[Spacewalk-list] Unable to retrieve stage2.img ... but able to retrieve it. Weird.

[Michael Schwager]

Hello Astronauts,
We are using Spacewalk release 0.5.  We are Kickstarting our machines from
it.  We have created a channel for Kickstarting our RHEL5.3 machines, called
"rhel5.3_x86_64_clone".

When we try to Kickstart, it stops with the aforementioned "Unable to
retrieve stage2.img" error.  But if we move on without changing anything, it
works.  Details follow...

Details:
We create an iso file with our initrd.img and ks.cfg in it.  We mount that
and boot off of it.  In there, we have specified

url --url http://10.52.151.21/ks/dist/rhel5.3_x86_64_clone

Our Kickstart advances to the point where it says "Unable to retrieve
http://10.52.151.21//ks/dist/rhel5.3_x86_64_clone/images/stage2.img".
(notice the double slash, but I don't think it's a problem).

If I look at console 3, I can see things like

01:02:28 INFO   : transferring
http://10.52.151.21//ks/dist/rhel5.3_x86_64_clone/disk1/images/product.imgto
a fd
01:02:31 INFO   : 12296768 kB are available
01:02:31 INFO   : transferring
http://10.52.151.21//ks/dist/rhel5.3_x86_64_clone/images/stage2.img to a fd
01:02:34 INFO   : transferring
http://10.52.151.21//ks/dist/rhel5.3_x86_64_clone/disk1/images/stage2.img to
a fd

...and there it is stopped.  So I go back to Screen 1, I hit OK, It takes me
to the HTTP Setup screen.  I don't change a thing.  I notice that in the
"Red Hat Enterprise Linux Server directory:" box, the path ends in
"...dist/rhel5.3_x86_64_clone".  Regardless, I hit "OK".  Then I switch
immediately to console 3 and take a screenshot.

It does a bunch of transfers again, starting with
00:54:53 INFO  : trying to mount CD device scd0
... (then it tries to transfer images/updates.img "to a fd",
disc1/images/updates.img, images/product.img, and disc1/images/product.img.
The prefix to all of these is
http://10.52.151.21//ks/dis/rhel5.3_x86_64_clone/ .  Finally, we get to:)
00:54:54 INFO  : 12296768 kB are available
00:54:54 INFO  : transferring
http://10.52.151.21//ks/dist/rhel5.3_x86_64_clone/images/stage2.img to a fd

...and the Kickstart completes successfully.

Notice that that URL is the same as the one above, at 01:02:31.  I don't
know why it complained up above.

Now we have a 4-port NIC installed in our machine, and this has made Linux
think that our motherboard NIC is at "eth4", but on the network line in our
ks.cfg we specify "--device=eth4", and regardless the point that bugs me is
I don't change anything but I still need to do that manual step.  How
annoying.

Any ideas?  Thanks.
-- 
-Mike Schwager
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/spacewalk-list/attachments/20090828/89061bcc/attachment.htm>