[newbie] CoreOS iso failed to extract and install.

[leam hall]

Trying to install Container Linux on a Dell 980. Should be enough disk space, RAM, etc, and installing on bare metal. The ISO image from https://coreos.com/os/docs/latest/booting-with-iso.html  seemed to have issues. I downloaded it, did a "dd" to a 2GB memory stick, and then it had issues with unpacking. The journal said there was a tar file extraction issue. The second download was the same size as the first. 

I was using the latest stable coreos_iso_production_image.iso but it seems I missed a step somewhere. Ideas on how I can dig into this more? 

Thanks!

Leam

[David Michael]

On Tue, Jul 31, 2018 at 2:30 PM, leam hall <leam...@gmail.com> wrote:
> Trying to install Container Linux on a Dell 980. Should be enough disk
> space, RAM, etc, and installing on bare metal. The ISO image from
> https://coreos.com/os/docs/latest/booting-with-iso.html seemed to have
> issues. I downloaded it, did a "dd" to a 2GB memory stick, and then it had
> issues with unpacking. The journal said there was a tar file extraction
> issue. The second download was the same size as the first.

The only default tar extraction I recall is extracting the Docker
version, which could fail on the ISO if your system doesn't have
enough memory. The documentation says that at least 2GiB is required,
but more is always better.

> I was using the latest stable coreos_iso_production_image.iso but it seems I
> missed a step somewhere. Ideas on how I can dig into this more?

Can you post the exact error messages?

Thanks.

David

[Leam Hall]

David, I appreciate any help you can give! Here's a quick pull of the
"fail|error" lines. The entire log is at:

https://github.com/LeamHall/Ansible_Test/blob/master/tmp/song.rdsosreport.txt

#####

[leam@shaphan docker]$ egrep -i "fail|error" rdsosreport.txt
[ 0.304179] localhost kernel: acpi PNP0A03:00: _OSC failed
(AE_NOT_FOUND); disabling ASPM
[ 0.343685] localhost kernel: ACPI: \: failed to evaluate _DSM (0x1001)
[ 1.133854] localhost kernel: Initramfs unpacking failed:
uncompression error
[ 1.946875] localhost mount[319]: mount: /sysroot/usr: wrong fs type,
bad option, bad superblock on /dev/loop0, missing codepage or helper
program, or other error.
[ 1.946724] localhost kernel: squashfs: SQUASHFS error: unable to
read xattr id index table
[ 1.952694] localhost systemd[1]: sysroot-usr.mount: Failed with
result 'exit-code'.
[ 1.953106] localhost systemd[1]: Failed to mount /sysroot/usr.
[ 1.953502] localhost systemd[1]: Dependency failed for Initrd Root
File System.
[ 1.953816] localhost systemd[1]: Dependency failed for Reload
Configuration from the Real Root.
[ 1.954245] localhost systemd[1]: initrd-parse-etc.service: Job
initrd-parse-etc.service/start failed with result 'dependency'.
[ 1.954713] localhost systemd[1]: initrd-parse-etc.service:
Triggering OnFailure= dependencies.
[ 1.966215] localhost systemd[1]: initrd-root-fs.target: Job
initrd-root-fs.target/start failed with result 'dependency'.
[ 1.967152] localhost systemd[1]: initrd-root-fs.target: Triggering
OnFailure= dependencies.
[ 1.979308] localhost systemd[1]: Dependency failed for Root
filesystem setup.
[ 1.979578] localhost systemd[1]: initrd-setup-root.service: Job
initrd-setup-root.service/start failed with result 'dependency'.
[ 4.324602] localhost kernel: sd 6:0:0:0: [sdc] Asking for cache data
failed

[David Michael]

On Tue, Jul 31, 2018 at 4:10 PM, Leam Hall <leam...@gmail.com> wrote:
> On 07/31/2018 02:38 PM, David Michael wrote:
>>
>> On Tue, Jul 31, 2018 at 2:30 PM, leam hall <leam...@gmail.com> wrote:
>>>
>>> Trying to install Container Linux on a Dell 980. Should be enough disk
>>> space, RAM, etc, and installing on bare metal. The ISO image from
>>> https://coreos.com/os/docs/latest/booting-with-iso.html seemed to have
>>> issues. I downloaded it, did a "dd" to a 2GB memory stick, and then it
>>> had
>>> issues with unpacking. The journal said there was a tar file extraction
>>> issue. The second download was the same size as the first.
>>
>>
>> The only default tar extraction I recall is extracting the Docker
>> version, which could fail on the ISO if your system doesn't have
>> enough memory. The documentation says that at least 2GiB is required,
>> but more is always better.
>>
>>> I was using the latest stable coreos_iso_production_image.iso but it
>>> seems I
>>> missed a step somewhere. Ideas on how I can dig into this more?
>>
>>
>> Can you post the exact error messages?
>>
>> Thanks.
>>
>> David
>>
>
> David, I appreciate any help you can give! Here's a quick pull of the
> "fail|error" lines. The entire log is at:
>
> https://github.com/LeamHall/Ansible_Test/blob/master/tmp/song.rdsosreport.txt

The later failures are from the "SQUASHFS error: unable to read xattr
id index table" error, which I assume originated from the "Initramfs
unpacking failed" error. The ISO is working fine for me, though. Can
you verify your download was not corrupted by checking the signature?
Maybe also verify that it was written correctly to the memory stick.

https://stable.release.core-os.net/amd64-usr/current/coreos_production_iso_image.iso
https://stable.release.core-os.net/amd64-usr/current/coreos_production_iso_image.iso.sig

Thanks.

David

[Leam Hall]

I wrote the USB stick with:
dd if=coreos_production_iso_image.iso of=/dev/sdb


Here's the GPG output. Looks good but I'm not well versed in it.

###
[leam@shaphan Downloads]$ gpg --import CoreOS_Image_Signing_Key.asc
gpg: keyring `/home/leam/.gnupg/secring.gpg' created
gpg: /home/leam/.gnupg/trustdb.gpg: trustdb created
gpg: key 93D2DCB4: public key "CoreOS Buildbot (Offical Builds)
<buil...@coreos.com>" imported
gpg: Total number processed: 1
gpg: imported: 1 (RSA: 1)
[leam@shaphan Downloads]$ gpg --verify
coreos_production_iso_image.iso.sig coreos_production_iso_image.iso
gpg: Signature made Sat 28 Jul 2018 07:13:25 PM EDT using RSA key ID
0638EB2F
gpg: Good signature from "CoreOS Buildbot (Offical Builds)
<buil...@coreos.com>"
gpg: WARNING: This key is not certified with a trusted signature!
gpg: There is no indication that the signature belongs to the
owner.
Primary key fingerprint: 0412 7D0B FABE C887 1FFB 2CCE 50E0 8855 93D2 DCB4
Subkey fingerprint: 4D72 41B1 4AA4 7290 515D 6A8D 7FB3 2ABC 0638 EB2F

[David Michael]

The signature looks fine, but it would be good to verify it wasn't
corrupted when writing to the USB stick. Maybe something like this:

cmp --bytes=$(stat -c%s coreos_production_iso_image.iso) /dev/sdb
coreos_production_iso_image.iso

Or verify it works in a VM:

qemu-kvm -cpu host -m 4G -vga std -cdrom /dev/sdb

Thanks.

David

[Leam Hall]

There has to be something in how I'm copying it. If I mount the iso
image as a loopback and copy cpio.gz to the regular fileysystem, I can
gunzip it fine. Twice now I've tried to copy new installs of the ISO
from the USB stick and get the following error when trying to gunzip the
cpio.gz file:

gzip: cpio.gz: invalid compressed data--format violated

I should just use dd for the ISO write, right?

dd if=coreos_production_iso_image.iso of=/dev/sdb

Trying to avoid installing a couple dozen rpm for qemu-kvm...

Thanks!

Leam

[Benjamin Gilbert]

On Tue, Jul 31, 2018 at 7:05 PM, Leam Hall <leam...@gmail.com> wrote:

I should just use dd for the ISO write, right?

  dd  if=coreos_production_iso_image.iso  of=/dev/sdb

It might be worth trying the regular image instead.

--Benjamin Gilbert

[David Michael]

I think that should be okay. Do you have a different USB stick to test?

Thanks.

David

[David Michael]

Note that the linked image decompresses to over 4GiB, so you will need
a different USB stick to test this as well.

Thanks.

David

[Leam Hall]

Moved to a larger (8gb) USB stick and things work just fine. It could be
the age of the 2GB stick, but I don't think so. I was able to mount and
read it fine. Of course, I have been known to be wrong now and again. :P

Thanks guys!

Leam