nagios监控raid下磁盘和raid状态脚本实现

大家好，欢迎来到IT知识分享网。

Linux下服务器做了硬件raid之后，磁盘的状态比较难定位，windows则可以通过MegaRAID来检测，此脚本通过MegaCli来达到定位raid下哪块磁盘是坏块的功能，在nagios上面可以实现通过定期通过检测以邮箱或者短信等形式，来达到预警的功能，脚本在几台物理机上面测试过，是没问题的，分享给各位，也希望大家能相互讨论，学习。

一、安装Megacli:

 rpm-ivh megacli-8.00.46-2.x86_64.rpm

二、添加脚本到nagios监控：

执行visudo,然后在文件中root ALL=(ALL) ALL下面加入如下一行:

 nagios ALL=(ALL)NOPASSWD:/usr/local/nagios/libexec/check_raid.sh

并注释以下一行

 #Defaults requiretty

把脚本放在/usr/local/nagios/libexec目录下,chmod +x check_raid.sh,赋予x权限,并编辑
/usr/local/nagios/etc/nrpe.cfg加入

command[check_raid]=/usr/bin/sudo/usr/local/nagios/libexec/check_raid.sh

重启nrpe(根据安装方式的不同，可能有差异)

#pkill nrpe #/usr/local/nagios/bin/nrpe -c/usr/local/nagios/etc/nrpe.cfg -d

三、监控脚本说明：

#!/bin/sh #Program: # for monitor raid disk state #history: #------ First release #检测是否是LSI卡 rcexist=`dmesg| grep RAID | grep LSI` if [ ! -n"$rcexist" ]; then echo "not LSI or no raid" exit 2 fi OUTPUT='' #判断raid类型 R1=`/usr/sbin/MegaCli-cfgdsply -aALL | grep "RAID Level" |awk -F: '{print $2}' | sed -e"s/^*//" | grep -c "Primary-1, Secondary-0, RAID LevelQualifier-0"` R0=`/usr/sbin/MegaCli-cfgdsply -aALL | grep "RAID Level" |awk -F: '{print $2}' | sed -e"s/^*//" | grep -c "Primary-0, Secondary-0, RAID LevelQualifier-0"` R5=`/usr/sbin/MegaCli-cfgdsply -aALL | grep "RAID Level" |awk -F: '{print $2}' | sed -e"s/^*//" | grep -c "Primary-5, Secondary-0, RAID LevelQualifier-3"` R10=`/usr/sbin/MegaCli-cfgdsply -aALL | grep "RAID Level" |awk -F: '{print $2}' | sed -e"s/^*//" | grep -c "Primary-1, Secondary-3, RAID LevelQualifier-0"` if [ $R1-ge 2 ];then OUTPUT+="RAID10 " elif [ $R1-eq 1 ];then OUTPUT+="RAID1 " fi if [ $R0-ne 0 ];then OUTPUT+="RAID0 " fi if [ $R5-ne 0 ];then OUTPUT+="RAID5 " fi if [ $R10-ne 0 ];then OUTPUT+="RAID10 " fi #以上的if是根据资料和实际情况做了微调 #raid下面总的磁盘数 DiskNum=`/usr/sbin/MegaCli-cfgdsply -aALL | grep -c "Non Coerced Size"` OUTPUT+="TotalDisk:$DiskNum" #处于raid中的正常的盘数 OnlineDisk=`/usr/sbin/MegaCli-cfgdsply -aALL | grep "Online" | wc -l` OUTPUT+="online: $OnlineDisk" if [$DiskNum -ne $OnlineDisk ];then echo "CRITICAL:$OUTPUT" exit 2 fi #是否有坏的盘 FailDisk=`/usr/sbin/MegaCli-AdpAllInfo -aALL | grep "Failed Disks" | awk '{print $4}'` if [$FailDisk -eq 0 ];then OUTPUT+=" failed disk:0 " else OUTPUT+=" failed disk:$FailDisk" echo "CRITICAL: $OUTPUT" exit 2 fi #预警的盘以及位置 CriticalDisk=`/usr/sbin/MegaCli-AdpAllInfo -aALL | grep "Critical Disks" | awk '{print $4}'` if [$CriticalDisk -eq 0 ];then OUTPUT+="critiDisk is 0" else CriDisk=`/usr/sbin/MegaCli -cfgdsply -aALL| grep -E 'Predictive|Slot' | awk \ '{if(NR%3){printf$0":"}else{print $0}}'|awk -F':' '{if($4!=0){print $2+1}}'` OUTPUT+=" critidisk in $CriDiskslot" echo "WARNING: $OUTPUT" exit 1 fi #MediaErrcount检测坏块和哪块盘 MediaErrcount=`/usr/sbin/MegaCli-pdlist -aALL | grep -E "Media Error" |awk -F’:’ -v errcount=0 \ '{errcount+=$2}END{printerrcount}'` OtherErrcount=`/usr/sbin/MegaCli-pdlist -aALL | grep -E "Other Error" |awk -F’:’ -v errcount=0 \ '{errcount+=$2}END{printerrcount}'` #坏盘的位置 if [ $MediaErrcount-ne 0 -o $OtherErrcount -ne 0 ];then mDoD=`/usr/sbin/MegaCli -pdlist -aALL |grep -E "Media Error|Other Error|Slot" | awk \ '{if(NR%3){printf$0":"}else{print $0}}' | awk -F':' '{if($4!=0||$6!=0){print $2+1}}'` OUTPUT+=" bad block in $mDoD" echo "CRITICAL: $OUTPUT" exit 2 else OUTPUT+=" mediaerr:0 othererr:0" fi #raid状态是否正常 raidstate=`/usr/sbin/MegaCli-LDInfo -Lall -aAll | grep 'State' |awk -F':' '{print $2}' | \ sort |uniq | sed -e "s/^*//" | awk '{if($0 != "Optimal"){print"bad"}}'` if ["$raidstate" != "bad" ];then OUTPUT+=" raidstate:ok" else OUTPUT+=" raidstate:bad" echo "CRITICAL: $OUTPUT" exit 2 fi rm -rf./MegaSAS.log echo$OUTPUT

检测结果如下：

 RAID5 Total Disk: 4 online: 4 failed disk:0 critidisk is 0 mediaerr:0 othererr:0 raidstate:ok