文档介绍：----------------------------------------------------------------------------------------------------------------------1:对于Dell的机器并且配备了LSI的raid卡,那么Megacli自带的命令是最直观最可靠的硬盘检测方法。#/opt/MegaRAID/MegaCli/MegaCli64-pdlist-aall以其中的一块硬盘为例。EnclosureDeviceID:32SlotNumber:2Enclosureposition:0DeviceId:2//ID号SequenceNumber:2MediaErrorCount:0//如果error数值较多,那么需要更换硬盘OtherErrorCount:0PredictiveFailureCount:0//可理解为预知性的错误,这个数值较多意味着硬盘即将坏掉,现场会黄绿灯交叉闪烁LastPredictiveFailureEventSeqNumber:0PDType:SASRawSize:[0x11177328Sectors]NonCoercedSize:[0x11077328Sectors]CoercedSize:[0x11040000Sectors]Firmwarestate:Online,SpunUp//如果硬盘坏了那么会出现Failed或者是unconfigured(bad)SASAddress(0):0x500000e11716a112SASAddress(1):0x0ConnectedPortNumber:0(path0)InquiryData:FUJITSUMBD2147RCD809D0A4PB103ALAFDECapable:NotCapableFDEEnable:DisableSecured:UnsecuredLocked:UnlockedNeedsEKMAttention:NoForeignState:NoneDeviceSpeed:::HardDiskDeviceDriveTemperature:28C()一些对磁盘I/O操作比较密集的应用例如squid经常会有掉盘的现象,这时有时候通过上述命令会发现找不到其中的一块硬盘,即“消失了”,如果进入RAID卡界面会发现磁盘处于Foreign或者missing的状态,这种情况其实硬盘并没有坏掉或者说并没有完全坏掉(有诸多原因,包括RAID卡、背板、SAS线等等),如果是在线跑业务的机器不能重启,那么这时候就需要通过RAID卡的日志进行分析是否真的坏了。#/opt/MegaRAID/MegaCli/MegaCli64-fwtermlog-dsply-aall如果是硬盘坏了那么会发现有offline或者是某块硬盘有timeout的记录,需要认真分析日志。2:对于HP的机器也可通过hpacucli命令判断Hp的机器硬盘故障较为直观,如果是坏掉了那么status部分会直接显示为failed。#hpacuclictrlallshowdetailconfigSmartArrayP410iinSlot0(Embedded)//raid卡型号BusInterface:PCISlot:0SerialNumber:500**********F0CacheSerialNumber:PBCDH0CRH0E7UFRAID6(ADG)Status:DisabledControllerStatus:OK//控制器状态是ok的ChassisSlot:HardwareRevision:RevCFirmwareVersion::MediumExpandPriority:MediumSurfaceScanDelay:15secsSurfaceScanMode:IdleQueueDepth:AutomaticMonitorandPerformanceDelay:60minElevatorSort:EnabledDegradedPerformanceOptimization:DisabledInconsistencyRepairPolicy:DisabledWaitforCacheRoom:DisabledSurfaceAnalysisInconsistencyNotification:DisabledPostPromptTimeout:0secsCacheBoardPresent:TrueCacheStatus:eleratorRatio:25%Read/75%Writ