Subject: Re: [vserver] Kernel panic - not syncing
From: Herbert Poetzl <herbert@13thfloor.at>
Date: Wed, 28 Aug 2013 13:04:50 +0200

On Tue, Aug 27, 2013 at 08:40:29PM -0700, Roderick A. Anderson wrote:
> On 08/27/2013 03:28 PM, Herbert Poetzl wrote:
>> On Tue, Aug 27, 2013 at 03:25:20PM -0700, Roderick A. Anderson wrote:
>>> On 08/24/2013 09:00 AM, Herbert Poetzl wrote:
>>>> On Fri, Aug 23, 2013 at 03:33:15PM -0700, Roderick A. Anderson wrote:
>>>>> I am trying to build a new CentOS 6.4 64bit host.

>>>>> Started with a Minimal Desktop, followed the instruction on the
>>>>> wiki, but when I reboot I get:

>>>>>  kernel panic - not syncing. No init found.
>>>>>  Try passing init= option
>>>>>  Pid: 1. comm: init Not tainted
>>>>> 2.6.32-358.14.1.el6.vs2.3.0.36.29.6.31.x86_64

>>>> sounds like the rootfs cannot be properly mounted and
>>>> thus the designated init cannot be found/executed.

>>>>> There are Call Trace: lines but I only have them as a image
>>>>> with my phone.  If they will help I'll transpose them to text.

>>>> still no serial (or other remote) console on a server?

>>> Anyway here is my transcription of the kernel panic screen for:

>>> [<ffffffff01565328>] ? panic+0xa7/0x16f
>>> [<ffffffff010022ec>] ? init_post+0xbc/0x110
>>> [<ffffffff01c2d91b>] ? kernel_init+0x2ed/0x2f3
>>> [<ffffffff0100c0ca>] ? child_rip+0xa/0x20
>>> [<ffffffff01c2d62e>] ? kernel_init+0x0/0x2fe
>>> [<ffffffff0100c0c0>] ? child_rip+0x0/0x20

>> basically just says that init could not be started,
>> the reason is most likely documented in the lengthy
>> log prior to this ...

> What lengthy log?  :-)  This was it!  

> Before those lines was the "Kernel panic" lines.  
> "Not syncing: No init found.", "Try passing init= 
> option", etc. etc.

Everything up to "Not syncing: No init found."

Make sure that your kernel is not passed the 'quiet'
command line option from the bootloader.

Then capture every line exactly as it is and
upload the log to a pastebin.

best,
Herbert

> Then "Call Trace:' before those lines.

> I am not a kernel hacker nor have I played the part of one in movies of 
> TV :-) but I am thinking more and more there is a problem with the 
> initramfs image for the most current version.

>    That is why I hope Daniel will catch this thread and offer some 
> thoughts or a previous kernel (source or binary) RPM.  Even the patch 
> files he used would be nice.

>    I could see about using the current RPM source and back porting (?) 
> its patch file(s) to the older kernel.


> \\||/
> Rod
> -- 

>> best,
>> Herbert

>>>>> This kernel is from Daniel's repo.

>>>>>   2.6.32-358.14.1.el6.vs2.3.0.36.29.6.31.x86_64

>>>>> The CentOS kernel is (which boots just fine):

>>>>>   2.6.32-358.14.1.el6.x86_64

>>> The other new Vserver host I am working on fine:

>>> kernel-2.6.32-358.2.1.el6.vs2.3.0.36.29.6.31.x86_64
>>> kernel-2.6.32-358.el6.x86_64
>>> kernel-firmware-2.6.32-358.2.1.el6.vs2.3.0.36.29.6.31.noarch

>>> Is it possible to get the source (or binary) RPM for this version?
>>> Daniel? :-)


>>> Rod
>>> --
>>>>> Any suggestions on how to get this working?

>>>> well, if you provide the boot log, it's probably simple
>>>> to diagnose where it goes wrong, possible reasons include
>>>> but are not limited to:

>>>>  - bad initramfs (old initrd tools)
>>>>  - driver for rootdev/rootfs missing
>>>>  - device names changed (hda ->   sda)
>>>>  - incompatible hardware

>>>>> Additionally: I'm about 35 miles from the server, no IPMI
>>>>> (yet), so I have to travel for each reboot/test attempt.

>>>> you must really like travelling then :)

>>>> best,
>>>> Herbert

>>>>> TIA,
>>>>> Rod
>>>>> --