Issues with Swap file on importing OVF Template

Hi all,
I’m hoping you’ll be able to help, I have an OpenSuse virtual machine version 11.4 which I have running happily in VMWare workstation 9 and I export the server to an OVF template to import into ESXi 4.1.

It all imports fine and looks good but when I boot the OS it comes up saying:
Activating swap-devices in /etc/fstab failed

If I import the template back into VMWare workstation it works fine, have any of you got any ideas why this might be happening?

None of the disk sizes etc are changed during the transition and I’ve tried thick and thin provisioning.

Many thanks.

Don’t worry I’ve fixed this but wanted to share this with the forum.

It was all because the date on the server is wrong and the superblock was dated in the future.

trevrobwhite wrote:
> Don’t worry I’ve fixed this but wanted to share this with the forum.
>
> It was all because the date on the server is wrong and the superblock
> was dated in the future.

Well spotted!

Yes, as soon as you have more than one computer, keeping the time
synchronised is important. NTP is usually the best way.

On 2013-04-26 13:18, Dave Howorth wrote:
> Yes, as soon as you have more than one computer, keeping the time
> synchronised is important. NTP is usually the best way.

Yep.

But noticing that this is a virtual machine, it is the host that needs
be synced via NTP. All the guests must use, instead, guest tools and
equivalents, that sync the guest clock to the host clock, with much
lower resource usage, and it is also more reliable (you can not really
adjust the guest clock speed, no real hardware).

There is a vmware document about this.


Cheers / Saludos,

Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)

Completely agree, this is that will happen with the ESXi host being hooked up to the time server but it was the first time installation of ESXi and the first OVF failed to boot which stopped the engineer in his tracks as it looked really serious when actually it was something very minor.

It is very dramatic the way it doesn’t boot over just a time issue so I thought I’d reply it back in case someone else gets the same issue. :slight_smile:

On 2013-04-26 15:46, trevrobwhite wrote:

> Completely agree, this is that will happen with the ESXi host being
> hooked up to the time server but it was the first time installation of
> ESXi and the first OVF failed to boot which stopped the engineer in his
> tracks as it looked really serious when actually it was something very
> minor.

Possibly, if you were booting in text mode without the flash display,
you would have seen a text message in passing that explained the issue.
Provided you were actually reading the messages and they did not
disappear too fast…

> It is very dramatic the way it doesn’t boot over just a time issue so I
> thought I’d reply it back in case someone else gets the same issue. :slight_smile:

Yes, it is not something I ever have seen with a swap partition. Other
types, yes.


Cheers / Saludos,

Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)

On 2013-04-26 15:28, Carlos E. R. wrote:

> Yep.
>
> But noticing that this is a virtual machine, it is the host that needs
> be synced via NTP. All the guests must use, instead, guest tools and
> equivalents, that sync the guest clock to the host clock, with much
> lower resource usage, and it is also more reliable (you can not really
> adjust the guest clock speed, no real hardware).
>
> There is a vmware document about this.

In the file virtualmachinename.vmx, change this:


tools.syncTime = "TRUE"

By default it is false. I don’t have a link ready that documents this.


Cheers / Saludos,

Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)