I hope the CentOS people don't mind me cross-posting for their input.
I've just done a clean, text-mode minimal WBEL4 install on an IBM Netfinity 4500, essentially the same hardware as an xSeries 340.
The install went fine but now the box will boot, run through the initrd init sequence, perform its switchroot then hang here...
SELinux: Initializing. SELinux: Starting in permissive mode
I don't think it's an SELinux issue though, booting with selinux=0 the system just hangs after the switchroot because it's not initializing SELinux at all. It's what's coming next that's causing the problem.
On a working system (different hardware) the next stage is INIT...
INIT: version 2.85 booting
I've tried adding some debug echo lines into rc.sysinit incase it's something in there causing the problem but the init process doesn't appear to be getting that far. So, somewhere between the initrd's switchroot and spawning INIT something's broke.
It'll boot into rescue mount and I can chroot into the root filesystem and everything in there seems to be as it should. I've tried booting with selinux=0, noapic and noprobe with no change.
Where next? All suggestions appreciated.
Will.
On Fri, 2005-09-16 at 12:43 +0100, Will McDonald wrote:
I hope the CentOS people don't mind me cross-posting for their input.
I've just done a clean, text-mode minimal WBEL4 install on an IBM Netfinity 4500, essentially the same hardware as an xSeries 340.
The install went fine but now the box will boot, run through the initrd init sequence, perform its switchroot then hang here...
SELinux: Initializing. SELinux: Starting in permissive mode
I don't think it's an SELinux issue though, booting with selinux=0 the system just hangs after the switchroot because it's not initializing SELinux at all. It's what's coming next that's causing the problem.
On a working system (different hardware) the next stage is INIT...
INIT: version 2.85 booting
I've tried adding some debug echo lines into rc.sysinit incase it's something in there causing the problem but the init process doesn't appear to be getting that far. So, somewhere between the initrd's switchroot and spawning INIT something's broke.
It'll boot into rescue mount and I can chroot into the root filesystem and everything in there seems to be as it should. I've tried booting with selinux=0, noapic and noprobe with no change.
Where next? All suggestions appreciated.
try it with:
nousb
On 16/09/05, Johnny Hughes mailing-lists@hughesjr.com wrote:
On Fri, 2005-09-16 at 12:43 +0100, Will McDonald wrote:
The install went fine but now the box will boot, run through the initrd init sequence, perform its switchroot then hang here...
SELinux: Initializing. SELinux: Starting in permissive mode
I don't think it's an SELinux issue though, booting with selinux=0 the system just hangs after the switchroot because it's not initializing SELinux at all. It's what's coming next that's causing the problem.
On a working system (different hardware) the next stage is INIT...
INIT: version 2.85 booting
I've tried adding some debug echo lines into rc.sysinit incase it's something in there causing the problem but the init process doesn't appear to be getting that far. So, somewhere between the initrd's switchroot and spawning INIT something's broke.
It'll boot into rescue mount and I can chroot into the root filesystem and everything in there seems to be as it should. I've tried booting with selinux=0, noapic and noprobe with no change.
Where next? All suggestions appreciated.
try it with:
nousb
Same symptoms unfortunately. I tried nousb on its own then in combination with noprobe and noapic.
Will.
On 16/09/05, Will McDonald wmcdonald@gmail.com wrote:
On 16/09/05, Johnny Hughes mailing-lists@hughesjr.com wrote:
On Fri, 2005-09-16 at 12:43 +0100, Will McDonald wrote:
The install went fine but now the box will boot, run through the initrd init sequence, perform its switchroot then hang here...
SELinux: Initializing. SELinux: Starting in permissive mode
Right, after breaking the box open to see if removing the PCI 3com NIC and using the on-board AMD NIC (which used to cause all sorts of problems with older RH and FC releases) it seems to be sorted.
It wasn't the NIC though, it turns out the machine was booting but complaining during POST about one CPU. I assumed this would be because it had a VRM for a 2nd CPU but no actual processor in there. I was wrong. Previously it ran FC2 fine with a uni-processor kernel.
I reseated both processors which cleared the POST errors, reinstalled and that sorted the problem! The installer noted both processors and installed the correct SMP kernel, I couldn't be bothered fighting to get the RPM installed in a chrooted rescue environment. No doubt it's no-able but probably would've taken as long. :)
Will.
-
Johnny Hughes -
Will McDonald