5

I've a failed drive reported in my RAID5 running on a Dell R710 using RHEL5 with a DELL PERC controller. Is there a method to determine which drive it is? Can I do this at the command line?

ewwhite
  • 194,921
  • 91
  • 434
  • 799
dacracot
  • 469
  • 2
  • 12
  • 28
  • I'd normally look at the server's OMSA page at `https://server.ip:1311`, But that depends on the Dell agents being installed. Do you have the management agents installed? – ewwhite Jun 07 '12 at 22:24
  • There are no management agents installed. – dacracot Jun 07 '12 at 22:34

4 Answers4

4

I know this isn't the quickest solution, but I'd recommend downloading the OMSA management utilities to the system and using the status page I mentioned earlier at http://server.ip:1311 or using the omreport command line utility to obtain the RAID array's status.

This is assuming that you don't have any physical access to the system. If you do, there should be an amber light illuminated on the failed drive.

ewwhite
  • 194,921
  • 91
  • 434
  • 799
  • `omreport storage controller` to get a list of controllers, then `omreport storage pdisk controller=0`, or a more appropriate controller number. YMMV as this was done on a PE2950. Due to the system age, I used OMSA_5.4.1 instead of the latest OMSA_7.1 so some differences in commands could exist. Installing applicable packages from srvadmin-* folders beneath the most compatible OS subfolder of http://linux.dell.com/repo/hardware/2012_Q2/platform_independent/ and starting the dataeng service should get one going. Packages installed fine for me on an operating system not listed as supported. – kbulgrien Oct 19 '12 at 17:32
3

The orange light on the caddy should be illuminated on the bad drive. If you don't have management agents installed, this is the best way. Good old visual inspection.

MDMarra
  • 100,183
  • 32
  • 195
  • 326
1

If nothing else, you can try looking for a serial number:

smartctl -a /dev/sdX

where /dev/sdX is a failed device. You'll see a device serial number (near the top). The serial number should also be written on the sticker on the drive itself.

ps: you might need to install "smartmontools" to get "smartctl"

mulaz
  • 10,472
  • 1
  • 30
  • 37
  • oh :/ no idea then :( – mulaz Jun 07 '12 at 22:27
  • # smartctl -a /dev/sda smartctl version 5.38 [x86_64-redhat-linux-gnu] Copyright (C) 2002-8 Bruce Allen Home page is http://smartmontools.sourceforge.net/ Device: DELL PERC H700 Version: 2.10 DELL PERC controllers are not supported. – dacracot Jun 07 '12 at 22:28
  • @dacracot yeah, if you're using Dell PERC (or pretty much any hardware raid card) the `sdX` devices are going to be virtual drives and not the actual spinning disks. Some drivers can allow access to the spinning disks via `sg#` devices, but I don't think the H700 supports it. – DerfK Jun 08 '12 at 03:02
1

You can run the omreport from the command line to find failed disk on Dell servers. For example to display all physical disks on controller 1, run: 

omreport storage pdisk controller=1

References: http://stuff.mit.edu/afs/athena/dept/cron/documentation/dell-server-admin/en/Dosa/CLI/report.htm#wp1068065 http://linux.dell.com/wiki/index.php/OpenManage_command_line_interface

user189846
  • 11
  • 1