NOAPIC doesn't seem to work

We're doing a full move of our 900+ servers from RHEL satellite/cobbler to
foreman. Most of the machines have been migrated without issue. There are a
few that kernel panic on PXE boot with the 'try noapic' error. In cobbler
we add --kopts 'acpi:off'. The subsequent pxe config produced looks like
this:

default linux
prompt 0
timeout 1
label linux
kernel /images/ks-6-5-server:1:XXX/vmlinuz
ipappend 2
append initrd=/images/ks-6-5-server:1:XXX/initrd.img
ksdevice=bootif lang= sshd biosdevname=0 text kssendmac acpi=off
ks=http://10.176.6.206/cblr/svc/op/ks/system/foo-4

In foreman it took me forever to find where to add kernel opts. (In the PXE
Linux template). It produces this:

#kind: PXELinux
#name: Community Kickstart PXE
#oses:
#- CentOS 5
#- CentOS 6
#- Fedora 16
#- Fedora 17
#- Fedora 18
#- Fedora 19
#- RedHat 5
#- RedHat 6
default linux
prompt 0
timeout 1
label linux
kernel boot/RedHat-6.5-x86_64-vmlinuz
ipappend 2
append initrd=boot/RedHat-6.5-x86_64-initrd.img
ks=http://foreman.local/unattended/provision ksdevice=bootif noapic nolapic
acpi=off console=tty0 console=ttyS1,56700n8 sshd biosdevname=0 text network
kssendmac

However the cobbler system builds perfectly fine without kernel panics and
the Foreman one panics every time both on RHEL 6.5. If I remove the
acpi=off option in cobbler the machines panic with the exact error as
Foreman. So I feel like the opiton is being ignored in Foreman.

This happens on only a handful of our machines like Supermicros, VirtualBox
(even with Enable I/O APIC unchecked) and a few one offs. Has anyone else
run into these issues? I don't see similar threads in the forum.

Foreman - 1.3.2-1
RHEL - 6.5

So this was a misdirected diagnosis. Turns out that the acpi=off flag was
working just fine and the initrd error that shows up during the kickstart
wasn't fatal. What was fatal though was having 'rhgb' enabled in the
bootloader options. Turned this off and I can now provision HREL 6.5 on
Virtualbox VMs, weird supermicros and employee desktop Dells even though
about 600 Dell servers in our multiple machine rooms had no issue with the
flag.

··· On Tuesday, June 17, 2014 1:25:28 PM UTC-4, Edson Manners wrote: > > We're doing a full move of our 900+ servers from RHEL satellite/cobbler to > foreman. Most of the machines have been migrated without issue. There are a > few that kernel panic on PXE boot with the 'try noapic' error. In cobbler > we add --kopts 'acpi:off'. The subsequent pxe config produced looks like > this: > > default linux > prompt 0 > timeout 1 > label linux > kernel /images/ks-6-5-server:1:XXX/vmlinuz > ipappend 2 > append initrd=/images/ks-6-5-server:1:XXX/initrd.img > ksdevice=bootif lang= sshd biosdevname=0 text kssendmac acpi=off ks= > http://10.176.6.206/cblr/svc/op/ks/system/foo-4 > > In foreman it took me forever to find where to add kernel opts. (In the > PXE Linux template). It produces this: > > #kind: PXELinux > #name: Community Kickstart PXE > #oses: > #- CentOS 5 > #- CentOS 6 > #- Fedora 16 > #- Fedora 17 > #- Fedora 18 > #- Fedora 19 > #- RedHat 5 > #- RedHat 6 > default linux > prompt 0 > timeout 1 > label linux > kernel boot/RedHat-6.5-x86_64-vmlinuz > ipappend 2 > append initrd=boot/RedHat-6.5-x86_64-initrd.img ks= > http://foreman.local/unattended/provision ksdevice=bootif noapic nolapic > acpi=off console=tty0 console=ttyS1,56700n8 sshd biosdevname=0 text network > kssendmac > > > > However the cobbler system builds perfectly fine without kernel panics and > the Foreman one panics every time both on RHEL 6.5. If I remove the > acpi=off option in cobbler the machines panic with the exact error as > Foreman. So I feel like the opiton is being ignored in Foreman. > > This happens on only a handful of our machines like Supermicros, > VirtualBox (even with Enable I/O APIC unchecked) and a few one offs. Has > anyone else run into these issues? I don't see similar threads in the forum. > > Foreman - 1.3.2-1 > RHEL - 6.5 >

So hmmmmm. How do I add [SOLVED] to the subject line.

··· On Tuesday, June 17, 2014 1:25:28 PM UTC-4, Edson Manners wrote: > > We're doing a full move of our 900+ servers from RHEL satellite/cobbler to > foreman. Most of the machines have been migrated without issue. There are a > few that kernel panic on PXE boot with the 'try noapic' error. In cobbler > we add --kopts 'acpi:off'. The subsequent pxe config produced looks like > this: > > default linux > prompt 0 > timeout 1 > label linux > kernel /images/ks-6-5-server:1:XXX/vmlinuz > ipappend 2 > append initrd=/images/ks-6-5-server:1:XXX/initrd.img > ksdevice=bootif lang= sshd biosdevname=0 text kssendmac acpi=off ks= > http://10.176.6.206/cblr/svc/op/ks/system/foo-4 > > In foreman it took me forever to find where to add kernel opts. (In the > PXE Linux template). It produces this: > > #kind: PXELinux > #name: Community Kickstart PXE > #oses: > #- CentOS 5 > #- CentOS 6 > #- Fedora 16 > #- Fedora 17 > #- Fedora 18 > #- Fedora 19 > #- RedHat 5 > #- RedHat 6 > default linux > prompt 0 > timeout 1 > label linux > kernel boot/RedHat-6.5-x86_64-vmlinuz > ipappend 2 > append initrd=boot/RedHat-6.5-x86_64-initrd.img ks= > http://foreman.local/unattended/provision ksdevice=bootif noapic nolapic > acpi=off console=tty0 console=ttyS1,56700n8 sshd biosdevname=0 text network > kssendmac > > > > However the cobbler system builds perfectly fine without kernel panics and > the Foreman one panics every time both on RHEL 6.5. If I remove the > acpi=off option in cobbler the machines panic with the exact error as > Foreman. So I feel like the opiton is being ignored in Foreman. > > This happens on only a handful of our machines like Supermicros, > VirtualBox (even with Enable I/O APIC unchecked) and a few one offs. Has > anyone else run into these issues? I don't see similar threads in the forum. > > Foreman - 1.3.2-1 > RHEL - 6.5 >