signal connector transmits the interrupt signal to the programmable logic device through the I2C interface, triggering the baseboard management controller to read the light state signal of the target light sensing cable.

[0021] In some embodiments, the baseboard management controller is further configured to:

[0022] In response to the interrupt signal sent by the programmable logic device, repeat the acquisition of the light state signal of the target light sensing cable N times according to a preset second time interval; wherein N is a positive integer;

[0023] If the N times of acquired illumination state signals are same, it is determined that the in-place state of the wind shielding component corresponding to the target light sensing cable is the first state.

[0024] In some embodiments, the baseboard management controller is further configured to:

[0025] When the in-place state is the first state representing that the wind shielding component has fallen off, if the error between the set rotation speed value and the measured rotation speed value is not greater than an error threshold value, and after the rotation speed of the target fan is controlled to increase, the temperature change value of the region corresponding to the target fan after the first time length is not less than a temperature drop threshold value, a third abnormality detection result representing light sensing abnormality is output.

[0026] In some embodiments, the baseboard management controller is further configured to:

[0027] Before the rotation speed of the target fan is increased, if the measured rotation speed value of the target fan is not less than a preset maximum rotation speed threshold value, the measured rotation speed value is maintained, and after a second time length, it is determined that the temperature of the region corresponding to the target fan is greater than a preset temperature threshold value, a second abnormality detection result representing that the wind shielding component has fallen into the server case is output; wherein the second time length is greater than the first time length.

[0028] In some embodiments, the baseboard management controller is further configured to:

[0029] According to the use time length of the target fan, the error threshold value is determined; wherein the use time length is positively correlated with the size of the error threshold value.

[0030] In some embodiments, the baseboard management controller is further configured to:

[0031] When the in-place state is the first state representing that the wind shielding component has fallen off, at least one target fan is determined according to the fan number associated with the wind shielding component.

[0032] In a second aspect, the embodiments of the present application provide a server, comprising the server heat dissipation abnormality detection system provided in the first aspect.

[0033] The server heat dissipation abnormality detection system and the server provided in the embodiments of the present application can realize the logic of judging whether the wind shielding component is in place by arranging the light sensing cable below the wind shielding component, using the indicator light of the fan, the light shielding property of the wind shielding component, and the photosensitive property of the light sensing cable, combining the difference between the set rotation speed value and the actual rotation speed value of the fan, judging the specific influence of the falling of the wind shielding component on the server heat dissipation, thereby quickly positioning the abnormality in the server heat dissipation system caused by the falling of the wind shielding component, reducing the detection cost of the server heat dissipation abnormality, and improving the maintenance efficiency of the heat dissipation system. BRIEF DESCRIPTION OF DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, the drawings described below only illustrate some of the embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort based on these drawings.

[0035] Figure 1 A schematic diagram of a server heat dissipation anomaly detection system provided by an embodiment of the present application;

[0036] Figure 2 A schematic diagram of a hardware architecture of a server heat dissipation anomaly detection system related to a signal connector provided by an embodiment of the present application;

[0037] Figure 3 A schematic diagram of a flow of a detection method performed by a basic management controller provided by an embodiment of the present application;

[0038] Figure 4 A schematic diagram of a flow related to in-place state judgment provided by an embodiment of the present application;

[0039] Figure 5 A schematic diagram of a hardware architecture of a server heat dissipation anomaly detection system related to an interface expander provided by an embodiment of the present application. DETAILED DESCRIPTION

[0040] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort fall within the protection scope of the present application.

[0041] It should be noted that, in the description of the present application, the terms "comprise", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment comprising a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or equipment. The terms "first", "second" and the like in the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence.

[0042] In order to make those skilled in the art better understand the present application, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0043] In a first aspect, as Figure 1As shown, the embodiments of the present application provide a server heat dissipation anomaly detection system, which includes a processor 110, a programmable logic device 120, and a plurality of light sensing cables 130 (light sensing cable 131, light sensing cable 132, …, light sensing cable 13n). The plurality of light sensing cables 130 are used to determine the in-place state of the corresponding wind blocking component of each light sensing cable 130. The processor 110 can communicate with the programmable logic device 120 to read the light state signals of the plurality of light sensing cables 130 stored in the programmable logic device 120, and is used to execute the detection method.

[0044] The processor 110 can be implemented in the form of a baseboard management controller (BMC) or a general-purpose processor, or other hardware forms.

[0045] The programmable logic device 120 can be implemented in the form of at least one of a programmable logic array (PLA), a field programmable gate array (FPGA), a complex programmable logic device (CPLD), or other programmable logic devices.

[0046] Specifically, the server heat dissipation anomaly detection system can include a hardware architecture as shown in the following figure: Figure 2 The hardware architecture includes a plurality of light sensing cables Sensor Cable, a signal connector CONN (Connector), a programmable logic device (FPGA / CPLD), and a baseboard management controller BMC. The plurality of light sensing cables Sensor Cable are used to determine the in-place state of the corresponding wind blocking component of each light sensing cable. The wind blocking component is initially installed within the light range of the indicator light of the fan and provides light blocking for the light sensing segment of the light sensing cable. The CONN is used to receive the light state signals transmitted by the plurality of light sensing cables. The BMC is used to receive and save the light state signals of the plurality of light sensing cables sent by the signal connector.

[0047] The light sensing cable includes a light-sensitive resistor Rg and a 100K voltage dividing resistor, and the light-sensitive resistor Rg is connected in parallel with an anti-interference capacitor C. The light sensing cable is powered by P3V3_STBY. In the voltage dividing state of the light-sensitive resistor Rg and the voltage dividing resistor, the voltage is about 3.3V (high level) without light and the voltage is lower than 0.2V (low level) with light. The light-sensitive resistor can be a small dark light-sensitive resistor, which can detect low-intensity light. The light surface resistor can face the wind-blocking component and be fixed on the server case. The n light sensing cables Sensor Cable are connected with the ports Senor_GPIO0 to Senor_GPIOn of CONN, and the light illumination state signals are transmitted to the FPGA / CPLD through the connection ports GPIO0 to GPIOn (general input / output ports) of the programmable logic device (FPGA / CPLD). The light illumination state signals are saved in the registers in the FPGA / CPLD. The baseboard management controller BMC is connected with the FPGA / CPLD through BMC_I2C (Inter-Integrated Circuit, Inter-Integrated Circuit). The BMC can poll the registers in the FPGA / CPLD through the BMC_I2C interface to read the light illumination state signals of each light sensing cable.

[0048] When there is no strong electromagnetic interference or vibration interference in the connection path of CONN, the system architecture can reduce the hardware cost.

[0049] As shown in Figure 3 , the baseboard management controller in the server heat dissipation anomaly detection system can be used to perform the detection method corresponding to steps S301, S302 and S303.

[0050] Step S301: According to the light illumination state signals of each light sensing cable, the in-place state of the wind-blocking component corresponding to each light sensing cable is determined.

[0051] The wind-blocking component is initially installed in the light illumination range of the indicator light of the target fan, and provides light shielding for the light sensing segment of the light sensing cable.

[0052] The wind-blocking component is usually a wind-blocking bubble cotton, which is a functional component made of high molecular foam material. In the server heat dissipation system, it is usually pasted on the side wall of the server case and other positions to block wind and prevent air backflow. The wind-blocking bubble cotton can be processed into any shape through die cutting, laser cutting and other processes to accurately match the complex installation space in the server case.

[0053] For example, part of the server cooling system guides the fan airflow to blow towards the central processing unit, graphics card and other core heat generating components through the fairing. The connection between the fairing and the case is prone to gaps. The gaps can be closed by pasting a special-shaped wind-blocking bubble cotton (such as L-shaped or U-shaped) that matches the shape of the gap, so as to ensure that the airflow flows through the heat generating components, thereby smoothly cooling the heat generating components.

[0054] Each fan on the server fan plate has a status indicator light (such as blue or green) for reflecting the running state of the fan. The installation position of the wind-blocking component is usually within the light range of the fan indicator light.

[0055] The light sensing cable is a signal transmission cable integrated with a light sensing element (such as a photoresistor or a photodiode). The light sensing section (e.g., the part containing the photoresistor) can form a pasting assembly relationship with the wind-blocking component, for example, the surface of the wind-blocking component completely covers the light sensing section, thereby blocking the indicator light of the target fan. As can be understood, in addition to the wind-blocking property, the wind-blocking component also needs to have the light-blocking property, for example, the surface of the wind-blocking bubble cotton can be coated with black ink to meet the light-blocking requirement. In addition, the wind-blocking component can also be a silica gel wind-blocking and light-blocking pad, a nylon fiber felt pad, a high-temperature-resistant rubber composite, etc. that simultaneously have the wind-blocking and light-blocking properties. The present application does not limit this.

[0056] The target fan is one or more fans whose indicator lights can illuminate the wind-blocking component. During the hardware and software design process, each fan and each wind-blocking component can be numbered, and the number of the wind-blocking component can be associated with the number of the fan whose indicator light can illuminate the wind-blocking component, so as to determine the target fan according to the number association relationship. At the same time, the light sensing cable can be associated with the wind-blocking component, so as to determine the fallen wind-blocking component according to the association relationship when the light sensing cable detects light.

[0057] For example, all the fans in the server case are uniquely numbered. The fan number can include a physical position identifier and a logical identifier. The physical position identifier is defined based on the installation area of the fan in the case, such as “CPU0 area fan 1”. The logical identifier is a unique digital code that can be recognized by the baseboard management controller, such as “Fan_01”. The fan number is bound to the hardware interface (speed control interface, speed feedback interface, indicator light interface) of the fan and stored in the non-volatile memory of the baseboard management controller.

[0058] All the wind-blocking components in the server case are uniquely numbered. The wind-blocking component number can include a corresponding cooling area identifier and a wind-blocking component serial number, such as “CPU0 air duct wind-blocking component 1”. The wind-blocking component number is marked on the wind-blocking component body and the identification sticker of the corresponding installation position inside the server case.

[0059] The light sensing cable of each wind shielding component is uniquely numbered, the light sensing cable number is consistent with the corresponding wind shielding component number, such as "Cable_01", corresponding to "CPU0 air duct wind shielding component 1", and the light sensing cable light state signal output interface is bound and stored to the non-volatile memory of the baseboard management controller.

[0060] A correlation mapping table of wind shielding component number and fan number is established, and the correlation relationship is determined based on the light range of the fan indicator light that the wind shielding component can receive. For example, for each wind shielding component, identify all the fans corresponding to its installation position (i.e. the fans that can be illuminated by the indicator light of the wind shielding component), record the numbers of these fans and the number of the wind shielding component in the mapping table, forming a "one wind shielding component - multiple fans" correlation relationship.

[0061] A one-to-one mapping table of wind shielding component number and light sensing cable number is established to determine the unique wind shielding component corresponding to each light sensing cable. The corresponding relationship is physically associated through the consistency of the cable number and the component number, and the logical association data is stored in the baseboard management controller.

[0062] The above two mapping tables are integrated into a "fan number - wind shielding component number - light sensing cable number" three-dimensional correlation database, which is stored in the storage module of the baseboard management controller, so that the baseboard management controller can query the other two correlation numbers through any one number.

[0063] The light state signal of the light sensing cable is used to indicate whether the light sensing cable detects light. Generally, when the light state signal is low, it indicates that the light sensing cable detects light, and it can be determined that the wind shielding component covering the light sensing cable has fallen off; on the contrary, when the light state signal is high, it indicates that the light sensing cable does not detect light, and it can be determined that the wind shielding component covering the light sensing cable has not fallen off.

[0064] In some optional embodiments, the high level of the light state signal can also be used to indicate that the light sensing cable detects light, and correspondingly, the low level can be used to indicate that the light sensing cable does not detect light. The actual application can be determined according to the needs, and the present application does not limit this.

[0065] Step S302: When the in-place state is a first state indicating that the wind shielding component has fallen off, if the error between the set speed value and the measured speed value of the target fan is greater than the preset error threshold, output a first abnormal detection result indicating that the wind shielding component has fallen into the target fan.

[0066] After the wind shielding component falls off, it may be stuck in the fan blades, affecting the rotation of the fan, or it may fall into the server case except the fan blades, blocking the air outlet or other ventilation gaps, affecting the airflow, and ultimately affecting the heat dissipation effect of the server.

[0067] In some specific embodiments, the baseboard management controller obtains the set speed value of the target fan by reading the PWM (Pulse Width Modulation) signal of the target fan, and obtains the measured speed value of the target fan by reading the Tach (Tachometer) signal fed back by the target fan, and determines whether there is an inconsistency between the set speed value and the measured speed value. Generally, due to factors such as voltage fluctuation or environmental interference, the set speed value and the measured speed value of the target fan are not completely equal, and the baseboard management controller can allow a certain error when making consistency determination, that is, the baseboard management controller compares the difference between the set speed value and the measured speed value with a preset error threshold.

[0068] If the error between the set speed value and the measured speed value is greater than the error threshold, it can be determined that the wind-blocking component has fallen into the target fan, causing the fan to rotate to be stuck or abnormal, and the baseboard management controller outputs a first abnormality detection result.

[0069] Generally, the longer the fan is used, the higher the aging degree of the fan, and the error between the set speed value and the measured speed value of the fan may increase due to the length of use of the fan. Therefore, the baseboard management controller can determine the error threshold according to the length of use of the target fan, wherein the length of use is positively correlated with the size of the error threshold.

[0070] The first abnormality detection result can include log information of the target fan, such as identification information of the target fan, time when the target fan first triggers a fault, fault event code, fault severity level, etc.

[0071] If there is more than one target fan, the baseboard management controller compares the set speed value and the measured speed value of each target fan respectively, and finally determines that the fan with an abnormality may also include multiple fans.

[0072] Step S303: When the in-place state is the first state representing that the wind-blocking component has fallen off, if the error between the set speed value and the measured speed value is not greater than the error threshold, and after the speed of the target fan is increased, the change value of the temperature in the region corresponding to the target fan after the first time length is less than a preset temperature drop threshold, a second abnormality detection result representing that the wind-blocking component has fallen into the server case is output.

[0073] In some specific embodiments, the wind blocking component falls off, but the error between the set speed value of the target fan and the measured speed value is less than the error threshold value, and it can be determined that the wind blocking component does not fall into the target fan. In order to further determine whether the wind blocking component falling off blocks the air outlet or other ventilation gaps and affects the airflow, the baseboard management controller increases the speed of the target fan and detects the temperature of the region corresponding to the target fan after a period of time (i.e., the first time length). Under normal circumstances, the temperature of the region corresponding to the target fan will decrease within a certain range, otherwise it indicates that the air duct inside the server is affected, i.e., the wind blocking component falling off affects the heat dissipation. At this time, the baseboard management controller outputs a second abnormality detection result, which feeds back that the server may have heat dissipation abnormalities due to the wind blocking component falling into the case.

[0074] For example, the baseboard management controller controls the target fan speed to increase by 10%, waits for 40s, and then reads the temperature of the region corresponding to the target fan. If the change value of the current temperature relative to the temperature before 40s is less than a preset temperature drop threshold value, the second abnormality detection result is output.

[0075] The region corresponding to the target fan can be a region extending 5 to 10 cm (the specific range depends on the fan and the internal structure of the server case) along the airflow direction from the fan air outlet as the starting point. A temperature sensor can be installed in the region to obtain the temperature of the region corresponding to the target fan. When the speed of the target fan is increased, the speed of the target fan after the increase is usually not more than a maximum speed threshold value, which can be 100% of the rated speed of the target fan, or any value between 80% and 100%, so as to avoid excessive fan speed and accelerate wear and tear, thereby reducing the service life.

[0076] By arranging the light sensing cable below the wind blocking component, using the indicator light of the fan, the light blocking property of the wind blocking component, and the photosensitive property of the light sensing cable, the logic of determining whether the wind blocking component is in place can be realized. Combined with the difference between the set speed value and the actual speed value of the fan, the specific influence of the wind blocking component falling off on the server heat dissipation can be determined, so as to quickly locate the abnormality of the server heat dissipation system caused by the wind blocking component falling off, reduce the detection cost of the server heat dissipation abnormality, and improve the maintenance efficiency of the heat dissipation system.

[0077] In some embodiments, as shown in Figure 4 When the hardware architecture shown in Figure 2 is adopted, step S301 can further include steps S3011, S3012 and S3013.

[0078] Step S3011: Poll the light state signal of each light sensing cable according to a preset first time interval until the light state signal of a target light sensing cable indicates that the light sensing cable detects light.

[0079] The target light sensing cable is one or more of the plurality of light sensing cables.

[0080] Specifically, the light illumination state signals (e.g., high level signals and low level signals) of the respective light sensing cables can be stored in registers of the programmable logic device, and the baseboard management controller reads the respective light illumination state signals stored in the registers every time interval (i.e., a first time interval) to determine whether the respective light sensing cables detect light illumination.

[0081] For example, the respective light illumination state signals are polled every 10 seconds, and when one or more of the light illumination state signals represent that the light sensing cable detects light illumination, the subsequent steps are performed.

[0082] Step S3012: The light illumination state signal of the target light sensing cable is repeatedly obtained N times according to a preset second time interval.

[0083] N is a positive integer.

[0084] Step S3013: If the light illumination state signals obtained N times are the same, it is determined that the in-place state of the windshield bubble cotton corresponding to the target light sensing cable is a first state.

[0085] The first time interval and the second time interval are only used to distinguish different time parameters in the two steps, and the two time intervals can be the same or different, which is not limited in the present application.

[0086] In some specific embodiments, the light illumination state signal may be false due to electromagnetic interference or mechanical vibration interference during transmission. In order to prevent false judgment of the falling of the windshield component, the light illumination state signal of the light sensing cable can be repeatedly obtained every time interval (i.e., a second time interval) after the light illumination state signal representing that the light sensing cable detects light illumination is first found, so as to perform repeated verification and avoid wasting time and maintenance costs due to a false alarm.

[0087] For example, the light illumination state signal is read every 1 second, and if the light illumination state signals of the light sensing cable represent that the light sensing cable detects light illumination for 3 consecutive times, it is determined that the in-place state of the windshield component is a first state representing that the windshield component has fallen.

[0088] In some optional embodiments, the light illumination state signal can also be repeatedly obtained every time interval within a preset third time interval, and if the light illumination state signals read for N times in total represent that the light sensing cable detects light illumination, it is determined that the in-place state of the windshield component is a first state representing that the windshield component has fallen.

[0089] For example, the light state signal is read every 1s within 5s, and if 3 of the 5 readings indicate that the light sensing cable detects light, it is determined that the in-place state of the wind shielding component is the first state indicating that the wind shielding component has fallen off.

[0090] By repeatedly reading the light state signal, the waste of maintenance cost caused by false positives due to transient interference can be reduced at a lower time cost.

[0091] The hardware architecture of the server heat dissipation anomaly detection system can further include an interface expander connected to the plurality of light sensing cables, configured to receive the light state signals transmitted by the plurality of light sensing cables, and when the light state signal indicates that the target light sensing cable detects light, send an interrupt signal to the signal connector through the I2C interface, so that the signal connector transmits the interrupt signal to the programmable logic device through the I2C interface, triggering the baseboard management controller to read the light state signal of the target light sensing cable.

[0092] As shown in Figure 5 , the interface expander can be an IO Expander (Input Output Expander), mainly used to increase the number of input and output ports of the system, thereby connecting more light sensing cables. Moreover, the IO Expander can convert the I2C bus protocol to the GPIO interface, which is used to expand the function of the I2C bus.

[0093] Specifically, each light sensing cable is connected to a port of the IO Expander and transmits the light state signal to the FPGA / CPLD through the CONN. In the IO Expander, an interrupt signal can be triggered when the light state signal is at a low level (indicating that light is detected), and the interrupt signal (IRQ_GPIO_I2C) is transmitted to the FPGA / CPLD, which is then transmitted to the BMC, so that the BMC actively reads the light state signal, avoiding the problem of slow anomaly detection caused by too long polling time.

[0094] For example, the BMC can repeatedly obtain the light state signal of the target light sensing cable N times according to the preset second time interval in response to the interrupt signal sent by the FPGA / CPLD, and if the N times of obtained light state signals are the same, it is determined that the in-place state of the wind shielding component corresponding to the target light sensing cable is the first state.

[0095] Since the IO Expander can expand the I2C bus with strong anti-interference, the light state signal of the light sensing cable is transmitted to the FPGA / CPLD through the I2C interface, so when there is strong electromagnetic interference or vibration interference in the connection path of the CONN, or when the number of light sensing cables is large, the hardware architecture shown in Figure 4 can be selected.

[0096] In addition, the GPIO in the FPGA / CPLD can be set to a Schmitt trigger input mode to avoid logic judgment errors caused by unstable impedance changes of the photoresistor when there is no light to light.

[0097] In some embodiments, the baseboard management controller can be further configured to perform the following steps: if the light state signal of the light sensing cable represents that the light sensing cable does not detect light, determining that the in-place state is a second state representing that the wind shielding component is located at the initial installation position.

[0098] In some embodiments, the baseboard management controller can be further configured to perform the following steps: when the in-place state is the first state representing that the wind shielding component has fallen off, if the error between the set rotation speed value and the measured rotation speed value is not greater than the error threshold value, and after the rotation speed of the target fan is controlled, the temperature change value of the target fan corresponding area after the first time length is not less than the temperature drop threshold value, outputting a third abnormality detection result representing light sensing abnormality.

[0099] Specifically, after the rotation speed of the target fan is increased, the temperature change value of the target fan corresponding area meets the requirement of the ideal state, i.e., not less than the temperature drop threshold value, which means that the falling of the wind shielding component does not affect the rotation of the target fan and does not affect the passage of airflow. At this time, it can be considered that the brightness of the indicator light of the target fan is too strong, or the sensing function of the light sensing cable fails, causing the light state signal to represent that light has been detected. The baseboard management controller can output a third abnormality detection result representing light sensing abnormality to prompt maintenance personnel to check the brightness of the indicator light or maintain the function of the light sensing cable.

[0100] In some embodiments, the baseboard management controller can be further configured to perform the following steps: before the rotation speed of the target fan is increased, if the measured rotation speed value of the target fan is not less than the maximum rotation speed threshold value, maintaining the measured rotation speed value, and after a second time length, determining that the temperature of the target fan corresponding area is greater than a preset temperature threshold value, outputting a second abnormality detection result representing that the wind shielding component has fallen into the server case; wherein the second time length is greater than the first time length.

[0101] Specifically, the measured rotation speed of the target fan can have reached the maximum rotation speed threshold value. At this time, the waiting time can be extended, for example, after 60s to 80s, the temperature of the target fan corresponding area is detected. If the temperature is greater than the preset temperature threshold value, it can be determined that the server cooling is abnormal due to the falling of the wind shielding component, and the second abnormality detection result is outputted.

[0102] It should be understood that, although Figure 3 and Figure 4 the flowcharts in the respective steps are displayed in the order indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Figure 3 andFigure 4 The steps shown and steps involved in other embodiments need not be performed in the order shown unless explicitly stated herein. Moreover, at least some of the steps of the aforementioned embodiments can include multiple sub-steps or stages, which need not necessarily be performed at the same time, but can be performed at different times, and need not necessarily be performed in the order shown, but can be performed in a different order, or in an alternating or interleaved order, with other steps or sub-steps or stages of other steps.

[0103] In a second aspect, embodiments of the present application provide a server comprising the server abnormal heat dissipation detection system according to any of the embodiments of the first aspect.

[0104] It is understood by a person of ordinary skill in the art that all or part of the processes in the above embodiments can be implemented by a computer program instructing relevant hardware, and the aforementioned computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above embodiments. Any reference to memory, storage, database or other medium used in the embodiments of the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct RAM bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0105] The technical features of the above embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not contradict, they should be considered within the scope of the present application.

[0106] The above embodiments only express some implementation manners of the present application, which are described in a more specific and detailed manner, but should not be understood as a limitation to the protection scope of the present application. It should be noted that, for those of ordinary skill in the art, some modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application.

Claims

1. A server heat dissipation anomaly detection system, characterized in that, include: Multiple optical sensing cables are used to determine the position status of the windproof component corresponding to each optical sensing cable; wherein, the windproof component is initially installed within the illumination range of the fan's indicator light and provides light blocking for the photosensitive section of the optical sensing cable; A signal connector is used to receive the illumination status signal transmitted by the plurality of optical sensing cables; A programmable logic device for receiving and storing the illumination status signals of the plurality of optical sensing cables sent by the signal connector; A substrate management controller is used to access the programmable logic device via an I2C interface and perform the following detection method: Based on the illumination status signals of each optical sensor cable, determine the position status of the windproof component corresponding to each optical sensor cable; When the in-situ state is the first state indicating that the windshield component has fallen off, if the error between the set speed value and the measured speed value of the target fan is greater than the preset error threshold, then the first abnormal detection result indicating that the windshield component has fallen into the target fan is output. When the in-situ state is the first state indicating that the windshield component has fallen off, if the error between the set rotation speed value and the measured rotation speed value is not greater than the error threshold, and after controlling and increasing the rotation speed of the target fan, the temperature change value of the area corresponding to the target fan after the first time period is less than the preset cooling threshold, then the second abnormal detection result indicating that the windshield component has fallen into the server chassis is output.

2. The system according to claim 1, characterized in that, The substrate management controller is also used for: According to a preset first time interval, the illumination status signal of each optical sensor cable is polled until the illumination status signal of a target optical sensor cable indicates that the optical sensor cable has detected illumination; the target optical sensor cable is one or more of the plurality of optical sensor cables. The illumination status signal of the target optical cable is repeatedly acquired N times according to a preset second time interval; where N is a positive integer. If the illumination status signals obtained N times are the same, then the in-situ state of the windproof component corresponding to the target light-sensing cable is determined to be the first state.

3. The system according to claim 2, characterized in that, The substrate management controller is also used for: If the illumination status signal of the target optical sensing cable indicates that the target optical sensing cable has not detected illumination, then the in-situ state is determined to be the second state indicating that the windbreak component is in the initial installation position.

4. The system according to claim 1, characterized in that, The system also includes: An interface extender connected to the plurality of optical sensing cables is used to receive the illumination status signals transmitted by the plurality of optical sensing cables, and when there is an illumination status signal indicating that a target optical sensing cable has detected illumination, it sends an interrupt signal to the signal connector through the I2C interface, so that the signal connector transmits the interrupt signal to the programmable logic device through the I2C interface, triggering the substrate management controller to read the illumination status signal of the target optical sensing cable.

5. The system according to claim 4, characterized in that, The substrate management controller is also used for: In response to the interrupt signal sent by the programmable logic device, the illumination status signal of the target optical cable is repeatedly acquired N times according to a preset second time interval; where N is a positive integer. If the illumination status signals obtained N times are the same, then the in-situ state of the windproof component corresponding to the target light-sensing cable is determined to be the first state.

6. The system according to any one of claims 1 to 5, characterized in that, The substrate management controller is also used for: When the in-situ state is the first state indicating that the windshield component has fallen off, if the error between the set rotation speed value and the measured rotation speed value is not greater than the error threshold, and after controlling and increasing the rotation speed of the target fan, the temperature change value of the area corresponding to the target fan after the first time period is not less than the cooling threshold, then the third abnormality detection result indicating light sensitivity abnormality is output.

7. The system according to any one of claims 1 to 5, characterized in that, The substrate management controller is also used for: Before increasing the speed of the target fan, if the measured speed value of the target fan is not less than the preset maximum speed threshold, the measured speed value is maintained, and after a second duration, if it is determined that the temperature of the area corresponding to the target fan is greater than the preset temperature threshold, a second abnormal detection result indicating that the windshield component has fallen into the server chassis is output; wherein, the second duration is longer than the first duration.

8. The system according to any one of claims 1 to 5, characterized in that, The substrate management controller is also used for: The error threshold is determined based on the usage time of the target fan; wherein the usage time is positively correlated with the magnitude of the error threshold.

9. The system according to any one of claims 1 to 5, characterized in that, The substrate management controller is also used for: When the in-situ state is a first state indicating that the windshield component has fallen off, at least one target fan is identified based on the fan number associated with the windshield component.

10. A server, characterized in that, Including the server heat dissipation anomaly detection system as described in any one of claims 1 to 9.

Images