Otdfbt

I have re-inserted a RAID disk, on a DELL server with Windows Server 2008. The drive-status indicator was changing between a green and amber light, and the monitor gave the following message:

 There are offline or missing virtual drives with preserved cache.
 Please check the cables and ensure that all drives are present.
 Press any key to enter the configuration utility.

I pressed a key and the PERC 6/I Integrated BIOS Configuration Utility showed that the RAID Status for that disk was Offline.

After reinsertion of the disk the monitor is giving the following message:

 Foreign configuration(s) found on adapter.
 Press any key to continue or ‘C’ load the configuration utility,
 or press ‘F’ to import foreign configuration(s) and continue.

After checking around on the net I am uncertain if I should choose import or clear. I cannot find out if an import means importing information from the array/system to the now foreign disk or the other way, i.e. importing information from the foreign disk to the array/system that was actually working fine. Also; if clear is a necessary thing to do ahead of a rebuild of that disk, or if clear means to clear the system to somehow make it ready to import the information from the foreign disk to the array/system. I imagine that making the wrong choice here might be fatal.

Please help clearing this out by telling what to choose and why.

EDIT:
I have found some more information, on dell.com support troubleshooting:

• Clear Foreign if array is working, Import Foreign if array is offline.

On the same page dell support also says:

Rule of Thumb

• If the impacted array is currently functional and data is accessible, clear the Foreign configuration.




• If the impacted array is currently non-functional, import the Foreign configuration.

In my case Windows will not start up, I’m only given the above mentioned text and when hitting a key opens up the PERC which is showing disks as online except the foreign disk.

So I need to know how this is defined. Is my array functional or non-functional, and how to tell?

@JimNim, Thank you for a detailed answer. After hitting the ESC while being in the PERC menu the server told me to use Ctrl+Alt+Del and rebooted. The message with the Ctrl+R appeared, and I tried to use that key combination. Don’t know if I got where you wanted me, since the monitor showed the previously mentioned messages and after hitting the keys I got back to the PERC menu. Assuming that this is where Ctrl+R would get me. Please help to analyze what to do and if there still is a hope for getting the server up and running or if you need some more details.

The PERC menu is showing the following:

VD Mgmt main page:

• Controller 0


• Disk Group 0


• Virtual Disks
  
  
  • Virtual Disk 0 (RED)
  
  
  • Virtual Disk 1 (RED)
  
  


• Physical Disks


• Space allocation


• Hot spares

VD Mgmt right side, while highlighting the Controller 0:

Foreign Config Present
Controller Prop.:

• DG Count: 1


• VD Count: 2


• PD Count: 6

VD Mgmt right side, while highlighting the Disk Group 0:

Foreign Config Present
Disk Group 0 Prop:

• VD Count: 2


• PD Count: 5


• Space Avl.: 0.000MB


• Free Seg.: 0


• Dedicated HS: 0

VD Mgmt right side, while highlighting the Virtual Disk 0:

Foreign Config Present
Virtual Disk 0 Prop:

• RAID Level: 5


• RAID Status: Offline


• Size: 97.656GB


• Operation: None

VD Mgmt right side, while highlighting the Virtual Disk 1:

Foreign Config Present
Virtual Disk 1 Prop:

• RAID Level: 5


• RAID Status: Offline


• Size: 4.450TB


• Operation: None

PD Mgmt main page:

• 00: SEAGATE, Online


• 01: WD, Online


• 02: WD, Online


• 03: SEAGATE, Foreign


• 04: SEAGATE, Failed


• 05: SEAGATE, Online

Welcome to serverfault! Based on the information you supplied in your "answer" post (you should edit your question to supply extra information, not post an answer):

Oh dear. Looks like you had a failed disk, and then the array ejected another disk for whatever reason - given that the outcome of whatever happened was fatal, it either ejected one of the "working" drives during the rebuild onto the hotspare (taking you to two failed/unavailable disks from the current working set), or you didn't have a hotspare and the ejection took you to two offline disks. Two disks offline in RAID5 = sudden death.

You can probably kiss the data on this array goodbye - at best you're already going to have some filesystem corruption from the data that couldn't be flushed out from cache to disk.

In any event, your raid set is "not functional" (your virtual disks are offline) so based on Dell's instructions that you have found, you should "import" the foreign array and hope it adds it back in to the existing raid5 array without too much damage resulting from the unclean state - but basically you should already be preparing to start from scratch, because there's every chance the controller will just create an additional set of offline virtual disks with all but one drive missing based on the "imported" "foreign" disk.

If by some miracle it appears to work okay you should do extensive checks that everything you need to be readable/functional on this server is in fact readable/functional - or better yet, plan to reinstall it anyway, because after this kind of failure, the integrity of the data on the volume is under serious doubt.

You make no mention of if you were aware that one of the drives was already failed, so as a follow-up action, ensure you have monitoring configured for all your RAID arrays to try and minimise risk going forward.

You have 2 failed disks. RAID5 + 2 failed disks = failed array.

You have two choices from here:

1. Send the drives to a recovery company and let them recover your data. You are likely to get almost all (if not all) of your data back. This will probably cost you $2500, plus the cost of new drives because you should not use the old drives again.


2. Continue to monkey around with your configuration, risking total loss of all of your data.

You will need to use Ctrl+R during startup to enter the configuration screen of the PERC adapter. From there, you should be able to see whether or not the drive in question is the only drive in a "foreign" state, and whether or not the RAID set is degraded/online, or failed.
If only one drive is foreign, you almost certainly do not want to import it; This could be a sure-fire method to end up with corrupted data within the RAID set, and you'd have to do some file-system repairs at minimum, if not reinstall your OS and restore data from backup.
If MULTIPLE drives are foreign and the RAID shows as failed, you should try an import. This often has a good chance of bringing the RAID back online.
Clearing is typically only a good idea if the RAID is simply degraded/online, and only a single drive shows as foreign - you'd clear in this scenario because the drive in question no longer has accurate up-to-date data for the RAID set, and a rebuild is necessary.
If importing does not resolve the issue, a last-resort option would be a "retag", recreating the RAID set from scratch without initializing.

I would highly recommend contacting support for additional help if your system is still under warranty (not sure how helpful out-of-warranty support would be with this kind of issue).