CN113608901B - NVMe SSD constant temperature reliability test method and device - Google Patents

NVMe SSD constant temperature reliability test method and device Download PDF

Info

Publication number
CN113608901B
CN113608901B CN202110722973.6A CN202110722973A CN113608901B CN 113608901 B CN113608901 B CN 113608901B CN 202110722973 A CN202110722973 A CN 202110722973A CN 113608901 B CN113608901 B CN 113608901B
Authority
CN
China
Prior art keywords
fio
workload
temperature value
nvme ssd
nvme
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.)
Active
Application number
CN202110722973.6A
Other languages
Chinese (zh)
Other versions
CN113608901A (en
Inventor
刘迎
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Suzhou Inspur Intelligent Technology Co Ltd
Original Assignee
Suzhou Inspur Intelligent Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Suzhou Inspur Intelligent Technology Co Ltd filed Critical Suzhou Inspur Intelligent Technology Co Ltd
Priority to CN202110722973.6A priority Critical patent/CN113608901B/en
Publication of CN113608901A publication Critical patent/CN113608901A/en
Application granted granted Critical
Publication of CN113608901B publication Critical patent/CN113608901B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/008—Reliability or availability analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/3003—Monitoring arrangements specially adapted to the computing system or computing system component being monitored
    • G06F11/3037—Monitoring arrangements specially adapted to the computing system or computing system component being monitored where the computing system component is a memory, e.g. virtual memory, cache
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/3051—Monitoring arrangements for monitoring the configuration of the computing system or of the computing system component, e.g. monitoring the presence of processing resources, peripherals, I/O links, software programs
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/3058—Monitoring arrangements for monitoring environmental properties or parameters of the computing system or of the computing system component, e.g. monitoring of power, currents, temperature, humidity, position, vibrations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computing Systems (AREA)
  • Quality & Reliability (AREA)
  • Mathematical Physics (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Software Systems (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

The invention provides an NVMe SSD constant temperature reliability test method and device, wherein the method comprises the following steps: determining a temperature value and FIO workload relation model through an artificial neural network algorithm; acquiring a required constant temperature value, and calculating a required FIO workload through a relation model of the temperature value and the FIO workload; and performing reliability test on the FIO workload required by the NVMe SSD operation through a FIO tool, detecting a real-time temperature value through the NVMe CLI tool every set time period in the test process, and adjusting the real-time FIO workload according to a relation model of the temperature value and the FIO workload when the real-time temperature value is not matched with the required constant temperature value. According to the invention, the corresponding relation model of the temperature value of the NVMe SSD and the FIO work is determined through the artificial neural network algorithm, so that the reliability test of the NVMe SSD at a constant temperature is realized, and the test complexity is increased.

Description

NVMe SSD constant temperature reliability test method and device
Technical Field
The invention belongs to the technical field of hard disk reliability test, and particularly relates to an NVMe SSD constant temperature reliability test method and device.
Background
NVMe, short for Non-Volatile Memory express, is nonvolatile storage.
SSD is the abbreviation of Solid State Drive, solid state disk.
RDT, abbreviated as Reliability Demonstration Tests, is a reliability test.
FIO, an I/O tool used to stress test and verify hardware, supports 13 different I/O engines.
The conventional NVMe SSD RDT test controls the temperature of the NVMe SSD disc by changing the ambient temperature, so that the NVMe SSD disc performs the reliability test of the NVMe SSD at a certain required constant temperature.
This is a deficiency of the prior art, and therefore, it is necessary to provide a method and a device for testing the constant temperature reliability of NVMe SSD, aiming at the above-mentioned deficiency in the prior art.
Disclosure of Invention
Aiming at the defect that the reliability test result of the NVME SSD is inaccurate due to the fact that the real constant temperature is difficult to realize in the prior art by controlling the reliability test of the NVMe SSD to be tested at a constant temperature through changing the environmental temperature, the invention provides a constant temperature reliability test method and device for the NVMe SSD, and aims to solve the technical problems.
In a first aspect, the invention provides a method for testing the constant temperature reliability of an NVMe SSD, which comprises the following steps:
s1, determining a relation model of an NVMe SSD temperature value and a FIO workload through an artificial neural network algorithm;
s2, acquiring a required constant temperature value of the NVMe SSD, and calculating a required FIO workload through a relation model of the NVMe SSD temperature value and the FIO workload;
s3, performing reliability test on the FIO workload required by the operation of the NVMe SSD through a FIO tool, detecting the real-time temperature value of the NVMe SSD through the NVMe CLI tool every set time interval in the test process, and adjusting the real-time FIO workload according to a relation model of the temperature value of the NVMe SSD and the FIO workload when the real-time temperature value is not matched with the required constant temperature value.
Further, the specific steps of step S1 are as follows:
s11, creating a primary relation model of an NVMe SSD temperature value and an FIO workload based on an artificial neural network algorithm, taking FIO parameters as an input layer, taking the FIO workload as an hidden layer, and taking the NVMe SSD temperature value as an output layer;
s12, acquiring a sample data set, and training a primary relation model of the NVMe SSD temperature value and the FIO workload through the sample data set to obtain a relation model of the NVMe SSD temperature value and the FIO workload. And creating a primary relation model, determining the input and the output of the primary relation model, and acquiring the temperature value and the data corresponding to the FIO workload in the working process of the NVMe SSD as a sample data set.
Further, the specific steps of step S11 are as follows:
s111, based on an artificial neural network algorithm, using FIO parameters as an input layer, using FIO workload as an hidden layer, using an NVMe SSD temperature value as an output layer, and creating a primary relation model of the NVMe SSD temperature value and the FIO workload;
s112, setting the weight between the input layer and the hidden layer as a first weight, setting the weight between the hidden layer and the output layer as a second weight, and initializing the first weight and the second weight. And training the primary model through the sample set data, so as to continuously correct the values of the first weight and the second weight until the values meet the threshold requirements of each layer, namely, the model can reflect the real relation between the NVMe SSD temperature value and the FIO workload, and the model training is finished.
Further, FIO parameters include read-write mode, IO block size, number of working threads, and IO queue depth. Each set of FIO parameters corresponds to a FIO workload.
Further, the read-write mode includes a sequential write mode, a sequential read mode, a random write mode, a random read mode, a sequential hybrid read-write mode, and a random hybrid read-write mode;
the range of IO block size is larger than 4K and smaller than NVMe SSD capacity value;
the number of the working threads is less than or equal to 100;
IO queue depth is equal to or less than 4096. Setting a threshold value of the number of working threads based on the memory limit; the number of working threads and the depth of IO queues are integers. The IO block size is randomly valued from 4k to SSD capacity value; considering the memory limit of the test equipment, the working thread number range: integer random value between 0 and 100, IO queue depth range: integers between 0 and 4096 are randomly valued.
Further, the specific steps of step S12 are as follows:
s121, acquiring a sample data set, wherein the sample data set comprises FIO working parameter data, FIO workload data and NVMe SSD temperature values of all sampling time points of the NVMe SSD;
s122, sequentially inputting data in the sample data set into a primary relation model of the NVMe SSD temperature value and the FIO workload, and correcting the first weight and the second weight to obtain a relation model of the NVMe SSD temperature value and the FIO workload. One set of FIO operating parameter data corresponds to one FIO workload data. And continuously correcting the weight until the output meets the threshold requirements of each layer.
Further, the specific steps of step S2 are as follows:
s21, acquiring a required constant temperature value of the NVMe SSD according to a sample data set;
s22, inputting a required constant temperature value of the NVMe SSD into a relation model of the NVMe SSD temperature value and the FIO workload, and calculating the required FIO workload;
s23, adding the required constant temperature value and the required FIO workload of the NVMe SSD to the sample data set. And S1, determining a relation model, namely determining the corresponding relation between the NVMe SSD and the FIO workload, and inputting a required constant temperature value into the trained relation model to obtain the required FIO workload.
Further, the specific steps of step S3 are as follows:
s31, performing reliability test on FIO workload required by NVMe SSD operation through a FIO tool;
s32, monitoring a real-time temperature value of the NVMe SSD through an NVMe smart-log instruction of the NVMe CLI tool every set time period;
s33, judging whether the difference value between the real-time temperature value and the required constant temperature value exceeds a temperature threshold value;
if yes, go to step S34;
if not, returning to the step S32;
s34, inputting the required temperature value into a relation model of the NVMe SSD temperature value and the FIO workload, calculating the required FIO workload, and returning to the step S31. In the reliability test process, the FIO load is not constant, so that the NVMe SSD temperature value also changes, a monitoring time interval is required to be set, and the temperature is detected, so that the real constant temperature is realized, and the reliability test effect is ensured.
In a second aspect, the present invention provides an NVMe SSD thermostability test apparatus based on FIO, including:
the relation model determining module is used for determining a relation model of the NVMe SSD temperature value and the FIO workload through an artificial neural network algorithm;
the required FIO workload calculation module is used for acquiring a required constant temperature value of the NVMe SSD and calculating the required FIO workload through a relation model of the NVMe SSD temperature value and the FIO workload;
the reliability testing module is used for carrying out reliability testing on the FIO workload required by the operation of the NVMe SSD through the FIO tool, detecting the real-time temperature value of the NVMe SSD through the NVMe CLI tool every set time period in the testing process, and adjusting the real-time FIO workload according to the relation model of the NVMe SSD temperature value and the FIO workload when the real-time temperature value is not matched with the required constant temperature value.
Further, the relationship model determination module includes:
the primary relation model creation unit is used for creating a primary relation model of the NVMe SSD temperature value and the FIO workload based on an artificial neural network algorithm, taking the FIO parameter as an input layer, the FIO workload as an hidden layer and the NVMe SSD temperature value as an output layer;
the model training unit is used for acquiring a sample data set, training a primary relation model of the NVMe SSD temperature value and the FIO workload through the sample data set, and obtaining a relation model of the NVMe SSD temperature value and the FIO workload.
Further, the primary relationship model creation unit includes:
the primary model creation subunit is used for creating a primary relation model of the NVMe SSD temperature value and the FIO workload based on an artificial neural network algorithm, with the FIO parameter as an input layer, with the FIO workload as an hidden layer and with the NVMe SSD temperature value as an output layer;
the model weight setting and initializing subunit is configured to set a weight between the input layer and the hidden layer as a first weight, set a weight between the hidden layer and the output layer as a second weight, and initialize the first weight and the second weight.
Further, FIO parameters include read-write mode, IO block size, number of working threads, and IO queue depth.
Further, the read-write mode includes a sequential write mode, a sequential read mode, a random write mode, a random read mode, a sequential hybrid read-write mode, and a random hybrid read-write mode;
the range of IO block size is larger than 4K and smaller than NVMe SSD capacity value;
the number of the working threads is less than or equal to 100;
IO queue depth is equal to or less than 4096. The IO block size is randomly valued from 4k to SSD capacity value; considering the memory limit of the test equipment, the working thread number range: integer random value between 0 and 100, IO queue depth range: integers between 0 and 4096 are randomly valued.
Further, the model training unit includes:
the sample data set acquisition subunit is used for acquiring a sample data set, wherein the sample data set comprises FIO working parameter data, FIO workload data and NVMe SSD temperature values of all sampling time points of the NVMe SSD;
the model training subunit is used for sequentially inputting the data in the sample data set into a primary relation model of the NVMe SSD temperature value and the FIO workload, and correcting the first weight and the second weight to obtain a relation model of the NVMe SSD temperature value and the FIO workload.
Further, the required FIO workload calculation module includes:
the required temperature calculation unit is used for acquiring a required constant temperature value of the NVMe SSD according to the sample data set;
the required load calculation unit is used for inputting the required constant temperature value of the NVMe SSD into a relation model of the NVMe SSD temperature value and the FIO workload, and calculating the required FIO workload;
and a sample data adding unit for adding the required constant temperature value and the required FIO workload of the NVMe SSD to the sample data set.
Further, the reliability test module includes:
the reliability testing unit is used for carrying out reliability testing on the FIO workload required by the NVMe SSD operation through the FIO tool;
the temperature monitoring unit is used for monitoring the real-time temperature value of the NVMe SSD through an NVMe smart-log instruction of the NVMe CLI tool every set time period;
the temperature judging unit is used for judging whether the difference value between the real-time temperature value and the required constant temperature value exceeds a temperature threshold value;
and the required load recalculation unit is used for inputting the required temperature value into a relation model of the NVMe SSD temperature value and the FIO workload when the temperature difference exceeds the temperature threshold value, and calculating the required FIO workload.
The invention has the advantages that,
according to the method and the device for testing the constant-temperature reliability of the NVMe SSD, the corresponding relation model of the temperature value of the NVMe SSD and the FIO work is determined by using the artificial neural network algorithm of machine learning, so that the reliability test of the NVMe SSD at a certain constant temperature is realized, the test complexity of the NVMe SSD is increased, the application scene of the scheme is very abundant, and the RDT reliability test is also greatly optimized and improved.
In addition, the invention has reliable design principle, simple structure and very wide application prospect.
It can be seen that the present invention has outstanding substantial features and significant advances over the prior art, as well as the benefits of its implementation.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the description of the embodiments or the prior art will be briefly described below, and it will be obvious to those skilled in the art that other drawings can be obtained from these drawings without inventive effort.
Fig. 1 is a schematic flow chart of an NVMe SSD thermostability test method of the invention.
Fig. 2 is a flow chart diagram II of an NVMe SSD constant temperature reliability test method of the invention.
Fig. 3 is a schematic diagram of an NVMe SSD thermostability test apparatus of the present invention.
In the figure, a 1-relation model determining module; 1.1-a primary relationship model creation unit; 1.2-model training unit; 2-a required FIO workload calculation module; 2.1-a desired temperature calculation unit; 2.2-a required load calculation unit; 2.3-sample data increment unit; 3-a reliability test module; 3.1-a reliability test unit; 3.2-a temperature monitoring unit; 3.3-a temperature judgment unit; 3.4-required load recalculation unit.
Detailed Description
In order to make the technical solution of the present invention better understood by those skilled in the art, the technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
Example 1:
as shown in fig. 1, the invention provides a method for testing the constant temperature reliability of an NVMe SSD based on FIO, which comprises the following steps:
s1, determining a relation model of an NVMe SSD temperature value and a FIO workload through an artificial neural network algorithm;
s2, acquiring a required constant temperature value of the NVMe SSD, and calculating a required FIO workload through a relation model of the NVMe SSD temperature value and the FIO workload;
s3, performing reliability test on the FIO workload required by the operation of the NVMe SSD through a FIO tool, detecting the real-time temperature value of the NVMe SSD through the NVMe CLI tool every set time interval in the test process, and adjusting the real-time FIO workload according to a relation model of the temperature value of the NVMe SSD and the FIO workload when the real-time temperature value is not matched with the required constant temperature value.
Example 2:
as shown in fig. 2, the invention provides a method for testing the constant temperature reliability of NVMe SSD based on FIO, which comprises the following steps:
s1, determining a relation model of an NVMe SSD temperature value and a FIO workload through an artificial neural network algorithm; the method comprises the following specific steps:
s11, creating a primary relation model of an NVMe SSD temperature value and a FIO workload based on an artificial neural network algorithm, taking a FIO parameter as an input layer, taking the FIO workload as an hidden layer, and taking the NVMeSSD temperature value as an output layer; FIO parameters include read-write mode, IO block size, number of working threads, and IO queue depth; the read-write mode comprises a sequential write mode, a sequential read mode, a random write mode, a random read mode, a sequential mixed read-write mode and a random mixed read-write mode; the range of IO block size is larger than 4K and smaller than NVMe SSD capacity value; the number of the working threads is less than or equal to 100; IO queue depth is less than or equal to 4096;
s12, acquiring a sample data set, and training a primary relation model of the NVMe SSD temperature value and the FIO workload through the sample data set to obtain a relation model of the NVMe SSD temperature value and the FIO workload;
s2, acquiring a required constant temperature value of the NVMe SSD, and calculating a required FIO workload through a relation model of the NVMe SSD temperature value and the FIO workload; the method comprises the following specific steps:
s21, acquiring a required constant temperature value of the NVMe SSD according to a sample data set;
s22, inputting a required constant temperature value of the NVMe SSD into a relation model of the NVMe SSD temperature value and the FIO workload, and calculating the required FIO workload;
s23, adding a required constant temperature value and a required FIO workload of the NVMe SSD to a sample data set;
s3, performing reliability test on the FIO workload required by the operation of the NVMe SSD through a FIO tool, detecting the real-time temperature value of the NVMe SSD through the NVMe CLI tool every set time interval in the test process, and adjusting the real-time FIO workload according to a relation model of the temperature value of the NVMe SSD and the FIO workload when the real-time temperature value is not matched with the required constant temperature value; the method comprises the following specific steps:
s31, performing reliability test on FIO workload required by NVMe SSD operation through a FIO tool;
s32, monitoring a real-time temperature value of the NVMe SSD through an NVMe smart-log instruction of the NVMe CLI tool every set time period;
s33, judging whether the difference value between the real-time temperature value and the required constant temperature value exceeds a temperature threshold value;
if yes, go to step S34;
if not, returning to the step S32;
s34, inputting the required temperature value into a relation model of the NVMe SSD temperature value and the FIO workload,
the required FIO workload is calculated and the process returns to step S31.
Example 3:
as shown in fig. 2, the invention provides a method for testing the constant temperature reliability of NVMe SSD, which comprises the following steps:
s1, determining a relation model of an NVMe SSD temperature value and a FIO workload through an artificial neural network algorithm; the method comprises the following specific steps:
s11, creating a primary relation model of an NVMe SSD temperature value and a FIO workload based on an artificial neural network algorithm, taking a FIO parameter as an input layer, taking the FIO workload as an hidden layer, and taking the NVMeSSD temperature value as an output layer; the method comprises the following specific steps:
s111, based on an artificial neural network algorithm, using FIO parameters as an input layer, using FIO workload as an hidden layer, using an NVMe SSD temperature value as an output layer, and creating a primary relation model of the NVMe SSD temperature value and the FIO workload; FIO parameters include read-write mode, IO block size, number of working threads, and IO queue depth; the read-write mode comprises a sequential write mode, a sequential read mode, a random write mode, a random read mode, a sequential mixed read-write mode and a random mixed read-write mode; the range of IO block size is larger than 4K and smaller than NVMe SSD capacity value; the number of the working threads is less than or equal to 100; IO queue depth is less than or equal to 4096;
s112, setting the weight between the input layer and the hidden layer as a first weight, setting the weight between the hidden layer and the output layer as a second weight, and initializing the first weight and the second weight;
s12, acquiring a sample data set, and training a primary relation model of the NVMe SSD temperature value and the FIO workload through the sample data set to obtain a relation model of the NVMe SSD temperature value and the FIO workload; the method comprises the following specific steps:
s121, acquiring a sample data set, wherein the sample data set comprises FIO working parameter data, FIO workload data and NVMe SSD temperature values of all sampling time points of the NVMe SSD;
s122, sequentially inputting data in the sample data set into a primary relation model of the NVMe SSD temperature value and the FIO workload, and correcting the first weight and the second weight to obtain a relation model of the NVMe SSD temperature value and the FIO workload;
s2, acquiring a required constant temperature value of the NVMe SSD, and calculating a required FIO workload through a relation model of the NVMe SSD temperature value and the FIO workload; the method comprises the following specific steps:
s21, acquiring a required constant temperature value of the NVMe SSD according to a sample data set;
s22, inputting a required constant temperature value of the NVMe SSD into a relation model of the NVMe SSD temperature value and the FIO workload, and calculating the required FIO workload;
s23, adding a required constant temperature value and a required FIO workload of the NVMe SSD to a sample data set;
s3, performing reliability test on the FIO workload required by the operation of the NVMe SSD through a FIO tool, detecting the real-time temperature value of the NVMe SSD through the NVMe CLI tool every set time interval in the test process, and adjusting the real-time FIO workload according to a relation model of the temperature value of the NVMe SSD and the FIO workload when the real-time temperature value is not matched with the required constant temperature value; the method comprises the following specific steps:
s31, performing reliability test on FIO workload required by NVMe SSD operation through a FIO tool;
s32, monitoring a real-time temperature value of the NVMe SSD through an NVMe smart-log instruction of the NVMe CLI tool every set time period;
s33, judging whether the difference value between the real-time temperature value and the required constant temperature value exceeds a temperature threshold value;
if yes, go to step S34;
if not, returning to the step S32;
s34, inputting the required temperature value into a relation model of the NVMe SSD temperature value and the FIO workload, calculating the required FIO workload, and returning to the step S31.
In the above embodiment 3, in step S11, a primary relationship model of NVMe SSD temperature values and FIO workload is created based on an artificial neural network algorithm, with FIO parameters as an input layer, with FIO workload as an hidden layer,taking NVMe SSD temperature value as output layer, assuming that the unit numbers of input layer, hidden layer and output layer are I, J and K respectively, the input is (x 1 ,x 2 ，。。。，x I-1 ) The hidden layer output is (z 1 ,z 2 ，。。。，z J-1 ) The actual output of the network is (y 1 ,y 2 ，。。。，y K-1 )，(d 1 ,d 2 ，。。。，d K-1 ) Representing a desired output of the training sample;
let the weight of input layer element i to hidden layer element j be v ij The weight from the hidden layer unit j to the output layer unit k is w jk By P j And P k To represent thresholds of the hidden layer unit and the output layer unit, respectively;
in the forward information propagation stage, the output value of the hidden layer unit of the model is as follows:
the output values of the units of the output layer are as follows:
error back propagation stage:
if the activation function of each layer of the artificial neural network takes a monopole S-shaped function, namely:
the weight correction amount among the layers of the model can be conveniently calculated;
for weight value adjustment between the output layer and the hidden layer:
Δw jk ＝ηO j (d k -O k )f'(net k )＝ηO j (d k -O k )O k (1-O k )
for weight value adjustment between hidden layer and input layer:
according to Deltaw jk The first weight is adjusted according to Deltav ij And adjusting the second weight to obtain a final relation model of the NVMe SSD temperature value and the FIO workload.
Example 4:
as shown in fig. 3, the present invention provides an NVMe SSD constant temperature reliability test apparatus, including:
the relation model determining module 1 is used for determining a relation model of the NVMe SSD temperature value and the FIO workload through an artificial neural network algorithm;
the required FIO workload calculation module 2 is used for acquiring a required constant temperature value of the NVMe SSD and calculating the required FIO workload through a relation model of the NVMe SSD temperature value and the FIO workload;
the reliability testing module 3 is configured to perform reliability testing on a required FIO workload for running the NVMe SSD by using a FIO tool, detect a real-time temperature value of the NVMe SSD by using the NVMe CLI tool every set period of time in a testing process, and adjust the real-time FIO workload according to a relationship model of the NVMe SSD temperature value and the FIO workload when the real-time temperature value is not matched with the required constant temperature value.
Example 5:
as shown in fig. 3, the present invention provides an NVMe SSD constant temperature reliability test apparatus, including:
the relation model determining module 1 is used for determining a relation model of the NVMe SSD temperature value and the FIO workload through an artificial neural network algorithm; the relationship model determination module 1 includes:
the primary relation model creation unit 1.1 is used for creating a primary relation model of an NVMe SSD temperature value and a FIO workload based on an artificial neural network algorithm, wherein the FIO parameter is used as an input layer, the FIO workload is used as an implicit layer, and the NVMe SSD temperature value is used as an output layer;
the model training unit 1.2 is used for acquiring a sample data set, training a primary relation model of the NVMe SSD temperature value and the FIO workload through the sample data set, and obtaining a relation model of the NVMe SSD temperature value and the FIO workload;
the required FIO workload calculation module 2 is used for acquiring a required constant temperature value of the NVMe SSD and calculating the required FIO workload through a relation model of the NVMe SSD temperature value and the FIO workload; the required FIO workload calculation module 2 includes:
the required temperature calculating unit 2.1 is used for acquiring a required constant temperature value of the NVMe SSD according to the sample data set;
the required load calculation unit 2.2 is used for inputting the required constant temperature value of the NVMe SSD into a relation model of the NVMe SSD temperature value and the FIO workload, and calculating the required FIO workload;
a sample data adding unit 2.3 for adding the required constant temperature value and the required FIO workload of the NVMe SSD to the sample data set;
the reliability testing module 3 is used for carrying out reliability testing on the FIO workload required by the operation of the NVMe SSD through the FIO tool, detecting the real-time temperature value of the NVMe SSD through the NVMe CLI tool every set time period in the testing process, and adjusting the real-time FIO workload according to a relation model of the NVMe SSD temperature value and the FIO workload when the real-time temperature value is not matched with the required constant temperature value; the reliability test module 3 includes:
the reliability test unit 3.1 is used for performing reliability test on the FIO workload required by the NVMe SSD operation through the FIO tool;
the temperature monitoring unit 3.2 is used for monitoring the real-time temperature value of the NVMe SSD through an NVMe smart-log instruction of the NVMe CLI tool every set time period;
a temperature judging unit 3.3 for judging whether the difference value between the real-time temperature value and the required constant temperature value exceeds a temperature threshold value;
and the required load recalculation unit 3.4 is used for inputting the required temperature value into a relation model of the NVMe SSD temperature value and the FIO workload when the temperature difference exceeds the temperature threshold value, and calculating the required FIO workload.
Example 6:
as shown in fig. 3, the present invention provides an NVMe SSD constant temperature reliability test apparatus, including:
the relation model determining module 1 is used for determining a relation model of the NVMe SSD temperature value and the FIO workload through an artificial neural network algorithm; the relationship model determination module 1 includes:
the primary relation model creation unit 1.1 is used for creating a primary relation model of an NVMe SSD temperature value and a FIO workload based on an artificial neural network algorithm, wherein the FIO parameter is used as an input layer, the FIO workload is used as an implicit layer, and the NVMe SSD temperature value is used as an output layer; the primary relation model creation unit 1.1 includes:
the primary model creation subunit is used for creating a primary relation model of the NVMe SSD temperature value and the FIO workload based on an artificial neural network algorithm, with the FIO parameter as an input layer, with the FIO workload as an hidden layer and with the NVMe SSD temperature value as an output layer;
the model weight setting and initializing subunit is used for setting the weight between the input layer and the hidden layer as a first weight, setting the weight between the hidden layer and the output layer as a second weight, and initializing the first weight and the second weight;
the model training unit 1.2 is used for acquiring a sample data set, training a primary relation model of the NVMe SSD temperature value and the FIO workload through the sample data set, and obtaining a relation model of the NVMe SSD temperature value and the FIO workload; the model training unit 1.2 comprises:
the sample data set acquisition subunit is used for acquiring a sample data set, wherein the sample data set comprises FIO working parameter data, FIO workload data and NVMe SSD temperature values of all sampling time points of the NVMe SSD;
the model training subunit is used for sequentially inputting data in the sample data set into a primary relation model of the NVMe SSD temperature value and the FIO workload, and correcting the first weight and the second weight to obtain a relation model of the NVMe SSD temperature value and the FIO workload;
the required FIO workload calculation module 2 is used for acquiring a required constant temperature value of the NVMe SSD and calculating the required FIO workload through a relation model of the NVMe SSD temperature value and the FIO workload; the required FIO workload calculation module 2 includes:
the required temperature calculating unit 2.1 is used for acquiring a required constant temperature value of the NVMe SSD according to the sample data set;
the required load calculation unit 2.2 is used for inputting the required constant temperature value of the NVMe SSD into a relation model of the NVMe SSD temperature value and the FIO workload, and calculating the required FIO workload;
a sample data adding unit 2.3 for adding the required constant temperature value and the required FIO workload of the NVMe SSD to the sample data set;
the reliability testing module 3 is used for carrying out reliability testing on the FIO workload required by the operation of the NVMe SSD through the FIO tool, detecting the real-time temperature value of the NVMe SSD through the NVMe CLI tool every set time period in the testing process, and adjusting the real-time FIO workload according to a relation model of the NVMe SSD temperature value and the FIO workload when the real-time temperature value is not matched with the required constant temperature value; the reliability test module 3 includes:
the reliability test unit 3.1 is used for performing reliability test on the FIO workload required by the NVMe SSD operation through the FIO tool;
the temperature monitoring unit 3.2 is used for monitoring the real-time temperature value of the NVMe SSD through an NVMe smart-log instruction of the NVMe CLI tool every set time period;
a temperature judging unit 3.3 for judging whether the difference value between the real-time temperature value and the required constant temperature value exceeds a temperature threshold value;
and the required load recalculation unit 3.4 is used for inputting the required temperature value into a relation model of the NVMe SSD temperature value and the FIO workload when the temperature difference exceeds the temperature threshold value, and calculating the required FIO workload.
Although the present invention has been described in detail by way of preferred embodiments with reference to the accompanying drawings, the present invention is not limited thereto. Various equivalent modifications and substitutions may be made in the embodiments of the present invention by those skilled in the art without departing from the spirit and scope of the present invention, and it is intended that all such modifications and substitutions be within the scope of the present invention/be within the scope of the present invention as defined by the appended claims. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (10)

1. The NVMe SSD constant temperature reliability test method is characterized by comprising the following steps:
s1, determining a relation model of an NVMe SSD temperature value and a FIO workload through an artificial neural network algorithm;
s2, acquiring a required constant temperature value of the NVMe SSD, and calculating a required FIO workload through a relation model of the NVMe SSD temperature value and the FIO workload;
s3, performing reliability test on the FIO workload required by the operation of the NVMe SSD through a FIO tool, detecting the real-time temperature value of the NVMe SSD through the NVMe CLI tool every set time interval in the test process, and adjusting the real-time FIO workload according to a relation model of the temperature value of the NVMe SSD and the FIO workload when the real-time temperature value is not matched with the required constant temperature value.
2. The NVMe SSD constant temperature reliability test method of claim 1, wherein step S1 comprises the specific steps of:
s11, creating a primary relation model of an NVMe SSD temperature value and an FIO workload based on an artificial neural network algorithm, taking FIO parameters as an input layer, taking the FIO workload as an hidden layer, and taking the NVMe SSD temperature value as an output layer;
s12, acquiring a sample data set, and training a primary relation model of the NVMe SSD temperature value and the FIO workload through the sample data set to obtain a relation model of the NVMe SSD temperature value and the FIO workload.
3. The NVMe SSD constant temperature reliability test method of claim 2, characterized by step S11 comprising the specific steps of:
s111, based on an artificial neural network algorithm, using FIO parameters as an input layer, using FIO workload as an hidden layer, using an NVMe SSD temperature value as an output layer, and creating a primary relation model of the NVMe SSD temperature value and the FIO workload;
s112, setting the weight between the input layer and the hidden layer as a first weight, setting the weight between the hidden layer and the output layer as a second weight, and initializing the first weight and the second weight.
4. The NVMe SSD constant temperature reliability test method of claim 3, characterized by step S12 comprising the specific steps of:
s121, acquiring a sample data set, wherein the sample data set comprises FIO working parameter data, FIO workload data and NVMe SSD temperature values of all sampling time points of the NVMe SSD;
s122, sequentially inputting data in the sample data set into a primary relation model of the NVMe SSD temperature value and the FIO workload, and correcting the first weight and the second weight to obtain a relation model of the NVMe SSD temperature value and the FIO workload.
5. The NVMe SSD constant temperature reliability test method of claim 4, wherein step S2 comprises the specific steps of:
s21, acquiring a required constant temperature value of the NVMe SSD according to a sample data set;
s22, inputting a required constant temperature value of the NVMe SSD into a relation model of the NVMe SSD temperature value and the FIO workload, and calculating the required FIO workload;
s23, adding the required constant temperature value and the required FIO workload of the NVMe SSD to the sample data set.
6. The NVMe SSD constant temperature reliability test method of claim 1, wherein step S3 comprises the specific steps of:
s31, performing reliability test on FIO workload required by NVMe SSD operation through a FIO tool;
s32, monitoring a real-time temperature value of the NVMe SSD through an NVMe smart-log instruction of the NVMe CLI tool every set time period;
s33, judging whether the difference value between the real-time temperature value and the required constant temperature value exceeds a temperature threshold value;
if yes, go to step S34;
if not, returning to the step S32;
s34, inputting the required constant temperature value into a relation model of the NVMe SSD temperature value and the FIO workload, calculating the required FIO workload, and returning to the step S31.
7. An NVMe SSD constant temperature reliability test apparatus, comprising:
the relation model determining module (1) is used for determining a relation model of the NVMe SSD temperature value and the FIO workload through an artificial neural network algorithm;
the required FIO workload calculation module (2) is used for acquiring a required constant temperature value of the NVMe SSD and calculating the required FIO workload through a relation model of the NVMe SSD temperature value and the FIO workload;
the reliability testing module (3) is used for carrying out reliability testing on the FIO workload required by the operation of the NVMe SSD through the FIO tool, detecting the real-time temperature value of the NVMe SSD through the NVMe CLI tool every set time period in the testing process, and adjusting the real-time FIO workload according to the relation model of the NVMe SSD temperature value and the FIO workload when the real-time temperature value is not matched with the required constant temperature value.
8. The NVMe SSD constant temperature reliability test apparatus of claim 7, wherein the relationship model determination module (1) includes:
the primary relation model creation unit (1.1) is used for creating a primary relation model of the NVMe SSD temperature value and the FIO workload based on an artificial neural network algorithm, taking the FIO parameter as an input layer, the FIO workload as an implicit layer and the NVMe SSD temperature value as an output layer;
the model training unit (1.2) is used for acquiring a sample data set, training a primary relation model of the NVMe SSD temperature value and the FIO workload through the sample data set, and obtaining a relation model of the NVMe SSD temperature value and the FIO workload.
9. The NVMe SSD thermostability test apparatus of claim 7, characterized in that the required FIO workload calculation module (2) comprises:
a required temperature calculation unit (2.1) for obtaining a required constant temperature value of the NVMe SSD from the sample data set;
a required load calculation unit (2.2) for inputting the required constant temperature value of the NVMe SSD into a relation model of the NVMe SSD temperature value and the FIO workload, and calculating the required FIO workload;
a sample data adding unit (2.3) for adding the required thermostat value and the required FIO workload of the NVMe SSD to the sample data set.
10. The NVMe SSD constant temperature reliability test apparatus of claim 7, characterized in that the reliability test module (3) includes:
a reliability test unit (3.1) for performing reliability test on the FIO workload required by the NVMe SSD operation by the FIO tool;
a temperature monitoring unit (3.2) for monitoring real-time temperature values of the NVMe SSD through NVMe smart-log instructions of the NVMe CLI tool every set period of time;
a temperature judging unit (3.3) for judging whether the difference between the real-time temperature value and the required constant temperature value exceeds a temperature threshold value;
and the required load recalculation unit (3.4) is used for inputting the required constant temperature value into a relation model of the NVMe SSD temperature value and the FIO workload when the temperature difference exceeds the temperature threshold value, and calculating the required FIO workload.
CN202110722973.6A 2021-06-28 2021-06-28 NVMe SSD constant temperature reliability test method and device Active CN113608901B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110722973.6A CN113608901B (en) 2021-06-28 2021-06-28 NVMe SSD constant temperature reliability test method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110722973.6A CN113608901B (en) 2021-06-28 2021-06-28 NVMe SSD constant temperature reliability test method and device

Publications (2)

Publication Number Publication Date
CN113608901A CN113608901A (en) 2021-11-05
CN113608901B true CN113608901B (en) 2023-07-18

Family

ID=78303830

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110722973.6A Active CN113608901B (en) 2021-06-28 2021-06-28 NVMe SSD constant temperature reliability test method and device

Country Status (1)

Country Link
CN (1) CN113608901B (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7873885B1 (en) * 2004-01-20 2011-01-18 Super Talent Electronics, Inc. SSD test systems and methods
CN111696618A (en) * 2020-05-15 2020-09-22 苏州浪潮智能科技有限公司 Method and system for testing power-on and power-off stability of SSD (solid State disk)

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7873885B1 (en) * 2004-01-20 2011-01-18 Super Talent Electronics, Inc. SSD test systems and methods
CN111696618A (en) * 2020-05-15 2020-09-22 苏州浪潮智能科技有限公司 Method and system for testing power-on and power-off stability of SSD (solid State disk)

Also Published As

Publication number Publication date
CN113608901A (en) 2021-11-05

Similar Documents

Publication Publication Date Title
CN109817267B (en) Deep learning-based flash memory life prediction method and system and computer-readable access medium
JP2010530179A (en) Virtual sensor system and method
US20050273296A1 (en) Neural network model for electric submersible pump system
CN111814342B (en) Complex equipment reliability hybrid model and construction method thereof
CN112413812B (en) Refrigerant leakage detection method and system and air conditioner
CN112214880B (en) Multi-stage maintainability assessment method based on virtual-real fusion
CN112817524A (en) Flash memory reliability grade online prediction method and device based on dynamic neural network
CN107704704A (en) A kind of relay class unit Estimation of The Storage Reliability method based on bayesian theory
CN116089870A (en) Industrial equipment fault prediction method and device based on meta-learning under small sample condition
CN113983646A (en) Air conditioner interaction end energy consumption prediction method based on generation countermeasure network and air conditioner
CN113608901B (en) NVMe SSD constant temperature reliability test method and device
CN114511025A (en) Fan fault diagnosis method and device based on weighted multi-sensor fusion filtering
CN110543108B (en) Heating system measurement data correction method and system based on mechanism model prediction
CN112132324A (en) Ultrasonic water meter data restoration method based on deep learning model
CN109240927B (en) GDE 3-based multi-stage dynamic multi-target test resource allocation method
CN115408949B (en) Load model parameter identification method System, apparatus and medium
CN116292057A (en) Control method, device, equipment and medium for variable-speed pumped storage unit
CN115184055B (en) Test set determination method and system for hierarchical testability optimization
CN114296343B (en) Deep reinforcement learning-based aeroengine compression part characteristic correction method
CN116085245A (en) Online compressor performance prediction method and system based on OS-ELM
EP3955075A1 (en) Engine model construction method, program, and apparatus
CN115057006A (en) Distillation strategy evaluation method, device and medium based on reinforcement learning
CN112685954A (en) Method and device for predicting wind speed and fan rotating speed of automobile environment wind tunnel
CN116992222B (en) Method, device, equipment and medium for migration learning of wind element correction model
CN116069646B (en) Test set determination method and system for multi-objective testability optimization

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
GR01 Patent grant
GR01 Patent grant