Kickstarting with multiple network interfaces prevents kickstart from completing
Environment
- Red Hat Network Satellite or Proxy
- 5.4.1
- Kickstart
- Seen on Dell R710
Issue
-
Build from Satellite gives missing package information inconsistently.
-
Uncertain if this is a Satellite issue or a Red Hat Enterprise Linux (RHEL) issue. In Satellite, channels were cloned to a RHEL 5 build as well as another set for RHEL 6 build. These were created from the RHEL synchronized channels. Both channels were provisioned successfully a number of times. However, on certain hosts provisioning fails and the following error is presented:
libgomp-4.4.6-4.el6.x86_64.rpm cannot be opened. Missing file, corrupt package, or corrupt media
-
Every time a build is made on this particular Dell
R710
machine, it dies in the same place on thelibgomp
package as it follows:libjpeg-6b-37.x86_64.rpm cannot be opened. Missing file, corrupt package, or corrupt media.
Resolution
- The easier solution is to rerun the
satellite-sync
on your channels and checklog
files. More details on the Satellite based kickstart fails with "The file xyz.rpm cannot be opened. This is due to a missing file, a corrupted package or corrupted media" document. - A more in-depth way to clean up is to do as described in the following article, RHEL6 network installation failing with error 'The file ncurses-base-5.7-3.20090208.el6.x86_64.rpm cannot be opened.
-
Another option is to modify the
ksdevice=
variable as noted by one of the options below:- Add the
MAC
address of theNIC
tokickstart
from:ksdevice=A1:B2:C3:D4:E5:F6
- Use the "link" keyword to tell
kickstart
to use the interface that has a link status ofUP
:ksdevice=link
- Use the
bootif
keyword to tellkickstart
to use theboot
interface:ksdevice=bootif
(for this to work you needIPAPPEND 2
in yourpxelinux.cfg
) -
In order to prevent further issues, the following command can be run:
`ifconfig em2 down`
- Add the
Root Cause
During kickstart with two network interfaces cabled the network configuration is placed additionally on the non-pxe'd
network interface and thus causes the network to fail.
Diagnostic Steps
- Since the same
kickstart
is running on other machines, it is not a corruption issue. -
One other important thing is to always make sure to check in on your
db
tablespace. Use the following command:su - oracle db-control report
2.1. Keep it below 85-90% to be safe:
db-control extend
-
Check the
md5sum
of the file in question with one downloaded fromrhn
as shown from the second KCS article. - Assuming it is a network issue the retry should work, assuming you get the
drac tty
working usingctrl+alt+<F#>
, it will help todebug
. - Go through the RHEL 5 channel and clean up so it looks similar to the actual
DVD
. - The
NIC
gets configured twice. At boot thep1p1
,NIC
gets configured and that is how it starts its build. Later, as it is prepping for install, theem2
(embedded/onboard)NIC
gets configured with the sameIP
address so both stop responding.
This solution is part of Red Hat’s fast-track publication program, providing a huge library of solutions that Red Hat engineers have created while supporting our customers. To give you the knowledge you need the instant it becomes available, these articles may be presented in a raw and unedited form.
Comments