CN111352815A - Performance balance detection method, system and device of server system - Google Patents
Performance balance detection method, system and device of server system Download PDFInfo
- Publication number
- CN111352815A CN111352815A CN202010120841.1A CN202010120841A CN111352815A CN 111352815 A CN111352815 A CN 111352815A CN 202010120841 A CN202010120841 A CN 202010120841A CN 111352815 A CN111352815 A CN 111352815A
- Authority
- CN
- China
- Prior art keywords
- target
- hardware
- server system
- working
- power supply
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 44
- 230000015654 memory Effects 0.000 claims description 42
- 238000000034 method Methods 0.000 claims description 28
- 230000008569 process Effects 0.000 claims description 18
- 238000004590 computer program Methods 0.000 claims description 6
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 claims description 4
- 229910052802 copper Inorganic materials 0.000 claims description 4
- 239000010949 copper Substances 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011895 specific detection Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3409—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/505—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the load
Abstract
The invention discloses a performance balance detection method, a system and a device of a server system, which respectively obtain working parameters of hardware in the same type and in a working state in the server system; calculating the average parameter value of the same working parameter corresponding to each hardware, and respectively calculating the difference value between the target working parameter of each hardware and the average parameter value corresponding to each hardware; judging whether the absolute value of the difference value corresponding to the target working parameter of the target hardware is larger than a preset equalization threshold value or not; if so, determining that the performance of the target hardware is unbalanced under the target working parameters; and if not, determining the performance balance of the target hardware under the target working parameters. Therefore, the performance balance problem of each hardware can be analyzed according to the difference degree between the working parameter of each hardware and the average working parameter of the same type of hardware, and therefore more valuable reference information can be provided for system stability.
Description
Technical Field
The present invention relates to the field of servers, and in particular, to a method, a system, and an apparatus for detecting performance balance of a server system.
Background
With the development of computer technology, server systems are more and more widely used. At present, in order to ensure the stability of the performance of each hardware in the server system, the working parameters of each hardware are usually detected in the working process of the server system, and the specific detection means is as follows: judging whether the working parameters of the hardware are within a preset range, and if so, determining that the performance of the hardware is stable; if not, determining that the hardware performance is unstable. However, the performance detection method can only reflect the performance of a single hardware, and cannot reflect the performance balance problem of the same type of hardware in the server system, so that more valuable reference information cannot be provided for the system stability.
Therefore, how to provide a solution to the above technical problem is a problem that needs to be solved by those skilled in the art.
Disclosure of Invention
The invention aims to provide a performance balance detection method, a system and a device of a server system, which can analyze the performance balance problem of each hardware according to the difference degree between the working parameter of each hardware and the average working parameter of the same type of hardware, thereby providing more valuable reference information for the stability of the system.
In order to solve the above technical problem, the present invention provides a performance balance detection method for a server system, including:
respectively acquiring working parameters of hardware in the same type and in a working state in a server system;
calculating the average parameter value of the same working parameter corresponding to each hardware, and calculating the difference value between the target working parameter of each hardware and the average parameter value corresponding to the target working parameter; wherein the target working parameter is any one of the working parameters;
judging whether the absolute value of the difference value corresponding to the target working parameter of the target hardware is larger than a preset equalization threshold value or not; wherein the target hardware is any one of the hardware;
if so, determining that the performance of the target hardware is unbalanced under the target working parameters;
and if not, determining the performance balance of the target hardware under the target working parameters.
Preferably, the server system comprises PSU power supplies for supplying power to the server nodes one by one through copper bars;
correspondingly, the process of respectively obtaining the working parameters of each hardware of the same type and in the working state in the server system includes:
and respectively acquiring the power supply voltage and the power supply current of each PSU power supply through the RMC board card in the server system.
Preferably, the server system further includes a voltage and current adjusting device respectively disposed on a power supply line of each PSU power supply;
and the performance balance detection method further comprises the following steps:
after the target PSU power supply is determined to be not voltage-sharing or current-sharing, the supply voltage or the supply current of the target PSU power supply is correspondingly adjusted by controlling a voltage and current adjusting device on a power supply line of the target PSU power supply, so that the target PSU power supply is kept in a voltage-sharing and current-sharing state.
Preferably, the server system comprises a plurality of symmetrically distributed memories;
correspondingly, the process of respectively obtaining the working parameters of each hardware of the same type and in the working state in the server system includes:
and respectively acquiring the utilization rate of each memory.
Preferably, the performance balance detection method further includes:
after the imbalance of the utilization rate of the target memory is determined, the utilization rate of the target memory is adjusted by modifying the memory use allocation strategy of the server system, so that the target memory is kept in a utilization rate balance state.
Preferably, the server system comprises a plurality of GPUs;
correspondingly, the process of respectively obtaining the working parameters of each hardware of the same type and in the working state in the server system includes:
and respectively acquiring the work load of each GPU.
Preferably, the performance balance detection method further includes:
after determining that the workload of the target GPU is unbalanced, adjusting the workload of the target GPU by modifying a GPU task allocation strategy of the server system so as to keep the target GPU in a workload balancing state.
Preferably, the performance balance detection method further includes:
numbering each hardware in advance;
and after the performance imbalance of the target hardware is determined, uploading performance imbalance information containing the serial number of the target hardware to an upper computer.
In order to solve the above technical problem, the present invention further provides a performance balance detection system for a server system, including:
the acquisition module is used for respectively acquiring the working parameters of each hardware in the same type and in a working state in the server system;
the calculation module is used for calculating the average parameter value of the same working parameter corresponding to each hardware and calculating the difference value between the target working parameter of each hardware and the average parameter value corresponding to the target working parameter; wherein the target working parameter is any one of the working parameters;
the judging module is used for judging whether the absolute value of the difference value corresponding to the target working parameter of the target hardware is larger than a preset balance threshold value or not; wherein the target hardware is any one of the hardware; if yes, executing an imbalance determining module; if not, executing a balance determining module;
an imbalance determination module, configured to determine that performance of the target hardware is imbalanced under the target operating parameter;
and the balance determining module is used for determining the performance balance of the target hardware under the target working parameters.
In order to solve the above technical problem, the present invention further provides a performance balance detection apparatus for a server system, including:
a memory for storing a computer program;
and the processor is used for implementing the steps of the performance balance detection method of any one of the server systems when executing the computer program.
The invention provides a performance balance detection method of a server system, which respectively acquires working parameters of hardware in the same type and in a working state in the server system; calculating the average parameter value of the same working parameter corresponding to each hardware, and respectively calculating the difference value between the target working parameter of each hardware and the average parameter value corresponding to each hardware; judging whether the absolute value of the difference value corresponding to the target working parameter of the target hardware is larger than a preset equalization threshold value or not; if so, determining that the performance of the target hardware is unbalanced under the target working parameters; and if not, determining the performance balance of the target hardware under the target working parameters. Therefore, the performance balance problem of each hardware can be analyzed according to the difference degree between the working parameter of each hardware and the average working parameter of the same type of hardware, and therefore more valuable reference information can be provided for system stability.
The invention also provides a performance balance detection system and a device of the server system, and the system and the device have the same beneficial effects as the performance balance detection method.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed in the prior art and the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a flowchart of a performance balance detection method for a server system according to an embodiment of the present invention;
fig. 2 is a schematic power supply diagram of a server system according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a performance balance detection system of a server system according to an embodiment of the present invention.
Detailed Description
The core of the invention is to provide a method, a system and a device for detecting the performance balance of a server system, which can analyze the performance balance problem of each hardware according to the difference degree between the working parameter of each hardware and the average working parameter of the same type of hardware, thereby providing more valuable reference information for the stability of the system.
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, fig. 1 is a flowchart illustrating a performance balance detection method of a server system according to an embodiment of the present invention.
The performance balance detection method of the server system comprises the following steps:
step S1: and respectively acquiring the working parameters of the hardware in the same type and in the working state in the server system.
Specifically, the server system includes a plurality of pieces of hardware with the same structure and the same function, where the pieces of hardware belong to the same type of hardware, for example, a plurality of memories belong to the same type of hardware, and a plurality of hard disks belong to the same type of hardware. For the same type of hardware, the working parameters of the hardware in the same type and in a working state are respectively obtained.
It should be noted that, the present application may obtain the operating parameters of each hardware of multiple types and in an operating state at the same time, and then classify the operating parameters of the hardware of the same type together.
Step S2: and calculating the average parameter value of the same working parameter corresponding to each hardware, and respectively calculating the difference value between the target working parameter of each hardware and the average parameter value corresponding to each hardware.
It should be noted that the target operating parameter in step S2 is any one of the operating parameters of the same type of hardware mentioned in step S1.
Specifically, for a certain type of hardware, there may be more than one main working parameter, so for the working parameters of each hardware of the same type, the application first superimposes the same working parameter corresponding to each hardware to obtain a total parameter value of the same working parameter, and then divides the total parameter value of the same working parameter by the total amount of the corresponding hardware to obtain an average parameter value of the same working parameter.
Based on the above, for each hardware of the same type, the target working parameter of each hardware is differentiated from the corresponding average parameter value to obtain the difference value between the target working parameter of each hardware and the corresponding average parameter value, so as to determine whether the performance of the hardware is balanced subsequently.
Step S3: judging whether the absolute value of the difference value corresponding to the target working parameter of the target hardware is larger than a preset equalization threshold value or not; if yes, go to step S4; if not, step S5 is executed. Step S4: and determining the performance imbalance of the target hardware under the target working parameters. Step S5: and determining the performance balance of the target hardware under the target working parameters.
It should be noted that the target hardware in step S3 is any one of the same types of hardware mentioned in step S1.
Specifically, it can be understood that, if the absolute value of the difference value corresponding to the target operating parameter of the target hardware is greater than the preset equalization threshold, it indicates that the difference between the target operating parameter of the target hardware and the average parameter value corresponding to the target operating parameter is relatively large, that is, the performance of the target hardware under the target operating parameter is unbalanced; if the absolute value of the difference value corresponding to the target working parameter of the target hardware is not greater than the preset balance threshold, it is indicated that the difference degree between the target working parameter of the target hardware and the average parameter value corresponding to the target working parameter is smaller, that is, the performance of the target hardware under the target working parameter is balanced.
Based on the above, the method and the device judge whether the absolute value of the difference value corresponding to the target working parameter of the target hardware is larger than a preset balance threshold value, and if the absolute value is larger than the preset balance threshold value, the performance imbalance of the target hardware under the target working parameter is determined; and if the performance of the target hardware is not greater than the preset balance threshold, determining the performance balance of the target hardware under the target working parameters.
The invention provides a performance balance detection method of a server system, which respectively acquires working parameters of hardware in the same type and in a working state in the server system; calculating the average parameter value of the same working parameter corresponding to each hardware, and respectively calculating the difference value between the target working parameter of each hardware and the average parameter value corresponding to each hardware; judging whether the absolute value of the difference value corresponding to the target working parameter of the target hardware is larger than a preset equalization threshold value or not; if so, determining that the performance of the target hardware is unbalanced under the target working parameters; and if not, determining the performance balance of the target hardware under the target working parameters. Therefore, the performance balance problem of each hardware can be analyzed according to the difference degree between the working parameter of each hardware and the average working parameter of the same type of hardware, and therefore more valuable reference information can be provided for system stability.
On the basis of the above-described embodiment:
as an alternative embodiment, the server system comprises PSU power supplies for supplying power to each server node one by one through copper bars;
correspondingly, the process of respectively obtaining the working parameters of the hardware in the same type and in the working state in the server system comprises the following steps:
and respectively acquiring the power supply voltage and the power supply current of each PSU power supply through an RMC board card in the server system.
Specifically, referring to fig. 2, fig. 2 is a schematic power supply diagram of a server system according to an embodiment of the present invention. The server system is specifically a RACK server system, in which PSU (power supply Unit) power supplies for each server node one by one are inserted onto power shelf, and the power shelf is directly connected with the copper bar and then supplies power to each server node. The RMC (Rack Management Controller, centralized monitoring Management unit) board card is also inserted into the power shelf, and the RMC board card is used for collecting information of all server nodes in the server cabinet, controlling the on-off of all the server nodes, controlling the rotating speed of a fan in the cabinet, detecting whether the state of a PSU power supply is normal or not and the like.
Based on this, for a plurality of PSU power supplies (belonging to the same type of hardware), the application can obtain the supply voltage and the supply current of each PSU power supply through the RMC board in the server system. Specifically, the voltage-sharing judgment process of the PSU power supply comprises the following steps: superposing the power supply voltage of each PSU power supply to obtain total power supply voltage, and dividing the total power supply voltage by the total number of each PSU power supply to obtain average power supply voltage; respectively subtracting the power supply voltage of each PSU power supply from the average power supply voltage to obtain the voltage difference value of each PSU power supply; judging whether the absolute value of the voltage difference value of any PSU power supply is greater than a preset voltage balance threshold value or not, and if so, determining that the PSU power supply is not voltage-sharing; and if the voltage is not greater than the preset voltage balance threshold value, determining the PSU power voltage balance.
Similarly, the current-sharing judgment process of the PSU power supply comprises the following steps: superposing the power supply current of each PSU power supply to obtain total power supply current, and then dividing the total power supply current by the total number of each PSU power supply to obtain average power supply current; respectively subtracting the power supply current of each PSU power supply from the average power supply current to obtain the current difference value of each PSU power supply; judging whether the absolute value of the current difference value of any PSU power supply is greater than a preset current balance threshold value or not, and if so, determining that the PSU power supply is not uniform; and if the current is not greater than the preset current balance threshold value, determining the current balance of the PSU power supply.
As an optional embodiment, the server system further includes a voltage and current adjusting device respectively disposed on the power supply line of each PSU power supply;
and the performance balance detection method also comprises the following steps:
after the target PSU power supply is determined to be not voltage-sharing or current-sharing, the supply voltage or the supply current of the target PSU power supply is correspondingly adjusted by controlling a voltage and current adjusting device on a power supply line of the target PSU power supply, so that the target PSU power supply is kept in a voltage-sharing and current-sharing state.
Furthermore, the server system of the application is provided with a voltage and current adjusting device on a power supply line of each PSU power supply respectively, so that the power supply voltage and the power supply current of the PSU power supply can be adjusted. Based on this, after determining that the target PSU power supply (any PSU power supply) has uneven voltage, the application can adjust the power supply voltage of the target PSU power supply by controlling the voltage and current adjusting device on the power supply line of the target PSU power supply, so that the target PSU power supply is kept in a voltage-sharing state; similarly, after the target PSU power supply is determined to be non-current-sharing, the power supply current of the target PSU power supply is adjusted by controlling a voltage and current adjusting device on a power supply line of the target PSU power supply, so that the target PSU power supply is kept in a current-sharing state.
As an alternative embodiment, the server system includes a plurality of symmetrically distributed memories;
correspondingly, the process of respectively obtaining the working parameters of the hardware in the same type and in the working state in the server system comprises the following steps:
and respectively acquiring the utilization rate of each memory.
Specifically, the server system of the present application further includes a plurality of symmetrically distributed memories (symmetric distribution is favorable for balancing the usage rates of the memories). For a plurality of memories (belonging to the same type of hardware), the utilization rate of each memory can be respectively obtained. Specifically, the judgment process of the memory usage balance includes: overlapping the utilization rates of the memories to obtain a total utilization rate, and then dividing the total utilization rate by the total amount of the memories to obtain an average utilization rate; respectively differentiating the utilization rate of each memory with the average utilization rate to obtain the utilization rate difference value of each memory; judging whether the absolute value of the utilization rate difference value of any memory is greater than a preset utilization rate balance threshold, and if the absolute value of the utilization rate difference value of any memory is greater than the preset utilization rate balance threshold, determining that the utilization rates of the memories are unbalanced; and if the utilization rate is not greater than the preset utilization rate balance threshold value, determining that the utilization rate of the memory is balanced.
As an optional embodiment, the performance balance detection method further includes:
after the imbalance of the utilization rate of the target memory is determined, the utilization rate of the target memory is adjusted by modifying the memory use allocation strategy of the server system, so that the target memory is kept in a utilization rate balance state.
Further, after determining that the usage rate of the target memory (any memory) is unbalanced, the present application may adjust the usage rate of the target memory by modifying the memory usage allocation policy of the server system (adjusting the usage rate of the memory with the usage rate higher than the average usage rate to the usage rate of the memory with the usage rate lower than the average usage rate), so as to maintain the target memory in a usage rate balanced state.
As an alternative embodiment, the server system comprises a plurality of GPUs;
correspondingly, the process of respectively obtaining the working parameters of the hardware in the same type and in the working state in the server system comprises the following steps:
and respectively acquiring the work load of each GPU.
Specifically, the server system of the present application further includes a plurality of GPUs (Graphics Processing units). For multiple GPUs (belonging to the same type of hardware), the workload of each GPU can be obtained separately. Specifically, the process for determining the workload balance of the GPU includes: overlapping the working loads of the GPUs to obtain a total working load, and then dividing the total working load by the total quantity of the GPUs to obtain an average working load; respectively subtracting the working load of each GPU from the average working load to obtain the load difference value of each GPU; judging whether the absolute value of the load difference value of any GPU is larger than a preset load balancing threshold value or not, and if so, determining that the working load of the GPU is unbalanced; and if the load is not greater than the preset load balancing threshold value, determining the work load balance of the GPU.
As an optional embodiment, the performance balance detection method further includes:
after determining that the workload of the target GPU is unbalanced, adjusting the workload of the target GPU by modifying a GPU task allocation strategy of the server system so as to keep the target GPU in a workload balancing state.
Further, after determining that the workload of the target GPU (any GPU) is unbalanced, the GPU task scheduling policy of the server system may be modified (allocating tasks of GPUs with workloads higher than the average workload to GPUs with workloads lower than the average workload for execution) to adjust the workload of the target GPU so that the target GPU is maintained in a workload balancing state.
In addition, the server system of the application further comprises a plurality of hard disks. For a plurality of hard disks (belonging to the same type of hardware), the present application can respectively obtain IO (Input/Output) throughput of each hard disk. Specifically, the determining process of IO throughput balancing of the hard disk includes: superposing the IO throughput of each hard disk to obtain total IO throughput, and dividing the total IO throughput by the total number of each hard disk to obtain average IO throughput; respectively subtracting the IO throughput of each hard disk from the average IO throughput to obtain the throughput difference of each hard disk; judging whether the absolute value of the throughput difference value of any hard disk is larger than a preset throughput balance threshold, and if the absolute value of the throughput difference value of any hard disk is larger than the preset throughput balance threshold, determining that the IO throughput of the hard disk is unbalanced; and if the sum of the IO throughput of the hard disk is not greater than the preset throughput balance threshold, determining that the IO throughput of the hard disk is balanced.
As an optional embodiment, the performance balance detection method further includes:
numbering each hardware in advance;
and after the performance imbalance of the target hardware is determined, uploading performance imbalance information containing the serial number of the target hardware to an upper computer.
Furthermore, the hardware can be numbered in advance, and the purpose is to upload the performance imbalance information containing the serial number of the target hardware to an upper computer after determining that the performance of the target hardware is imbalanced, so that a user can check the performance imbalance information, and the user can know the position of the target hardware with imbalanced performance.
Referring to fig. 3, fig. 3 is a schematic structural diagram of a performance balance detection system of a server system according to an embodiment of the present invention.
The performance balance detection system of the server system comprises:
the acquisition module 1 is used for respectively acquiring working parameters of hardware of the same type and in a working state in the server system;
the calculating module 2 is used for calculating average parameter values of the same working parameter corresponding to each hardware, and calculating difference values between target working parameters of each hardware and the average parameter values corresponding to the target working parameters; wherein the target working parameter is any one working parameter;
the judging module 3 is used for judging whether the absolute value of the difference value corresponding to the target working parameter of the target hardware is larger than a preset balance threshold value or not; the target hardware is any hardware; if yes, executing an imbalance determining module 4; if not, executing the balance determining module 5;
the imbalance determining module 4 is used for determining the performance imbalance of the target hardware under the target working parameters;
and the balance determining module 5 is used for determining the performance balance of the target hardware under the target working parameters.
For introduction of the performance balance detection system provided in the present application, reference is made to the embodiments of the performance balance detection method described above, and details of the performance balance detection method are not described herein again.
The present application further provides a performance balance detection apparatus for a server system, including:
a memory for storing a computer program;
and the processor is used for implementing the steps of the performance balance detection method of any one of the server systems when executing the computer program.
For introduction of the performance balance detection apparatus provided in the present application, reference is made to the embodiments of the performance balance detection method described above, and details of the performance balance detection method are not described herein again.
It is further noted that, in the present specification, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A performance balance detection method for a server system is characterized by comprising the following steps:
respectively acquiring working parameters of hardware in the same type and in a working state in a server system;
calculating the average parameter value of the same working parameter corresponding to each hardware, and calculating the difference value between the target working parameter of each hardware and the average parameter value corresponding to the target working parameter; wherein the target working parameter is any one of the working parameters;
judging whether the absolute value of the difference value corresponding to the target working parameter of the target hardware is larger than a preset equalization threshold value or not; wherein the target hardware is any one of the hardware;
if so, determining that the performance of the target hardware is unbalanced under the target working parameters;
and if not, determining the performance balance of the target hardware under the target working parameters.
2. The method according to claim 1, wherein the server system includes PSU power supplies for supplying power to the server nodes one by one via copper bars;
correspondingly, the process of respectively obtaining the working parameters of each hardware of the same type and in the working state in the server system includes:
and respectively acquiring the power supply voltage and the power supply current of each PSU power supply through the RMC board card in the server system.
3. The method according to claim 2, wherein the server system further comprises a voltage/current regulator respectively provided on a power supply line of each PSU power supply;
and the performance balance detection method further comprises the following steps:
after the target PSU power supply is determined to be not voltage-sharing or current-sharing, the supply voltage or the supply current of the target PSU power supply is correspondingly adjusted by controlling a voltage and current adjusting device on a power supply line of the target PSU power supply, so that the target PSU power supply is kept in a voltage-sharing and current-sharing state.
4. The method according to claim 1, wherein the server system comprises a plurality of symmetrically distributed memories;
correspondingly, the process of respectively obtaining the working parameters of each hardware of the same type and in the working state in the server system includes:
and respectively acquiring the utilization rate of each memory.
5. The performance balance detection method of the server system as claimed in claim 4, wherein said performance balance detection method further comprises:
after the imbalance of the utilization rate of the target memory is determined, the utilization rate of the target memory is adjusted by modifying the memory use allocation strategy of the server system, so that the target memory is kept in a utilization rate balance state.
6. The method for detecting performance balance of a server system according to claim 1, wherein the server system includes a plurality of GPUs;
correspondingly, the process of respectively obtaining the working parameters of each hardware of the same type and in the working state in the server system includes:
and respectively acquiring the work load of each GPU.
7. The performance balance detection method of a server system according to claim 6, wherein the performance balance detection method further comprises:
after determining that the workload of the target GPU is unbalanced, adjusting the workload of the target GPU by modifying a GPU task allocation strategy of the server system so as to keep the target GPU in a workload balancing state.
8. The performance balance detection method of the server system according to any of claims 1 to 7, wherein the performance balance detection method further comprises:
numbering each hardware in advance;
and after the performance imbalance of the target hardware is determined, uploading performance imbalance information containing the serial number of the target hardware to an upper computer.
9. A system for detecting performance balance of a server system, comprising:
the acquisition module is used for respectively acquiring the working parameters of each hardware in the same type and in a working state in the server system;
the calculation module is used for calculating the average parameter value of the same working parameter corresponding to each hardware and calculating the difference value between the target working parameter of each hardware and the average parameter value corresponding to the target working parameter; wherein the target working parameter is any one of the working parameters;
the judging module is used for judging whether the absolute value of the difference value corresponding to the target working parameter of the target hardware is larger than a preset balance threshold value or not; wherein the target hardware is any one of the hardware; if yes, executing an imbalance determining module; if not, executing a balance determining module;
an imbalance determination module, configured to determine that performance of the target hardware is imbalanced under the target operating parameter;
and the balance determining module is used for determining the performance balance of the target hardware under the target working parameters.
10. A performance balance detecting apparatus for a server system, comprising:
a memory for storing a computer program;
processor for implementing the steps of the performance balancing detection method of the server system according to any of the claims 1-8 when executing said computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010120841.1A CN111352815A (en) | 2020-02-26 | 2020-02-26 | Performance balance detection method, system and device of server system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010120841.1A CN111352815A (en) | 2020-02-26 | 2020-02-26 | Performance balance detection method, system and device of server system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111352815A true CN111352815A (en) | 2020-06-30 |
Family
ID=71192478
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010120841.1A Withdrawn CN111352815A (en) | 2020-02-26 | 2020-02-26 | Performance balance detection method, system and device of server system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111352815A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112114647A (en) * | 2020-10-29 | 2020-12-22 | 苏州浪潮智能科技有限公司 | Power supply control method, system and device of server |
| WO2022257304A1 (en) * | 2021-06-09 | 2022-12-15 | 苏州浪潮智能科技有限公司 | Server tuning method, system and apparatus |
-
2020
- 2020-02-26 CN CN202010120841.1A patent/CN111352815A/en not_active Withdrawn
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112114647A (en) * | 2020-10-29 | 2020-12-22 | 苏州浪潮智能科技有限公司 | Power supply control method, system and device of server |
| CN112114647B (en) * | 2020-10-29 | 2022-06-10 | 苏州浪潮智能科技有限公司 | Power supply control method, system and device of server |
| WO2022257304A1 (en) * | 2021-06-09 | 2022-12-15 | 苏州浪潮智能科技有限公司 | Server tuning method, system and apparatus |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhu et al. | A three-dimensional virtual resource scheduling method for energy saving in cloud computing | |
| CN109800066B (en) | Energy-saving scheduling method and system for data center | |
| Cardosa et al. | Shares and utilities based power consolidation in virtualized server environments | |
| EP2417534B1 (en) | Optimized virtual machine migration mechanism | |
| Andreolini et al. | Dynamic load management of virtual machines in cloud architectures | |
| CN107124472A (en) | Load-balancing method and device, computer-readable recording medium | |
| CN103095599A (en) | Dynamic feedback weighted integration load scheduling method of cloud computing operating system | |
| Al-Shishtawy et al. | Elastman: elasticity manager for elastic key-value stores in the cloud | |
| US20150058844A1 (en) | Virtual computing resource orchestration | |
| CN107346264A (en) | A kind of method, apparatus and server apparatus of virtual machine load balance scheduling | |
| CN111147565B (en) | Cluster node control method, device and equipment and readable storage medium | |
| CN111352815A (en) | Performance balance detection method, system and device of server system | |
| CN102339233A (en) | Cloud computing centralized management platform | |
| CN106059940B (en) | A kind of flow control methods and device | |
| CN110659130A (en) | CPU utilization rate self-adaptive adjusting method, device, terminal and storage medium | |
| CN105872061A (en) | Server cluster management method, device and system | |
| CN105791166B (en) | A kind of method and system of load balancing distribution | |
| CN109284229A (en) | A kind of dynamic adjusting method and relevant device based on QPS | |
| WO2016171950A1 (en) | Multivariable control for power-latency management to support optimization of data centers or other systems | |
| Chang et al. | Energy efficient resource selection and allocation strategy for virtual machine consolidation in cloud datacenters | |
| CN107370783B (en) | Scheduling method and device for cloud computing cluster resources | |
| CN105045667A (en) | Resource pool management method for vCPU scheduling of virtual machines | |
| CN113961353A (en) | Task processing method and distributed system for AI task | |
| CN111935952B (en) | Large machine room energy consumption regulation and control method and device | |
| CN110647437B (en) | Method, device, terminal and medium for adjusting CPU utilization rate based on PID controller |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20200630 |