Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
netboot images for Debian 9 seems is broken?
1,306 views
Skip to first unread message
mizuki
Jun 14, 2018, 3:50:03 PM6/14/18
to
Hi,
The netboot image found in http://ftp.debian.org/debian/dists/Debian9.4/main/installer-amd64/current/images/netboot/ for Debian 9 seems broken, we use that for unattended installation.
The error msg returned says:
"No kernel modules were found. This probably is due to a mismatch between the kernel used by this version of the installer and the kernel version in the archive.
If you're installing from a mirror, you can work around this problem by choosing to install a different version of Ubuntu. The install will probably fail to work if you continue without kernel modules"
Can someone update the netboot image if it's a bug or any workaround helps to fix this situation would be greatly appreciated.
Thanks.
Mizuki
Andy Smith
Jun 14, 2018, 9:50:03 PM6/14/18
to
Hi,
On Thu, Jun 14, 2018 at 03:29:28PM -0400, mizuki wrote:
> "No kernel modules were found. This probably is due to a mismatch between
> the kernel used by this version of the installer and the kernel version in
> the archive.
In my experience, the above is correct. That is, I've experienced
this sort of thing 3 or 4 times and every time that has been the
reason for it.
> If you're installing from a mirror, you can work around this problem by
> choosing to install a different version of Ubuntu. The install will
> probably fail to work if you continue without kernel modules"
It's a bit odd that it says Ubuntu when you used a Debian netboot.
Use daily netboot image, pick a different mirror?
Cheers,
Andy
--
https://bitfolk.com/ -- No-nonsense VPS hosting
On Thu, Jun 14, 2018 at 03:29:28PM -0400, mizuki wrote:
> "No kernel modules were found. This probably is due to a mismatch between
> the kernel used by this version of the installer and the kernel version in
> the archive.
this sort of thing 3 or 4 times and every time that has been the
reason for it.
> If you're installing from a mirror, you can work around this problem by
> choosing to install a different version of Ubuntu. The install will
> probably fail to work if you continue without kernel modules"
Use daily netboot image, pick a different mirror?
Cheers,
Andy
--
https://bitfolk.com/ -- No-nonsense VPS hosting
Mike
Jun 18, 2018, 9:30:05 AM6/18/18
to
On Fri, Jun 15, 2018 at 01:43:50AM +0000, Andy Smith wrote:
> Hi,
>
> On Thu, Jun 14, 2018 at 03:29:28PM -0400, mizuki wrote:
> > "No kernel modules were found. This probably is due to a mismatch between
> > the kernel used by this version of the installer and the kernel version in
> > the archive.
>
> In my experience, the above is correct. That is, I've experienced
> this sort of thing 3 or 4 times and every time that has been the
> reason for it.
Forgive me if this is a stupid question but what exactly does that mean
> Hi,
>
> On Thu, Jun 14, 2018 at 03:29:28PM -0400, mizuki wrote:
> > "No kernel modules were found. This probably is due to a mismatch between
> > the kernel used by this version of the installer and the kernel version in
> > the archive.
>
> In my experience, the above is correct. That is, I've experienced
> this sort of thing 3 or 4 times and every time that has been the
> reason for it.
ie what is "the archive"? Presumably that means a mismatch between the
linux kernel and the kernel modules in the initrd or am I very much
mistaken?
> > If you're installing from a mirror, you can work around this problem by
> > choosing to install a different version of Ubuntu. The install will
> > probably fail to work if you continue without kernel modules"
>
> It's a bit odd that it says Ubuntu when you used a Debian netboot.
described when I added Debian 9 to my PXE server. I think I managed to
find a different kernel and initrd so I tried that too which got a
little further but as far as I can tell, didn't have the ext drivers, so
couldn't mkfs when it got the file system creation part of the install
process. I have to confess, I didn't get any further, as a work-around,
I found something else to occupy my time.
>
> Use daily netboot image, pick a different mirror?
try to remember to have a got at fixing this when I get home, as it's
rekindled my interest in getting it working.
Are the PXE installers known to be a little flaky? I've found them a
little hit and miss in the past but never put enough effort into working
out if it was me or someone else that was at fault :-)
Mike.
Andy Smith
Jun 18, 2018, 3:30:04 PM6/18/18
to
Hello,
On Mon, Jun 18, 2018 at 02:26:20PM +0100, Mike wrote:
> On Fri, Jun 15, 2018 at 01:43:50AM +0000, Andy Smith wrote:
> > On Thu, Jun 14, 2018 at 03:29:28PM -0400, mizuki wrote:
> > > "No kernel modules were found. This probably is due to a mismatch between
> > > the kernel used by this version of the installer and the kernel version in
> > > the archive.
> >
> > In my experience, the above is correct. That is, I've experienced
> > this sort of thing 3 or 4 times and every time that has been the
> > reason for it.
>
> Forgive me if this is a stupid question but what exactly does that mean
> ie what is "the archive"? Presumably that means a mismatch between the
> linux kernel and the kernel modules in the initrd or am I very much
> mistaken?
Yes. When you PXE boot, the kernel and initramfs comes from the
netboot image that you downloaded in the past, but the archive will
be different after a point release which can result in the
installer's failure to load certain essential modules like
filesystems.
So in my experience every time there is a point release you need to
re-download the netboot images that you are serving over PXE as
well.
> Are the PXE installers known to be a little flaky?
I've got along with them okay as long as I remember to download a
new one after every point release.
I'm not 100% sure this is OP's problem (the Ubuntu reference is
alarming), but if I was OP I'd certainly try doing that first.
On Mon, Jun 18, 2018 at 02:26:20PM +0100, Mike wrote:
> On Fri, Jun 15, 2018 at 01:43:50AM +0000, Andy Smith wrote:
> > On Thu, Jun 14, 2018 at 03:29:28PM -0400, mizuki wrote:
> > > "No kernel modules were found. This probably is due to a mismatch between
> > > the kernel used by this version of the installer and the kernel version in
> > > the archive.
> >
> > In my experience, the above is correct. That is, I've experienced
> > this sort of thing 3 or 4 times and every time that has been the
> > reason for it.
>
> Forgive me if this is a stupid question but what exactly does that mean
> ie what is "the archive"? Presumably that means a mismatch between the
> linux kernel and the kernel modules in the initrd or am I very much
> mistaken?
netboot image that you downloaded in the past, but the archive will
be different after a point release which can result in the
installer's failure to load certain essential modules like
filesystems.
So in my experience every time there is a point release you need to
re-download the netboot images that you are serving over PXE as
well.
> Are the PXE installers known to be a little flaky?
new one after every point release.
I'm not 100% sure this is OP's problem (the Ubuntu reference is
alarming), but if I was OP I'd certainly try doing that first.
Mike
Jun 19, 2018, 6:10:04 AM6/19/18
to
On Mon, Jun 18, 2018 at 07:24:25PM +0000, Andy Smith wrote:
> Hello,
Previously I've updated the kernel and initrd each time there's a full
release. That might well explain why I've found it a bit unreliable in
the past.
Last night I downloaded the kernel and initrd that the OP was
complaining about (from the link they provided) and ironically it fixed
my issue :-) According to my preseed file, I'm using
ftp.uk.debian.org/debian. I guess the OP must be pointing at the wrong
archive.
Thanks again for the tip, I'll resolve to update with each Point Release
in future!
Mike.
> Hello,
>
> Yes. When you PXE boot, the kernel and initramfs comes from the
> netboot image that you downloaded in the past, but the archive will
> be different after a point release which can result in the
> installer's failure to load certain essential modules like
> filesystems.
>
> So in my experience every time there is a point release you need to
> re-download the netboot images that you are serving over PXE as
> well.
>
> > Are the PXE installers known to be a little flaky?
>
> I've got along with them okay as long as I remember to download a
> new one after every point release.
>
> I'm not 100% sure this is OP's problem (the Ubuntu reference is
> alarming), but if I was OP I'd certainly try doing that first.
>
Thanks for the tip about the point releases. I didn't know that.
> Yes. When you PXE boot, the kernel and initramfs comes from the
> netboot image that you downloaded in the past, but the archive will
> be different after a point release which can result in the
> installer's failure to load certain essential modules like
> filesystems.
>
> So in my experience every time there is a point release you need to
> re-download the netboot images that you are serving over PXE as
> well.
>
> > Are the PXE installers known to be a little flaky?
>
> I've got along with them okay as long as I remember to download a
> new one after every point release.
>
> I'm not 100% sure this is OP's problem (the Ubuntu reference is
> alarming), but if I was OP I'd certainly try doing that first.
>
Previously I've updated the kernel and initrd each time there's a full
release. That might well explain why I've found it a bit unreliable in
the past.
Last night I downloaded the kernel and initrd that the OP was
complaining about (from the link they provided) and ironically it fixed
my issue :-) According to my preseed file, I'm using
ftp.uk.debian.org/debian. I guess the OP must be pointing at the wrong
archive.
Thanks again for the tip, I'll resolve to update with each Point Release
in future!
Mike.
David Wright
Jun 20, 2018, 5:50:04 PM6/20/18
to
On Wed 20 Jun 2018 at 11:31:20 (-0400), mizuki wrote:
> Hi,
>
> Apologies for my lousy copy and paste, it is Debian (not Ubuntu), see
> attached screenshot.
> When this error, drop to a shell and run 'uname -r' returned *4.9.0-4.amd64*
> ,
> I believe the kernel in Archive Mirror is higher 4.9.0-6. That's why this
> is likely a bug to me.
What is the file modification timestamp on the image you downloaded?
Cheers,
David.
> Hi,
>
> Apologies for my lousy copy and paste, it is Debian (not Ubuntu), see
> attached screenshot.
> When this error, drop to a shell and run 'uname -r' returned *4.9.0-4.amd64*
> ,
> I believe the kernel in Archive Mirror is higher 4.9.0-6. That's why this
> is likely a bug to me.
What is the file modification timestamp on the image you downloaded?
Cheers,
David.
mizuki
Jun 20, 2018, 7:10:03 PM6/20/18
to
The image dated at Mar 4th 2018 which is the latest avaialble at upstream ftp.debian.org (file sizes, time stamps all matched with the upstream).
Thanks
Mizuki
David Wright
Jun 20, 2018, 9:30:03 PM6/20/18
to
On Wed 20 Jun 2018 at 19:03:33 (-0400), mizuki wrote:
> On Wed, Jun 20, 2018, 17:43 David Wright <deb...@lionunicorn.co.uk> wrote:
> > On Wed 20 Jun 2018 at 11:31:20 (-0400), mizuki wrote:
> > > Hi,
> > >
> > > Apologies for my lousy copy and paste, it is Debian (not Ubuntu), see
> > > attached screenshot.
> > > When this error, drop to a shell and run 'uname -r' returned
> > *4.9.0-4.amd64*
> > > ,
> > > I believe the kernel in Archive Mirror is higher 4.9.0-6. That's why this
> > > is likely a bug to me.
> >
> > What is the file modification timestamp on the image you downloaded?
>
> On Wed, Jun 20, 2018, 17:43 David Wright <deb...@lionunicorn.co.uk> wrote:
> > On Wed 20 Jun 2018 at 11:31:20 (-0400), mizuki wrote:
> > > Hi,
> > >
> > > Apologies for my lousy copy and paste, it is Debian (not Ubuntu), see
> > > attached screenshot.
> > > When this error, drop to a shell and run 'uname -r' returned
> > *4.9.0-4.amd64*
> > > ,
> > > I believe the kernel in Archive Mirror is higher 4.9.0-6. That's why this
> > > is likely a bug to me.
> >
> > What is the file modification timestamp on the image you downloaded?
>
> The image dated at Mar 4th 2018 which is the latest avaialble at upstream
> ftp.debian.org (file sizes, time stamps all matched with the upstream).
My guess is that somehow you used a stale file when you got that
> ftp.debian.org (file sizes, time stamps all matched with the upstream).
message. There are several ways to do that accidentally.
I've downloaded the .iso from the address you posted, and it has the
timestamp you just reported; to be precise (and precision helps):
40894464 Mar 4 13:38 /home/debian/mini.iso
Booting from this .iso reports this kernel version on VC2:
~ # uname -a
Linux (none) 4.9.0-6-amd64 #1 SMP Debian 4.9.82-1+deb9u3 (2018-03-02) x86_64 GNU/Linux
~ #
which should be compatible with the current version running the same
machine:
$ uname -a
Linux wren 4.9.0-6-amd64 #1 SMP Debian 4.9.88-1+deb9u1 (2018-05-07) x86_64 GNU/Linux
$
Perhaps you could reproduce your "bug" as before, and report the full
line as above.
Cheers,
David.
mizuki
Jun 21, 2018, 2:50:04 PM6/21/18
to
That was a good point, that made me re-download the latest netboot tar ball from upstream and extracted the kernel from that, made comparison against the new and existing kernels. although both were dated same timestamps/sizes, but they actually differ!
Went ahead and ran an actual install, ta-da, the kernel matched and it went through~!
That was awesome! Thanks for the hint and tips, very much appreciated!!!
Cheers.
Mizuki
0 new messages