OS2 World.Com Forum
2012.02.10, 02:48:17 *
Welcome, Guest. Please login or register.

Login with username, password and session length
News:
 
   Home   Help Search Login Register  
Pages: [1]
  Print  
Author Topic: RAID 5 Recovery - Dell PowerEdge 4400 - Perc 3Di  (Read 2802 times)
NNYITGuy
Jr. Member
**
Posts: 85

IT Consultant - eCS & W32 Solutions for Businesses


View Profile
« on: 2010.02.23, 13:20:24 »

Hi,

Does anyone have any ideas for recovering a RAID 5 container?

The client had physical disk 3 of 4 fail out in a RAID 5 container.  The client then used a physical disk from another identical spare system.  The physical disk was labeled as a member of 0 of a RAID 1 container.  Unknown by the client, this apparently, when initialized wipes out the other containers on the PERC controller.

I have exhausted my internal technical resources on how to retrieve the RAID 5 data.  Can someone give me a path forward?

Very Greatful,
-Todd
Logged

Todd J Simpson (IT Consultant)
Business Automation Technologies
(Professional Products and Services.  Business to Business...)
rwklein
Full Member
***
Posts: 142


View Profile
« Reply #1 on: 2010.02.23, 20:58:28 »

Is he is realy screwed and can't do anything to recover the container, you could consider contacting maybe this company. www.retrodata.co.uk

I don't know if this is an OS/2 box you want recovered they just received a NFR copy of eComStation to look how they can support OS/2/eCS.

Roderick Klein
Mensys
Logged
RobertM
Global Moderator
Hero Member
*****
Posts: 1936



View Profile WWW
« Reply #2 on: 2010.02.23, 21:46:17 »

Unfamiliar with those cards, but on some, you can re-read the RAID configuration from the drives themselves. That would require removing the replacement drive first.

Additionally, it is always best to wipe a replacement drive before putting it in a system. That just requires a SCSI controller with that feature, or the use of certain RAID tools on controllers that support it.
Logged

|
|
Kirk's 5 Year Mission Continues at:
Star Trek New Voyages
|
|

Sev0600
Newbie
*
Posts: 2


View Profile
« Reply #3 on: 2010.02.24, 02:50:13 »

Sound like the data might be gone or corrupt at this point. You can always call Dell Support on this, they will take your call even though the server is out of warranty.

If the disks were not initalized, you should be able to clear the controllers current configuration, recreate the RAID 5 array, save the new configuration, then reboot without initializing disks.

I've just successfully tested this using a Dell Power Connect Cache Server with 2 drives configured as RAID 1 and an old Dell workstation that has a PERC II controller in it.
Logged
RobertM
Global Moderator
Hero Member
*****
Posts: 1936



View Profile WWW
« Reply #4 on: 2010.02.24, 04:45:27 »

Sound like the data might be gone or corrupt at this point. You can always call Dell Support on this, they will take your call even though the server is out of warranty.

If the disks were not initalized, you should be able to clear the controllers current configuration, recreate the RAID 5 array, save the new configuration, then reboot without initializing disks.

I've just successfully tested this using a Dell Power Connect Cache Server with 2 drives configured as RAID 1 and an old Dell workstation that has a PERC II controller in it.

In his particular case, wont this cause a problem if he does not clear the new drive, or actually (the better option?) reinstall the old one? If he recreates the RAID configuration with the new (unwiped) drive in it, how will it know which drive was bad in order to rebuild the array correctly?

Asking because I am not sure, as I have not ever had issues on the rare few machines I have handled that had the PERC II controllers.
Logged

|
|
Kirk's 5 Year Mission Continues at:
Star Trek New Voyages
|
|

Sev0600
Newbie
*
Posts: 2


View Profile
« Reply #5 on: 2010.02.24, 06:21:29 »

No, it should not cause a problem assuming disk 3 is in a failed state when he inserts the new disk.

In this particular case, I would recreate the array with the old drive (forcing the failed disk online if it shows failed), save the configuration then reboot. The server should boot even if the disk fails again (may prompt to press F1 to continue or F2  for setup).
Logged
mobybrick
Full Member
***
Posts: 104



View Profile
« Reply #6 on: 2010.02.25, 12:31:04 »

This is going to sound daft, but I wonder... did your colleague pull out the wrong disk?  When disks go bad on the PERC, the orange failed light should flash. Was this the disk that he pulled? If he pulled it when the system was off, then he might have pulled the wrong one. If he did, then the container will have been destroyed and needing re-initializing.

On some of the dell chasis, the discs are labelled 1...8 from the right hand side... but of course, in the RAID bios the disks are labelled 0...7.

PERC2s sometimes flagged as disk as 'bad' inadvertantly, but this was fixed in the PERC3 which is based on a different family/manufacturer of chipset.

Regards,
Moby
Logged
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.14 | SMF © 2006-2011, Simple Machines LLC
Joomla Bridge by JoomlaHacks.com
Valid XHTML 1.0! Valid CSS!