Joeyg2391201110
Contributor
Contributor

Transfer VM to another host with the same exact hardware setup issues

Hello,

Im running free esxi 5.1.0 and a windows 2008 r2 x64 vm which im trying to transfer to another host with the exact same hardware/esxi setup.  Ive used several methods of trying to transfer the vm but all have failed.  I suspect it has something to do with the vm config file corruption from what im reading.  I have used vm converter and it fails at 1% with no specific error, ovatools fails with

SHA1 digest of file **.vmdk does not match manifest.  I even used veambackup fscp free and it fails as well.  This is a very important exchange server 2010 vm and hardware is fine. The VM is also running fine when turned on.  I believe the cause of my issues transferring is because the VM is thin based..  I also think this was a clone i did a long time ago.  Again, I cant afford to lose this vm and really appreciate the help.

Thanks

0 Kudos
14 Replies
Joeyg2391201110
Contributor
Contributor

i think i figured it out.  For some reason i had my vm set up as 2 sockets which the host doesnt even have but the vm still ran fine.  I shut it down and change it to single socket single core but i have not tested the transfer yet.  Can someone confirm if they have seen this before?

Thanks

0 Kudos
Joeyg2391201110
Contributor
Contributor

negative.  Still fails at moving the vm

0 Kudos
Madmax01
Expert
Expert

Hello,

hm strange.  i know that only for OVF Appliances.  Could be Kind of Corruption.

So i would now avoid shutting down the VM.  i would try out to use the Converter to make a V2V  of it to create a new xxx_new. (to have that VM Covered)

maybe you could arrange that in a Time where not much activity is happening ;(.

Best would be to open  a Vmware Prio 1 Call to get direct help. Because it's an Incident ;(

Best regards

Max

0 Kudos
Joeyg2391201110
Contributor
Contributor

i have tried converter and it fails at 1% with this error

Capture.JPG

using ovftool command i get this

Capture2.JPG

I also tried browsing the datastore and doing a normal copy and that fails.

Im running out of ideas.  I even reboot the host and still no good.  The VM is running perfectly fine it's just that i want to move it to a raid 1 group.  I really need assistance if possible.

Thanks

0 Kudos
Madmax01
Expert
Expert

Hi theire,

ah so it's an OVF?

So then i found this  virtuallyGhetto: How to Deploy an OVF Located On ESXi Datastore Using ovftool

theire is an entry from "Jerome OrtegaJuly 28, 2013 at 7:19 PM"

he noted having same Problem > and he used "--skipManifestCheck switch"  to have a Workaround.

maybe it helps in future > once downloading OVF or any other Binaries > checking the Hash md5/sh1  against the Vendors ones Smiley Wink

Best regards

Max

0 Kudos
Joeyg2391201110
Contributor
Contributor

it's  not a OVF file and is currently a VM.  I used ovftool.exe successfully before to move the vm from one host to another successfully.  How can i check exactly what wrong with the VM config?  I like to fix it beforehand instead of using skipManifestCheck switch.

Thanks

0 Kudos
Joeyg2391201110
Contributor
Contributor

when i browse the datastore and try to copy i get this error.  Again, this VM is running perfectly fine and i shut down everything before i do the move.

Thanks Capture4.PNG

0 Kudos
Madmax01
Expert
Expert

Hm very strange. never saw such conditions.  Do you have San connection in background? Maybe theire is a Kind of Block Level Corruption,.... ?.

Anyway for this Case i think it would be important to contact Vmware for a Direct Support.

i don't sense it as a "normal" Problem.

Best regards

Max

0 Kudos
Joeyg2391201110
Contributor
Contributor

The email vm is on a sata drive.  I would contact VMWare but they are charging me 300 bucks which is something i dont have.  ill see what i can do.  Thanks

0 Kudos
Joeyg2391201110
Contributor
Contributor

Turns out the disk has a lot of bad sectors so it's only a matter of time before this vm dies.  I ran a disk util program.  I found it weird because esxi 5 event viewer doesnt show anything regarding HD errors

0 Kudos
EV_Simon
Enthusiast
Enthusiast

Surely that will depend on whether these errors are reporting via SMART, if not then I wouldn't expect the events to show up.

One word of advice would be to get this backed up anyway you can because if\when it fails you will be in a world of hurt.

0 Kudos
Joeyg2391201110
Contributor
Contributor

backing up is what im trying to do but it's impossible to copy the vm using any method from above

0 Kudos
EV_Simon
Enthusiast
Enthusiast

Ok you may have to see about taking a backup of the data inside the VM and build up a new VM on the other host and restoring the data that way.

The problem you may face is that the more attempts you make to take data off the disk 'may' cause the disk to fail sooner rather than later.

This is why having a decent backup solution in place can be the difference between a successful company or one that fails because they lose their data and have to stop trading.

I don't know how many critical vm's you have running on that disk but you may want to shut down the non critical ones and see about getting that in VM backup kicked off ASAP.

0 Kudos
Joeyg2391201110
Contributor
Contributor

yes im aware of possible failure trying to do activity on the disk.  It's only 1 exchange cas server and dag member.  I can survive but would rather copy the vm.  Im going to deploy a cas array to further protect myself.

Thanks

0 Kudos