RAID Drive Question
December 31st, 2025, 12:14
Hello I have a client with a very old Dell PowerEdge T310 running a PERC6/i controller. The system is running RAID5 (4 hard 146GB drives) and has VMWare installed. I think the controller MIGHT be bad, and I am pretty sure that one drive has gone bad.
The client contacted us saying that they were unable to access their shared drive (from a VM) and upon trying to access the Windows OS remotely, the system showed that it was online, however it was non-responsive when I remoted into it. So the OS was still somewhat functional because it did accept the connection from our ScreenConnect software.
When I dispatched a technician, the standard VMWare ESXi screen was up.
Upon restarting the physical server, we were presented with the "Foreign Configuration Found" message, and rather than load the foreign config, we went into the controller's configuration to see what it was showing. The 2nd drive (drive number 1) showed as failed. There was no utility to run a drive test. Upon rebooting the server and going back into the config utility, it showed that 2 drives were failed (1 & 2), and that the only drives not in an error state were Drives 0 & 3....Rebooting the server again, it then showed that all drives were failed with the exception of drive 0...and it shows as foreign.
Subsequent reboots have produced different information on the Foreign View tab (first it showed the physical drives along with the logical drive, and now it shows only one physical disk and Virtual Disk 255)...very inconsistent behavior, which is why I am thinking the actual controller might be bad, opposed to the all 3 disks being bad.
I am not sure if there is a utility that I can possibly use to verify if the disks are good or not first before making a decision to purchase recovery software or send off to a recovery center. Maybe taking each drive out (will probably have to remove from the drive tray as well...they are SAS drives) and testing on a separate machine.
If there is a utility to do that, and we find that only the single drive is bad opposed to all 3, do we first replace the controller, reinstall the drives and attempt to import the foreign config if it's found on the drives, or is there another route we should go. I don't want to do anything to increase the risk of not being able to recover the data.
This client doesn't have a backup of this data as they were under the impression it had been moved to a different location.
I have attached a couple of photos showing what the controller is showing now.
Please advise.
The client contacted us saying that they were unable to access their shared drive (from a VM) and upon trying to access the Windows OS remotely, the system showed that it was online, however it was non-responsive when I remoted into it. So the OS was still somewhat functional because it did accept the connection from our ScreenConnect software.
When I dispatched a technician, the standard VMWare ESXi screen was up.
Upon restarting the physical server, we were presented with the "Foreign Configuration Found" message, and rather than load the foreign config, we went into the controller's configuration to see what it was showing. The 2nd drive (drive number 1) showed as failed. There was no utility to run a drive test. Upon rebooting the server and going back into the config utility, it showed that 2 drives were failed (1 & 2), and that the only drives not in an error state were Drives 0 & 3....Rebooting the server again, it then showed that all drives were failed with the exception of drive 0...and it shows as foreign.
Subsequent reboots have produced different information on the Foreign View tab (first it showed the physical drives along with the logical drive, and now it shows only one physical disk and Virtual Disk 255)...very inconsistent behavior, which is why I am thinking the actual controller might be bad, opposed to the all 3 disks being bad.
I am not sure if there is a utility that I can possibly use to verify if the disks are good or not first before making a decision to purchase recovery software or send off to a recovery center. Maybe taking each drive out (will probably have to remove from the drive tray as well...they are SAS drives) and testing on a separate machine.
If there is a utility to do that, and we find that only the single drive is bad opposed to all 3, do we first replace the controller, reinstall the drives and attempt to import the foreign config if it's found on the drives, or is there another route we should go. I don't want to do anything to increase the risk of not being able to recover the data.
This client doesn't have a backup of this data as they were under the impression it had been moved to a different location.
I have attached a couple of photos showing what the controller is showing now.
Please advise.
- Attachments
-
-
Re: RAID Drive Question
July 15th, 2026, 5:41
I had.. well still have as Disks are not quick to source in South Australia.. a similar issue on a Lenovo server. The Lenovo servers have a very comprehensive logging system and I was able to download the logs of the RAID. In my case it was RAID10, 6x 4TB disks. Of course the 0 and 1 failed of the same stripe. This is the first time Ive dealt with a RAID failure of this nature and it took a fair bit of reading before I did anything, so I understand your hesitation. This confirmed it was the drives. Damn Seagate!
I have read that poor connections can sometimes lead to the foreign config message, but to be honest poor connection/dirty pins have not really ever been an issue anytime over the years. but worth cleaning them to rule it out as a simple first step.
I am in no way recommending anything here as I am not any sort of expert in RAID and SAS drives. If I was in your situation, what I might look at doing is buying a SAS controller card for a different PC and attempt to image each drive. At least then I guess there is RAID recovery software that could work on the images. Then swap out the controller, lastly look at the disks or hire someone to recover.
In my case I was able to use the Lenovo Xclarity utility to put one of the drives in the stripe back online. I bought a 16TB external and started copying data most important first. When that drive reported failed, I stopped and sent the 6 disks to a DR lab. I have 6x 10TB drives coming and have to rebuild 5 VM servers after that.
I would say that likely a disk started going bad, the added strain on the raid could have led to other disks also going bad or starting to. It is easy not to notice unless you have some kind of reporting (unlikely on a server that old) and someone watching it
I have read that poor connections can sometimes lead to the foreign config message, but to be honest poor connection/dirty pins have not really ever been an issue anytime over the years. but worth cleaning them to rule it out as a simple first step.
I am in no way recommending anything here as I am not any sort of expert in RAID and SAS drives. If I was in your situation, what I might look at doing is buying a SAS controller card for a different PC and attempt to image each drive. At least then I guess there is RAID recovery software that could work on the images. Then swap out the controller, lastly look at the disks or hire someone to recover.
In my case I was able to use the Lenovo Xclarity utility to put one of the drives in the stripe back online. I bought a 16TB external and started copying data most important first. When that drive reported failed, I stopped and sent the 6 disks to a DR lab. I have 6x 10TB drives coming and have to rebuild 5 VM servers after that.
I would say that likely a disk started going bad, the added strain on the raid could have led to other disks also going bad or starting to. It is easy not to notice unless you have some kind of reporting (unlikely on a server that old) and someone watching it
Powered by phpBB © phpBB Group.