Bare Metal Recovery Fails To Use Correct Driver For 2016

[AlexLeadingEdge]

We have made a backup of a server, something wasn't quite right and it would freeze intermittently, which we suspect is software related, so we want to move the OS and data to a new machine. We made a backup of all the drives and then tried to do a Bare Metal Recovery to a new system, with the same RAID configuration, but Veeam Agent Beta won't use the correct Windows Server 2016 RAID drivers, instead it wants to use the Windows 7 RAID drivers. If we try to use anything other than the Windows 7 RAID drivers it says "Failed to install driver. Error code: ERROR_NO_DEVICE_ID"

When we boot up the recovered Server 2016 on the new machine it doesn't like the default Windows RAID driver and we get a BSOD with INACCESSIBLE_BOOT_DEVICE. At least that is what I think is happening.

I have tried both injecting the drivers and tried not injecting the drivers, both give the same BSOD.

How do I insert the correct driver into the new Windows Server 2016 install and get it to start using that one instead?

[AlexLeadingEdge]

Anyone???

[Dima P.]

Alex,

Please share the recovery logs (you can find it in backup file location) with me via PM.

P.S. On which operating system Recovery Media was created?

[AlexLeadingEdge]

Recovery Media was created on a Windows Server 2016 machine. I will PM you the file location.

[Dima P.]

Thanks! The original OS in backup is Windows 7, is that correct? Or thats the only driver that works in BMR?

[AlexLeadingEdge]

Thanks! The original OS in backup is Windows 7, is that correct? Or thats the only driver that works in BMR?

Original OS was Server 2016. The only RAID driver that seems to work with Veeam Agent is the Windows 7 one, which doesn't work for Windows Server 2016, and I couldn't find a way to inject the Windows Server 2016 RAID driver after the recovery ISO had already been created.

As this company doesn't use a paid-for Veeam product I didn't feel it was appropriate to call Veeam for help.

We tried using another backup solution (StorageCraft's Shadow Protect) to do the same thing, but we get the same BSOD "INACCESSIBLE_BOOT_DEVICE". Not sure what is causing it. According to our records this company had an old but still valid licence with StorageCraft, so I called them for technical support. After sending off heaps of logs they say the boot partition of the old server may be partially corrupted. We tried several recovery options but all failed with the same BSOD that Veeam was showing. We tried installing Server 2016 with the hope of doing a restore of only the OS partition, keeping the installed boot partitions, but instead of a GPT install we got an MBR install; Microsoft says probably because of the format of the USB drive that the ISO was installed to, which wasn't the original USB stick that was used to install the old server.

After having wasted two days on this with no progress my boss pulled the plug on this. I did a clean install of Server 2016 on the new hardware and migrated all the emails and data over. Not an ideal solution but I had no options as the original server was randomly freezing, with no pattern other than it could be anywhere between 1 hour and 5 days until the next freeze.

Given that this was our first Bare Metal Restore of a server-class machine with RAID drives it wasn't a good result

[Dima P.]

Alex,

Thanks for sharing the update and sorry to hear that you have to build this server from scratch.

As this company doesn't use a paid-for Veeam product I didn't feel it was appropriate to call Veeam for help.

You are always welcome to open a support case regarding VEB/VAW. We do provide best effort support for all free products.

I did a clean install of Server 2016 on the new hardware and migrated all the emails and data over.

It should be possible to perform a file level or volume level restore from existing backup file to new OS.

Given that this was our first Bare Metal Restore of a server-class machine with RAID drives it wasn't a good result

Any chance you still have the recovery logs left in your repository? Can you share them with me via PM?

[AlexLeadingEdge]

Any chance you still have the recovery logs left in your repository? Can you share them with me via PM?

I sent them to you by PM the other day

[Dima P.]

My bad, sorry. Passed to QA folks!

[AlexLeadingEdge]

My bad, sorry. Passed to QA folks!

No problem, I administer 500 computers, I know how easy it is for things to fall through the cracks

[ejenner]

In case anybody is looking for an answer here... I had this problem and managed to work out what was going on out of necessity.

My server in this case was a HP blade server with no KVM. Had to do all through iLo.

I had replaced the two hard drives as I wanted to increase the available disk space on the server once recovered.

It was replacing the drives which caused the problem. Even though I didn't expect second-hand hard drives (from one of our decommissioned systems) to contain any data... especially since they weren't in the original server and were now part of a new array completely unrelated to the previous configuration they had retained the OS from the server they had previously been removed from.

To save messing about in this case I took the drives out and used another pair of different hard drives... trying my luck. The second set of hard drives didn't remember anything about where they had come from and I was able to boot to the Veeam recovery ISO. I created the new array in the same way I had done before but just had a better set of disks.