Want to protect your cyber security and still get fast solutions? Ask a secure question today.Go Premium

x
?
Solved

Intel RAID - All Disks Ofline.

Posted on 2009-12-23
18
Medium Priority
?
655 Views
Last Modified: 2016-12-08
Hi all,
Major panic today. Reinstalled a tape drive last Friday at a site, & later noticed that the Intel Storage manager was reporting that a drive had failed.
I suspected maybe a dislodged connection etc.

Powered the system down a while ago & checked all connections & restarted & now NO GO.

When I go into the Storage Manager BIOS Settings it says no RAID Volumes Defined.

It Lists all 4 Drives as "Offline Members".

Their Backups are less than perfect so really need to get this array back up.

All this is on an S3210SHLX Intel board with 4 drives connected as RAID10.
ROM Version 5.0.1017 is mentioned in the BIOS.

I read this which really made my heart sink:-
http://download.intel.com/support/motherboards/server/ta_445.pdf
The bit on page 3 about DO NOT REBOOT THE SERVER.

4pm on Christmas Eve.......

Help greatly appreciated

Thanks

Phil

0
Comment
Question by:PhilPorritt
  • 10
  • 7
18 Comments
 
LVL 10

Expert Comment

by:wmeerza
ID: 26117539
Phil,
this is not looking good. recovery procedure 1 on page 6 of the link below seems to be your only option.
http://download.intel.com/support/motherboards/server/RAID_vol_recover_proc.pdf



0
 

Author Comment

by:PhilPorritt
ID: 26117674
Thanks wmeerza,

Looked at that doc a little whie ago.
Currently trying to determine if I have that CD or where I can obtain it.
Motherboard shipped with one called "Intel Derver Deployment Toolkit 2.0"

Tried to boot it in one of my own servers (Identical M/B & array setup), but then I remembered ho these boards seem to refuse to boot from the DVD when a RAID array is configured.

That's something I discovered earlier this year & have not yet had to resolve.

Phil
0
 
LVL 5

Expert Comment

by:dmbgo
ID: 26118770
Hi Phil, when will you be working on this next? If the data is important the first step will be to image the drives, and then attempt recovery from the image, that way you wont be risking further damage to the data. All hope is not lost, but there is a fair amount of time involved.
Merry Xmas

Dave
0
Who's Defending Your Organization from Threats?

Protecting against advanced threats requires an IT dream team – a well-oiled machine of people and solutions working together to defend your organization. Download our resource kit today to learn more about the tools you need to build you IT Dream Team!

 

Author Comment

by:PhilPorritt
ID: 26120084
Thanks Dave,

Got it back at my Home office over the weekend. Imaging the drives is the first issue. Was preparing to rob drives from 9 new systems that arived yesterday, only to realise they were only supplied with 250 Gig & I need 4 by 500 Gig.

Thought of picking up externals from a retailer on the weekend, but then realised I would still be working with the originals.

Ghost Images to USB Drives, would that be considered safe or stupid?

Phil
0
 
LVL 5

Accepted Solution

by:
dmbgo earned 2000 total points
ID: 26120895
Hi Phil, I will be working as well over the weekend, so I should be around to help. What you have at the present is two raid 0 stripe sets (raid 10), one of which has a failed member. Since the system was working at power down, it is highly likely that two of the drives contain the data, with hopefully only minor corruption. As I said yesterday the first step is to image the drives, if they are 500 gb each, you could grab a couple of 1tb drives from a pc shop, or Dick Smith etc. It doesn't matter if they are USB, but it would be faster if you had them on a sata interface.
Another important thing, do you have the original raid parameters such as stripe size? This makes it easier to rebuild from images, since the recovery process doesn't involve guesswork an trial and error. You may be able to get these from the raid controller.
I suggest you purchase Raid reconstructor for this job, you can use the trial version to determine the extent of the damage, then pay the $99.00 US to recover the data. You can find the software at: http://www.runtime.org/raid.htm 
Ghosting to USB drives probably wont be necessary as the Raid reconstructor software will create the images for you, then reconstruct from the images, leaving the original drives alone. It doesnt matter if the drives are usb or not.

Cheers
Dave
0
 

Author Comment

by:PhilPorritt
ID: 26120943
Thanks again Dave,

Should I be making 4 clones & trying to rebuild first on them?
Or should I be backing up the drives to image files on the other drives I
find.

Harvey Norman is as good as it gets here. Might be open to-morrow.
This document refers to ARCU on page 6, but I can't find much info on that.
Not a very current document either, Dec 2001.

http://www.intel.com/support/motherboards/server/sb/CS-010767.htm

Have also read clearing the CMOS & reconnecting the drives one at a time may work,
but don't even want to go there till I've imaged them.

Phil
0
 
LVL 5

Expert Comment

by:dmbgo
ID: 26120988
Yes you should try to rebuild from drive images first. You probably only need two of the cloned drive to rebuild from, since raid 10 is a mirror of two raid 0 stripe sets, however you have to have the right two drive images.  
Do you know which drive failed originally? If so you can exclude that from any attempts to rebuild. If you try to rebuild with the failed drive in the mix you will almost certainly end up with corrupt data.

After giving it a bit more thought (and reading the intel document) I see where you are going with this. I have been giving you advice that is probably one step further down the track. I think we should proceed as follows.

1: Ghost the original 3 drives that the system was running on when powered down, leaving the original failed drive alone (assuming that you know which one this was)

2: Obtain the Raid CD from intel (I'll have a look for this)

3: Put the 3 drives (clones preferably) back into the server and use the ARCU to mark the drives back online.

4: Replace the failed drive once the array is up and running.

Here is a link to the ARCU
http://downloadcenter.intel.com/detail_desc.aspx?ProductID=369&DwnldID=1175
0
 
LVL 5

Expert Comment

by:dmbgo
ID: 26120996
Just a bit more, dont use the ARCU to mark all 4 of the original drives online, if you dont know which was the original failed drive, or you will definitely get corrupt data.
The goal is to get the 3 drives online that were running prior to the last shutdown, which will create a failed, but running array, then rebuild from there.

Cheers
Dave
0
 

Author Comment

by:PhilPorritt
ID: 26121060
It was the drive on Port0 which appeared failed. Cloning the drives will be the issue. We a a small regional community (Pop 10,000), so getting drives may be a bit difficult at this time of year. 3.5" external USB drives may be a source of drives if I can get 4 to-morrow & disassemble them. That's assuming they have SATA drives in them, presume they would.

Assistance is greatly appreciated.
Thanks Very Much Dave.
0
 
LVL 5

Expert Comment

by:dmbgo
ID: 26121108
In that case you probably only need 3 drives, you could just put the original one from port 0 aside.
Cheers
Dave
0
 

Author Comment

by:PhilPorritt
ID: 26121185
Thanks Again Wmeerza:

Was just reading one of your older posts the links to another document that may be of help:-

http://www.experts-exchange.com/Storage/Storage_Technology/Q_23506499.html

ftp://download.intel.com/support/motherboards/server/srcsas18e/sb/d29305_005_raid_swg.pdf

Not sure of it's relevance but I will have a read.

Phil
0
 

Author Comment

by:PhilPorritt
ID: 26135096
Still at a standstill with this job. Hoping to obtain a couple of 1T drives later today, create Ghost image files for 2 of each drives on each of the 1000Gig's. That's the only option I have up here unless I wait another week to get some 500Gigs.
Ghost 11.5 is what I've got on hand & I'm sure if options are set correctly It can create images to files.
Any better & alternative suggestions are welcome.
Thinking about using Raid Reconstructor also incase the array cant be recovered, presume I'd have to attache the drives to another machine as this one thinks they are assigned to an array but offline.
Presume I could do this with one of the new systems I have on hand at the moment, either to an image file or to a new drive which may be bootable.
Phil
 
0
 
LVL 5

Expert Comment

by:dmbgo
ID: 26136815
I think even the trial version rr will create images for you. Make sure you know what the original raid parameter were (stripe size etc) Once you have the drives safely imaged, I personally would be tempted to put the 3 working original drives back in the original system and force them online. It is likely that all will be well then.

BUT and this is a definite but, only if you are very sure which were the three working drives when you shut the system down.
 under no circumstances force all 4 drives on line at once or data chaos will prevail.
0
 

Author Comment

by:PhilPorritt
ID: 26140712
Sure Drive0 was the problem drive. Making recoverys with RR ath the moment. Just not 100% sure which drive is Drive0's partner in it's Stripe set. 2 & 3 both seem valid.
 
0
 

Author Comment

by:PhilPorritt
ID: 26140912
Forcing drives back can only be done with ARCU? Or are there other ways? Did manage to get ARCU on a DOS formatted XP floppy last night, but still doing my other recoverys first.
Posted another Question Re Which drives are partners in the Array.
0
 
LVL 5

Expert Comment

by:dmbgo
ID: 26141180
the original raid controller should know which drives are partners, so as long as you dont put the drive that has been offline for longer than the other 3 back in you should be right. Yes the arcu is the only way I know of. It was raid 10 so it was probably was disk 0 +1 mirrored and disk 2+3 mirrored.

Cheers
Dave
0
 

Author Comment

by:PhilPorritt
ID: 26141973
Hmm
ARCU doesn't seem to do much. Just ran it on my test server & it says "No IOPs found". Was expecting to see it's 4 drives & options to deal with them, but nothing. Just the heading:-
"Integrated RAID Configuration Menu" followed by "No IOPs found'.
Same system board & Raid setup & fully functional.
Did see other suggestions around the place saying to Reset BIOS & boot with 3 of the 4 drives connected. Seemed to be various versions of this method with slight variations.
ARCU's failure didn't suprise given it's revision date.
Phil
0
 

Author Closing Comment

by:PhilPorritt
ID: 31675782
Using Raid Reconstructor was a real savior in this case as  securing the data before proceeding with any other approaches was the most crucial issue due to failed backup routines.

It turned out the array came back on-line with the failed drive removed. Unfortunately the OS would not boot & required a rebuild.
All data remained intact & the business became operational on the array minus the failed drive which will be replaced in due course.
0

Featured Post

Who's Defending Your Organization from Threats?

Protecting against advanced threats requires an IT dream team – a well-oiled machine of people and solutions working together to defend your organization. Download our resource kit today to learn more about the tools you need to build you IT Dream Team!

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Steps to fix “Unable to mount database. (hr=0x80004005, ec=1108)”.
Steps to fix error: “Couldn’t mount the database that you specified. Specified database: HU-DB; Error code: An Active Manager operation fail”
This tutorial will walk an individual through the steps necessary to configure their installation of BackupExec 2012 to use network shared disk space. Verify that the path to the shared storage is valid and that data can be written to that location:…
Despite its rising prevalence in the business world, "the cloud" is still misunderstood. Some companies still believe common misconceptions about lack of security in cloud solutions and many misuses of cloud storage options still occur every day. …
Suggested Courses

564 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question