HELP! "Two Filesystems with the same UUID have been detected" Occurs after turning on VT

[MichaelCorm]

I have been through everything i know about VM (not a great amount) but i have been trying to get this server to run for the last 4 days.

The server is a Dell R710

Newest Bios Update

ESXi4.1 installed

Everything works fine until I try to turn on VT to install my 64-bit Machines. Then this message, "Two Filesystems with the same UUID have been detected. Makes sure you do not have two ESXi installations."

I am willing to try anything!

I have tried removing the additional NIC also to no avail.

[tsce]

I have exactly the same problem after upgrading my Dell R710 from ESXi 4.0 to 4.1. I also upgraded a bunch of other servers running on other hardware (Dell R510, Sun X2100 and even a Dell Precision T7400 test machine) without problems.

I wonder whether this may be a raid issue. Do you have a raid controller in your R710?

Nico

[tsce]

Hmm, my R710 server has 2 disk systems: 1 single disk which I use to run VMware ESXi and 1 3-disk raid which I actually don't use as my vmware images are stored on the SAN. ESXi is notorious for mixing up disks during the installation so I probably installed ESXi on the raid first and then reinstalled using the single disk. I never used the raid so I never reformatted it. I guess ESXi 4.1 checks the disks more thorough and discovers the second installation.

I'm going to try wiping the raid completely (I would remove it but I don't have any empty disk trays.)

Nico

[tsce]

Yep, that did the trick for me:

I rebooted the server and came back still running ESXi 4.0 as this was an upgrade. I added a new datastore and discovered the raid indeed had VMware partitions on it so told it to destroy the partitions and create new ones. Tried upgrading to 4.1 again and now it comes up properly.

I you are not doing an upgrade then I suggest physically removing the second disk and wiping it in another server or boot a linux live cd and wipe the second disk.

Nico

[vinayak111]

Re-insert ESXi install disk and this time dont install but repair the ESX4i installation using the option that appears on screen...worked out for me.

[tekenconsulting]

Just to add some fresh informations..

I just had same issue on IBM HS22 blades.

Upgraded to 4.1 booted - came up with the initial error. The server then came up with 4.0 after reboot.

The solution for me was to shut down server, remove one of the two mirrored disks, start the server.

Insert the disk again when the server was up.

Now I am awaiting a complete sync, and after that I will try again.

This indicates that it is RAID related since I could not boot on mirrored disks, but it worked fine on broken mirror.

[tekenconsulting]

Server came up as 4.0 after reboot so I ended up reinstalling to 4.1 from CD, and after that no issues.

I would have liked to find what caused this error, but again it is easier and faster to reinstall.. guess that saves VMware support lots of hours..

[vinayak111]

This solution will help all :-

Make a copy of Hiren Boot CD latest one.

Boot with it and choose Mini Linux to boot up.

Once that gets booted right click on desktop and you will be able to find lot of partition tools .

I used the Gparted Tool.

I found ESX4i was installed multiple time on my hd and original installation had not been cleaned up.

So I just cleared all partitions on my local disk till entrie disk was shown unallocated.

Then started ESX4.1i installation and boom it worked.

Consider awarding points if you feel this solution is helpful.

Thanks

Vinayak

[Phylum]

I have also tried to update from ESX4.0 build-294855 embedded (via the USB port on the board) to 4.1 on a Dell T710 with only 3 drives in a non-raid configuration. Thinking it was as simple as running the host utility, I applied all the patches, saw it was still running 4.0 and after a bit of research discovered that the Host Update Utility was no longer an option. (See this page for more information.)

Following the instructions here, I downloaded the ESXi 4.1 upgrade ZIP from ESXi 4.0 (upgrade-from-ESXi4.0-to-4.1.0-0.0.260247-release.zip) from somewhere, then proceeded with the upgrade via CLI. As far as I could tell, everything seemed to go well and the query reported 4.1. Good show - or so I thought. I bounced the box received the PSOD with the error we're all discussing: "Two Filesystems with the same UUID have been detected"

There must be a way, right, that doesn't involve removing hardware or standing on one leg on he ceiling while balancing fine china on my nose. I'm not thrilled about the potential 'remove my hard drives' method of fixing this. I'm willing to try suggestions as long as it doesn't require pulling drives.

For now I'll stick with 4.0.

[J1mbo]

I came across this issue recently with an initial install. Server had 3x 2TB LUNs (each mirrors) on a Perc 6i. The installer would very definitely write the ESXi image to the wrong (and a random) LUN regardless of that selected, hence two times out of three the server could not be booted and when it eventually did, it showed this message.

Similar to the above, I had to run the installer and updater with only one LUN defined, and then add the others once everything had been installed and configured.

http://blog.peacon.co.uk

Please award points to any useful answer.

Unofficial List of USB Passthrough Working Devices

[virtusjester]

You probably got your problem fixed but maybe this can help somebody else out who is getting that error.

I encountered this problem with ESXi 4.1 on a Dell PowerEdge R710 server and it appears it was the USB pass through that was causing it.  I had selected both the EHCI controllers for pass through and it seems that 00:1a.7 is for the internal USB which my thumb drive is connected to.  Selecting that one for pass through must have inserted a new UUID (not sure how that works).  I used the install disk to repair the installation and then only enabled pass through on one of the EHCI controllers (00:1d.7) and have not seen that error again.

[marrosen]

Thanks, Vinayak, your solution worked for me.  I downloaded the utility CD you suggested from http://www.hirensbootcd.org/download/ and used the linux boot to run Gparted.  As you said, I deleted all the partitions and I was able to re-install ESXi 4.1 without the PSOD repeating.

[BigGeorge]

Hi virtusjester, how you've repaired esxi installation? what procedure you've followed? I've a Dell Poweredge R510 with ESXi 5.0 and I've the same problem after that I've enabled the USB passthrough on 00:1a.7  tank you  BigGeorge

[DSTAVERT]

Welcome to the Communities

The problem occurs when you have two installs of ESXi. If ESXi is installed on a USB stick and also on the hard drive the boot process will see the two installs and stop with the Duplicate UUID message. You can use the following method to temporarily boot http://kb.vmware.com/kb/1035107

Since this happened after USB passthrough try disconnecting the USB disk before rebooting.

[Phylum]

I posted about the same problem here and ultimately fixed it

http://communities.vmware.com/message/1765808

The duplicate UUID issue was more of an annoyance.  While I'm pleased to have this issue resolved, the most important issue remains to be addressed:

When installing ESXi onto and booting from an internal USB drive (as in inside the computer via the USB port baked into the board), how do you avoid running into this problem when upgrading?

[BigGeorge]

Hi DSTAVERT,In my case the cause is the activation of USB Passthrough, no doble ESXi installation.I've Made a New fresh installation and activate the USB Passthrough on the other device ID and it work properly!I think thath we have two different problem:1. Duplicate UID due to two ESXi installation2. duplicate UID due to USB Passthrough activation (may be only on Dell Hardware?) BigGeorge

[zuraz]

Guys,

I've had the same issue except in my case it wasnt during an ESXi upgrade but while performing a fresh install of ESXi 4.1 on a Dell T310 with disks in RAID1. I did not want to go about dissecting my server (unplugging disks) so after searching far and wide and trying different methods (including the Hiren Boot CD), a mish-mash of various suggestions/fixes did it for me. And here are the steps :

To resolve the "Two Filesystems with the same UUID have been detected" issue:

Restart the system.

During the reboot, when the hypervisor screen appears, press Shift+O.

In the boot option, type overrideDuplicateImageDetection and press Enter.

Once it ESXi successfully boots up -

1. Press F2 to set it up with a management IP & root password

2. Use vSphere Client to login to the ESXi host

3. Once logged in, goto Configuration > Storage > Datastore > Add Storage (to add the disk thats missing from the list, thats got the 'Duplicate Image')

4. While adding, make sure you fully format this disk **MOST IMPORTANT STEP**

5. Now Reboot the ESXi host to test if problem's been fixed (it most likely will be)

WOOHOO if it worked for you or just go boohoo and all the best !