4

As the title says, I have under my care a SuperMicro server running a hardware RAID 6 with a MegaRAID SAS 8708ELP on Linux. I have two failed disks that I have identified as the drives in slot 2 and 4, howerver, my boss tells me that the drives are somewhat scrambled and since the slots start in 0, I'm not sure which hard disks I'm supposed to be changing. I tried to find out the serial number of the drives so I could compare with the drives I remove, but all the commands I've been running to find such info haven't been working.

I tried using

#smartctl -d sat+megaraid,2 -a /dev/sda
#smartctl -d megaraid,2 -a /dev/sda

But both only give an invalid argument message.

I've tried running lshw but it only gives me the info of the Raid card itself.

And when I ran megacli -PDList -aAll I can identify my disks, but I'm not sure what is useful for what I need from all the info.

I'm guessing I only need to remove the 3rd and 5th disk from the server, but I just wanted to be absolutely sure of this before doing it.

I'd appreciate any advice, since I'm not really an IT guy.

Thanks.

TuturuJones
  • 41
  • 1
  • 2
  • Before you do **anything** further. Ensure you have a solid backup of the data on this server. There are a thousand ways you could manage to mess this up, and even if you do everything perfectly, there's a non-trivial chance of you having a third drive failure during the lengthy rebuild you'll need to go through. – EEAA Jan 23 '16 at 01:42
  • run `./lsi.sh drives` - the script is here https://calomel.org/megacli_lsi_commands.html – Marki555 Jul 01 '16 at 12:30

1 Answers1

7

Using storcli, the command is:

storcli /c0/eALL/sALL show all | egrep 'Device attributes|SN = '

will print something like:

Drive /c0/e252/s0 Device attributes :
SN =      WD-WMC1T4142955

The first line gives you the enclosure and slot, the second line is the serial number of the disk.

Aaron Digulla
  • 954
  • 1
  • 13
  • 24