CN104899119A

CN104899119A - Method for automatically detecting hard disk abnormality

Info

Publication number: CN104899119A
Application number: CN201510262060.5A
Authority: CN
Inventors: 刘凤刚
Original assignee: Inspur Electronic Information Industry Co Ltd
Current assignee: Inspur Electronic Information Industry Co Ltd
Priority date: 2015-05-21
Filing date: 2015-05-21
Publication date: 2015-09-09

Abstract

The present invention, belonging to the technical field of computers, provides a method for automatically detecting a hard disk abnormality. The method comprises steps of: preparing a test environment on a server, setting up a test platform, installing an operating system, installing a test tool, configuring test parameters, starting a test, monitoring a test process; and condition selection: if the test monitoring process fails, checking error report information and recording a hard disk SN code, conduct subsequent processing, and ending the test; and if the test monitoring process is subjected to no abnormality, checking a test result, and ending the test. The method for automatically detecting a hard disk abnormality achieves detection on the hard disk information and health state in combination with the smartctl command under Linux, integrates such detection into a restart test of the server, achieves the objective of testing stability of a storage subsystem by long-term restart test and the hard disk detection test each time; and additionally implements the hard disk reliability and stability tests.

Description

A kind of method of automatic detection hard disk exception

Technical field

The present invention relates to field of computer technology, specifically a kind of method of automatic detection hard disk exception.

Background technology

General, along with development and the growth of server, server provider has opened up the application of every field server, such as internet, finance ... Deng.The application of client, than the past, becomes complicated various.Hard disk is as one of server critical component, and once appearance hard disk more than once fell the problem of dish, had a strong impact on the data-storage applications of client, and this is the serious quality problems that client is difficult to accept.

Along with the continuous development of server industries, Rack is more obvious as the trend of internet development total solution; Simultaneously along with the increase of rack nodes, the increase of hard disk self-capacity, storage subsystem there has also been unprecedented growth, just becomes particularly important like this for the compatibility of hard disk, controller compatibility, backboard, this link of cable and stability.

As server product provider, how to meet the diversified demand of user, guarantee various applied environment stable, be reliably the problem that present stage must face and consider.Therefore hard disk is used on the server, test and the checking of comprehensively functional, stability and performance etc. must be carried out, thus guarantee the quality of this product.

Summary of the invention

Technical assignment of the present invention solves the deficiencies in the prior art, provides a kind of method of automatic detection hard disk exception.

Technical scheme of the present invention realizes in the following manner, and this detects the method for hard disk exception automatically, and the step of the method is:

Setup test environment, builds test platform on the server, installing operating system, installation testing instrument, configuration testing parameter, starts test, control and measuring process, and condition is selected:

If monitor test procedure failure, then check error information and recording disk SN code, make subsequent treatment, terminate test;

If monitor test process without any exception, then checks test result, terminate test.

Setup test environment on the server, build test platform, Linux or CentOS operating system is installed, installation testing instrument, copy testing tool to test machine, successively run check.sh program, hd.sh program, run the number of times amendment reboot.sh of restarting according to actual needs, and place it in root/.bashrc, configuration testing parameter, start test, control and measuring process, condition is selected:

Before starting test, write test script, first hard disk relevant information read before test and be automatically recorded in text, restart server at every turn and all can rerun this test script, detect all information of hard disk and make comparisons with recorded information before, if there be abnormal will automatically reporting an error, entry condition is selected.

Test script is arranged restarts number of times, also can automatically stop test according to prompting, when there being problem to occur, arranging red prompting, and automatically stopping test.

The beneficial effect that the present invention is compared with prior art produced is:

This method automatically detecting hard disk exception is in conjunction with the detection of smartctl order under Linux to hard disk information and health status, and be dissolved into server restart in test, by restarting test and each hard disk detection test for a long time, reach storage subsystem stability test object, also can take into account hard disk reliability and stability test simultaneously.

This method automatically detecting hard disk exception is simple to operate, and automaticity is high, and practicality is comparatively strong, can save manpower, effectively ensure that stability and the reliability of server, is the very effective method of checking product quality.

Accompanying drawing explanation

Accompanying drawing 1 is schematic flow sheet of the present invention.

Embodiment

Be described in detail below below in conjunction with the method for accompanying drawing to a kind of automatic detection hard disk exception of the present invention.

As shown in drawings, the method for automatic detection hard disk exception of the present invention, the step of the method is:

Hard disk SN code is the abbreviation of SerialNumber, is sometimes also SerialNo, namely product ID.

The script of test procedure, mainly in order to stability and the reliability of the health status and storage subsystem that detect hard disk, to reach the object improving subsystem stability and guarantee product quality.

S.M.A.R.T. be English Self-Monitoring Analysis and Reporting Technology(automatic detection analysis and reporting techniques) write a Chinese character in simplified form.It can be monitored the magnetic head unit of hard disk, hard disk temperature, disc surface dielectric material, motor and drive system, hard drive internal circuit etc., and immediate analysis also forecasts the contingent problem of hard disk.

Utilize smartctl order under Linux can check the function of hard disk model, Firmware version, amount of capacity, health status etc., write test script: first hard disk relevant information read before test and be automatically recorded in certain text, restart server at every turn and all can rerun this script, detect all information of hard disk and make comparisons with recorded information before, if there be abnormal will automatically reporting an error.Just can find according to error information is which block hard disk has problem, is which type of problem, and then is further analyzed and verifies.This script can be arranged restarts number of times, also can automatically stop test according to prompting, when there being problem to occur, having red prompting, and automatically stopping test.

Test procedure is as follows:

1), reboot.sh scripted code:

console=`/usr/bin/tty`

# .bashrc

# User specific aliases and functions

alias rm='rm -i'

alias cp='cp -i'

alias mv='mv -i'

# Source global definitions

if [ -f /etc/bashrc ]; then

. /etc/bashrc

fi

#source /opt/intel/composer_xe_2013/bin/compilervars.sh intel64

#source /opt/intel/mkl/bin/intel64/mklvars_intel64.sh

#source /opt/intel/impi/4.1.1.036/bin64/mpivars.sh

case $console in

/dev/tty1 | /dev/pts[0-100])

if [ -f /tmp/rcishu ];then

rcishu=`cat /tmp/rcishu`

else

/bin/touch rcishu

rcishu=0

fi

if [ $rcishu -lt 50 ];

then

sh /opt/scripts/hd.sh

sh /opt/scripts/hd.sh >> /tmp/hdtest.txt

read -t 5 -p "Press y Break Reboot " aa

if [ "${aa}" = "y" ]||[ "${aa}" = "Y" ] ;then

echo Breaked Reboot Cycle!!!!

else

rcishu=$(($rcishu+1))

echo "${rcishu}"> /tmp/rcishu

echo reboot $rcishu

/sbin/reboot

fi

esac

3), check.sh scripted code:

#!/bin/sh

rm /tmp/paizhi.txt -rf

/sbin/service ipmi restart > /dev/null

file=/proc/cpuinfo

num=`grep "physical id" $file |sort |uniq |awk '{print $4}' |wc -l`

corenum=`grep "cpu cores" $file |sort |uniq |awk '{print $4}'|sed -nr ' H;${x; s/\n/ /g;p}'`

clock=`grep name $file |awk -F CPU '{print $2}' |sed -e 's/ 0 //g' -e 's/ //g' -e 's/GHz//g'|awk -F '{print $2}'|sort |uniq|sed -nr ' H;${x; s/\n/ /g;p}'`

model=`grep name $file |awk -F CPU '{print $2}' |sed -e 's/ 0 //g' -e 's/ //g' |awk -F '{print $1}'|sort |uniq |sed -nr ' H;${x; s/\n/ /g;p}'|tr [a-z] [A-Z]`

cache=`grep "cache size" $file |sort |uniq |awk '{print $4}'|sed -nr ' H;${x; s/\n/ /g;p}'`

let cache=cache/1024

cache=$cache"M"

stepping=`grep stepping /proc/cpuinfo |uniq |sed 's/ //g' |awk -F: '{print $2}'|sed -nr ' H;${x; s/\n/ /g;p}'`

{

echo "cpu_number=\"$num\""

echo "cpu_core=\"$corenum\""

echo "cpu_type=\"$model\""

echo "cpu_cache=\"$cache\""

echo "cpu_clock=\"$clock\""

echo "cpu_stepping=\"$stepping\""

} >/tmp/cpu.txt

for i in {a..l}

do

if [ -b /dev/sd$i ];then

/usr/sbin/smartctl -a /dev/sd$i >hdinfo.$$

model=`grep Model hdinfo.$$ |sed 's/: */:/g' |awk -F: '{print $2}'`

fw=`grep Firmware hdinfo.$$ |sed 's/: */:/g' |awk -F: '{print $2}'`

serial=`grep Serial hdinfo.$$ |sed 's/: */:/g' |awk -F: '{print $2}'`

cap=`grep User hdinfo.$$ |sed 's/: */:/g' |sed 's/,//g' |awk -F: '{printf("%dG\n",$2/1000/1000/1000)}'`

smart=`grep "SMART support is: Enabled" hdinfo.$$`

smartlog=`grep "No Errors Logged" hdinfo.$$`

fi

done

{

echo "hdmodel=\"$model\""

echo "hdfw=\"$fw\""

echo "serial=\"$serial\""

echo "hdcap=\"$cap\""

echo "smart=\"$smart\""

echo "smartlog=\"$smartlog\""

} >/tmp/hd.txt

sysmem=`grep MemTotal /proc/meminfo |awk '{print $2}'`

sysmem=`echo "scale=2;$sysmem/1024/1024"|bc`

totalmem="$sysmem"

{

meminfo=`/usr/sbin/dmidecode |grep -A16 "Memory Device$" | sed -e '/./{H;$!d;}' \

-e 'x;/Memory Device\n/!d;' \

-e 's/: /:/g' \

-e 's/</{/g' \

-e 's/>/}/g' \

-e 's/[ \t]*\n/\n/g' \

| awk -F: '/Size|Form.Factor|[ ^]Locator|[\t]Speed|Serial|Manufacturer/{printf("|%s", $2)}/Part/{print "|" $2}'\

| sed -e 's/No Module Installed/{EMPTY}/' \

| awk -F'|' '{printf("|%s|%s|%s|%s|%s|%s|%s\n", $4, $7, $6, $3, $5,$2,$8)}'|sort |grep -v -E "EMPTY|BIOS"`

echo "${meminfo}" | awk -v stat="${stat}" -v loca="${loca}" -v manuf="${manuf}" -F "|" '

{

j=j+1

for (i=1;i<=NF;i++ )

{arr[j,i]=$i

if (i==2)

locator=(locator""$i)

#print arr[1,4]

}

END{

sw=0

for (i=4;i<=7;i++)

{

a=arr[1,i]

for (y=1;y<=j;y++)

{ if (a!=arr[y,i])

sw=1

}

if (sw==0)

{

print "mem="arr[1,7]

print "manuf="arr[1,4]

print "loca="locator

}

'

} >>/tmp/mem.check

cat >/tmp/meminfo.txt<<EOF

totalmem="$sysmem"

manuf="`cat /tmp/mem.check|awk -F= '/manuf/{print $2}'`"

mem="`cat /tmp/mem.check|awk -F= '/mem/{print $2}'`"

loca="`cat /tmp/mem.check|awk -F= '/loca/{print $2}'`"

EOF

/usr/bin/ipmitool sdr|sed 's/ //g'>/tmp/sdr1.log

cat >/tmp/sdr.txt <<EOF

CPUSensorNum=`cat /tmp/sdr1.log|grep CPU|grep ok|wc -l`

CPUSensor="`cat /tmp/sdr1.log|grep CPU|sed 's/ //g'|grep ok|sort|awk -F"|" '{for(j=1; j<=NF; j++); printf("%s "), $1}'`"

MemSensorNum=`cat /tmp/sdr1.log|grep Mem|grep ok|wc -l`

MemSensor="`cat /tmp/sdr1.log|grep Mem|sed 's/ //g'|grep ok|sort|awk -F"|" '{for(j=1; j<=NF; j++); printf("%s "), $1}'`"

PSUSensorNum=`cat /tmp/sdr1.log|grep PSU|grep ok|wc -l`

PSUSensor="`cat /tmp/sdr1.log|grep PSU|sed 's/ //g'|grep ok|sort|awk -F"|" '{for(j=1; j<=NF; j++); printf("%s "), $1}'`"

OtherSensor="`cat /tmp/sdr1.log|grep -v CPU|grep -v Mem|grep -v PSU|sed 's/ //g'|grep ok|sort|awk -F"|" '{for(j=1; j<=NF; j++); printf("%s "), $1}'`"

EOF

#cat /tmp/meminfo.txt >>/opt/jila/scripts/paizhi.txt

cat /tmp/meminfo.txt >>/tmp/paizhi.txt

#cat /tmp/cpu.txt >>/opt/jila/scripts/paizhi.txt

cat /tmp/cpu.txt >>/tmp/paizhi.txt

#cat /tmp/sdr.txt >>/opt/jila/scripts/paizhi.txt

cat /tmp/sdr.txt >>/tmp/paizhi.txt

cat /tmp/hd.txt >>/tmp/paizhi.txt

rm /tmp/mem.check -rf

rm /tmp/hd.txt -rf

rm /tmp/cpu.txt -rf

rm /tmp/meminfo.txt -rf

rm /tmp/sdr.txt -rf

3), hd.sh scripted code:

#!/bin/bash

source /tmp/paizhi.txt

function test ()

{

if [ "$2" = "$3" ];then

echo -e "\033[42m "$1" is "$3" .....................PASSED \033[0m"

else

echo -e "\033[41m "$1" current is "$3",Needed is "$2" .....................FAILED \033[0m"

fi

}

for i in {a..l}

do

if [ -b /dev/sd$i ];then

# obtains the model of hard disk

/usr/sbin/smartctl -a /dev/sd$i >hdinfo.$$

model=`grep Model hdinfo.$$ |sed 's/: */:/g' |awk -F: '{print $2}'`

#ST33000650NS

fw=`grep Firmware hdinfo.$$ |sed 's/: */:/g' |awk -F: '{print $2}'`

# obtains the FW of hard disk

serial=`grep Serial hdinfo.$$ |sed 's/: */:/g' |awk -F: '{print $2}'`

#Z293W51M

# obtains the capacity of hard disk

smart=`grep "SMART support is: Enabled" hdinfo.$$`

if [ $? = 0 ] ;then

echo -e "\033[42m HD /dev/sd$i Smart is Enabled..................Passed \033[0m"

else

echo -e "\033[41m HD /dev/sd$i Smart Not Open........................FAILED \033[0m"

read

fi

smartlog=`grep "No Errors Logged" hdinfo.$$`

if [ $? = 0 ] ;then

echo -e "\033[42m HD /dev/sd$i No Error Logged..................Passed \033[0m"

else

echo -e "\033[41m HD /dev/sd$i Error Logged........................FAILED \033[0m"

cat hdinfo.$$

read

fi

# obtains the state of hard disk self-inspection

#stat=`grep self-assessment hdinfo.$$ |sed 's/: */:/g' |awk -F: '{print $2}'`

test "HD sd$i Model" "${hdmodel}" "${model}"

test "HD sd$i Capacity" "$hdcap" "$cap"

test "HD sd$i Firmware" "$hdfw" "$fw"

rm hdinfo.$$

else

echo -e "\033[;41m ----sd$i not found----........................FAILED \033[0m"

read rrrr

fi

done

Operation steps:

1. build test platform.

2. linux or CentOS operating system is installed, and copies testing tool to test machine.

3. successively run check.sh, hd.sh.

4. run the number of times amendment reboot.sh of restarting according to actual needs, and place it in root/.bashrc.

5. start test.

6. check and collect test result.

This detects the method for hard disk exception automatically, is suitable for the red cap of x86 framework and Centos operating system 6 series.Utilize smartctl order under Linux can check the function of hard disk model, Firmware version, amount of capacity, health status etc., write test script: first hard disk relevant information read before test and be automatically recorded in certain text, restart server at every turn and all can rerun this script, detect all information of hard disk and make comparisons with recorded information before, if there be abnormal will automatically reporting an error.Does is just can find according to error information which block hard disk has problem? it is which type of problem? and then be further analyzed and verify.This script can be arranged restarts number of times, also can automatically stop test according to prompting, when there being problem to occur, having red prompting, and automatically stopping test.

Claims

1. automatically detect a method for hard disk exception, it is characterized in that the step of the method is:

2. automatically detect a method for hard disk exception, it is characterized in that:

3. the method for a kind of automatic detection hard disk exception according to claim 2, it is characterized in that: before starting test, write test script, first hard disk relevant information read before test and be automatically recorded in text, restart server at every turn and all can rerun this test script, detect all information of hard disk and make comparisons with recorded information before, if there be abnormal will automatically reporting an error, entry condition is selected.

4. the method for a kind of automatic detection hard disk exception according to claim 3, is characterized in that: