hardy->lucid upgrade using "linux-image-xen" breaks systems running in PV mode under xen/xenserver

I agree. Ubuntu is def. the Linux distribution of choice for our servers, and now that we have a XenServer setup it is a bit frustrating that the later distributions of Ubuntu seem to be the only ones that lack a simple a way to get it running paravirtualized under Xen.

I have been really looking forward to the new Lucid LTS Server, actually I deferred some server upgrades until the release of Lucid. I make heavy use of Xen to separate services on hardware in our network, and for Dapper and Hardy this has been working really great. I am very concerned to see these obstacles in the way ahead. And yes, this could lead me to the decision to look out for a new distribution to use on our servers.

Hi stiV,

Please be sure to confirm this issue exists with the latest development release of Ubuntu. ISO CD images are available from http://cdimage.ubuntu.com/releases/ . If the issue remains, please run the following command from a Terminal (Applications->Accessories->Terminal). It will automatically gather and attach updated debug information to this report.

apport-collect -p linux 517060

Also, if you could test the latest upstream kernel available that would be great. It will allow additional upstream developers to examine the issue. Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . Once you've tested the upstream kernel, please remove the 'needs-upstream-testing' tag. This can be done by clicking on the yellow pencil icon next to the tag located at the bottom of the bug description and deleting the 'needs-upstream-testing' text. Please let us know your results.

Thanks in advance.

    [This is an automated message. Apologies if it has reached you inappropriately; please just reply to this message indicating so.]

yeah well ... nice automated message :) unfortunately it has nothing to do with the underlying problem problem. The problem is not an error in the kernel package, the error is that there is no kernel package. So I urge you to read the inital report ;)

linux-image-server (64-bit) or linux-image-generic-pae (32-bit) in lucid work just fine paravirtualised in Xen domU.

Admittedly the existence of a -virtual kernel is somewhat misleading in this case, but the paravirtualised kernels do exist.

Thats true for the domU Part (using it already) - but i guess this bug is about dom0 support which is broken.

Since Oneiric, dom0 operation should be functional. That still leaves a gap in the upgrade path for dom0 running Hardy though.

Yeah. I guess in next LTS dom0 support will be there again, but the upgrade path will be hardy -> next LTS; don't know if that is supported.
Backporting the dom0 support from oneiric to lucid would be a nice alternative, opinions?

@Torsten, the upgrade path does not support to skip a LTS so Hardy to
Precise will not be officially supported. Also, backporting a kernel
"only" to ease the upgrade path isn't likely to happen considering the
effort required.

I don't have a Hardy dom0 so I cannot test but if I had to perform such
an upgrade, I'd do this (on a test machine first ;) :

1) manually install the Precise kernel (using the .deb directly)
2) reboot to see if dom0 and domUs are all working as normal
3) upgrade the host's userspace from Hardy to Lucid
4) reboot to see if dom0 and domUs are all working as normal
5) upgrade the host's userspace from Lucid to Precise
6) reboot to see if dom0 and domUs are all working as normal

On 12-01-25 03:06 AM, Torsten Krah wrote:
> Yeah. I guess in next LTS dom0 support will be there again, but the upgrade path will be hardy -> next LTS; don't know if that is supported.
> Backporting the dom0 support from oneiric to lucid would be a nice alternative, opinions?
>

This wont work.
From the Xen docs:
###################
Using older Xen versions is known to be problematic, for example Xen 4.0.0 libraries have problems with recent 2.6.32.x kernels, making xend fail to start due to evtchn/gntdev device node creation issues. Using Xen 3.4.2 or older won't work at all, since old hypervisor versions lack the new required IOAPIC setup hypercall and boot will fail with IRQ related issues.

It's recommended to run the latest Xen 4.0.x version, at least Xen 4.0.1.
###################

So step 1 would fail, because my hardy does "only" got a xen 3.3.0 or xen 3.2.x one; but those one would already fail to boot on this new kernel.
So i'll have to update to xen 4.0.1+ first, but there is no such paket for hardy *uggggh* - so still a "broken" upgrade path.

What now? Do you really want to tell me that i am stuck at hardy and have to do a complete reinstall of my host to get to a supported LTS release with Xen dom0 support?

Lucid is no more a supported version now