Boot failure

[Michael Powell]

Identry wrote:

>> Try downloading and booting the livefs environment (I think you need cd1
>> and the livefs cd or just the DVD) and see if you can mount it from
>> that, if not it could be a controller issue. If you can then its
>> probably your OS/kernel but at least you now have access to your
>> data/configs etc etc not to mention you could try extracting the GENERIC
>> kernel from the install media (use the install.sh script in the kernels
>> directory.)
> 
> Well, I am in the data center (2 hr drive, unfortunately)...
> 
> This is an Intel mother board. The front panel light labeled '!' is
> lit. It isn't lit on the working server. I'm googling right now for
> the meaning of this light, but if anyone knows off hand, please let me
> know...
> 
> I have the cd1 and cd2, but not the livefs cd. I'm going to try to
> find that right now.
> 

I do not know exactly what the light is indicating, such things are usually 
located in the hardware docs that came with the server. I would hazard a 
guess that it is indicating a hardware failure. If you are exceedingly lucky  
it might not be a FreeBSD issue as long as the data on the hard drive(s) has 
not been corrupted.

If it were me, the very first thing I'd do is power down and disconnect the 
drives. I'd install for temporary testing purposes any old spare blank hard 
drive I had laying around. If it is a brand name server there may have been 
included a diagnostics CD. Boot from that and see what happens. 

Next up is a boot to the BIOS configuration screen. When you power up the 
first item normally displayed is the text from the video ROM initialization. 
After this should be some form of announcement on how to get into BIOS 
config. Press whatever key and enter. Look for one of the preconfigured 
options such as "BIOS Defaults". If you can select this and save the board 
will be set to a fairly fail-safe set of defaults. Note that what this is 
telling you is the video and motherboard are initializing. If you cannot get 
to this point you may have a dead motherboard. Most boards will emit some 
form of beep code if the video ROM fails to initialize. 

Next up is if the board proceeds past this point watch for drive controller 
initialization. Most (most notably RAID) server controllers may have a 
message display when the controller ROM initializes indicating some form of 
key-press combo to enter the controller configuration. An example would be 
"press Ctrl-A" you'll see from an Adaptec card/chip.

If you cannot get to any of these stages consider either dead motherboard or 
power supply problem. Easiest way to confirm/eliminate a power supply is to 
substitute a known 100% functional one and see if you can now get to the 
afore mentioned stage(s) of boot. Power supply problems can sometimes 
manifest as hard drives that don't want to spin up. Listen and you can 
usually tell if they spin up, or not.

Your trouble sounds most like hardware failure. And because in your first 
email you did indicate a "Trying to mount root..." error most of the above 
described basic troubleshooting will end up being either dead hard drive(s) 
or malfunctioning controller. The reason I would have substituted a known 
good "scratch" drive earlier is twofold: if it can boot or install or 
otherwise initialize the controller it is indicative that the controller is 
OK and the problem is most likely dead drive(s). Secondly, you don't want to 
take any chances on damaging the data yourself with all this mucking about. 

If there has been drive failure you will need to replace, reinstall, and 
restore. If it has been a controller failure you will never have any success 
either booting, or installing a minimum system to the "scratch" drive. A 
controller failure also has the possibility of having already destroyed your 
data. Again, replace, reinstall, restore will be the order of the day.

This is just some quickly thrown out stuff in hopes it may be useful to you. 
It sounds like a hardware failure and not only do you have to deal with that 
first, you may also have to reinstall and/or restore from backup as well. I 
typed this up rather in a hurry, (so it's a little 'scrambled') but I think 
if read in totality you get the idea on things you can do to isolate and 
resolve. Good luck to you in any event.

-Mike