LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Newbie (https://www.linuxquestions.org/questions/linux-newbie-8/)
-   -   What log file reflects boot hang reason? (https://www.linuxquestions.org/questions/linux-newbie-8/what-log-file-reflects-boot-hang-reason-731633/)

Dims 06-09-2009 06:22 AM

What log file reflects boot hang reason?
 
I have system boot hang in specific hardware configuration. What log files should I analyse to determine hang reason?

I am booting the system two times. First time with "bad" config and having hang, second time with "good" config to be able to inspect logs.

Looking "messages" and "dmesg" files I see no messages belonging to bad boot (as I think). Do they stored somewhere or dropped due hang?

Thanks.

P.S. My distro is Fedora 9 and a hang occures before "starting udev" message.

pixellany 06-09-2009 06:27 AM

"boot" is sometimes ambiguous.....to some it mean the entire startup process, but to others is mean only the process leading up to handing control to the kernel.

For something like the GRUB bootloader, I don't think there are any error logs. Once the kernel takes over and runs "init", logs are written to /var/log.

In your case, something around "starting udev" should be in the logs.

JulianTosh 06-09-2009 06:29 AM

dmesg should indicate the failure. Mind posting the output? Also, what changed on the machine since the last boot?

Marko Hyvärinen 06-09-2009 07:35 AM

maby boot.msg etc, SuSE have it, does Fedora ?

malekmustaq 06-09-2009 07:54 AM

Dims,

Try analyse /var/log/syslog, Xorg.0.log, and /var/log/debug. There, at least whatever failed when kernel took over control it could be inferred easily.

By the way what were the messages when the bad config failed to boot?

We need more info regarding your problem so that we can provide help.

Goodluc.

johnsfine 06-09-2009 08:07 AM

Quote:

Originally Posted by Dims (Post 3567804)
I am booting the system two times. First time with "bad" config and having hang, second time with "good" config to be able to inspect logs.

Do you have a working liveCD?

If I were diagnosing a problem like that, I would want to preserve as much as possible of the hang, rather then have anything overwritten by the next startup. So after the hang, I would boot a liveCD and use that to look at any remains on the hard drive from the startup that failed.

Dims 06-09-2009 09:27 AM

I am trying to make graphicsless computer node. The situation specific is described here: http://www.linuxquestions.org/questi...onsole-730413/

Since it was impossible to find the specific answer, I would like to find the more general answer now: how to determine hardware responsible for hang in Linux?

Dims 06-09-2009 09:31 AM

Quote:

Originally Posted by malekmustaq (Post 3567880)
Try analyse /var/log/syslog,

Absent.

Quote:

Xorg.0.log,
I am not starting X11.

Quote:

and /var/log/debug.
Absent.

Quote:

By the way what were the messages when the bad config failed to boot?
I am looking at messages vis serial console and see some corrupted words during processors initializing.

The tail is

Code:

console [tty0] enabled
console [ttyS0] enabled
Checking aperture...
No AGP bridge found
PCI-DMA: Using software bounce buffering for IO (SWIOTLB)
Placing software IO TLB between 0x20000000 - 0x24000000
Memory: 6118344k/8388608k available (2852k kernel code, 164020k reserved, 1588k data, 1268k init)
SLUB: Genslabs=13, HWalign=64, Order=0-3, MinObjects=0, CPUs=8, Nodes=1
Calibrating delay loop (skipped), value calculated using timer frequency.. 5306.65 BogoMIPSIche hash tabl ns
subsyspuacct
U  Node
CPU: PhysiCcore ID: 0
CPU0:n hreads.
ExtINT Detected 8.29Iled: passed.
Bosor 2/Initializinr ic ro0 passed.
BootingInitializing CPU#3
Calibrating delayInitializing C007 new 0 passed.
Initializing CPU#5
p passed.
Booting priInitializing CPU#Initializing CPU#7
Calibratinae: redinterface drr:00:: PME# suppor0:ci 0000:00:1c, e ASPM for dI->A: PIC IRQ :

Quote:

We need more info regarding your problem so that we can provide help.
Thank you very much, would hope to find the cause with your help!

Dims 06-09-2009 09:34 AM

Quote:

Originally Posted by Admiral Beotch (Post 3567810)
dmesg should indicate the failure.

It seems to me that it shows data about last boot only. Since it is successful, I see no keys about hang.

Quote:

Also, what changed on the machine since the last boot?
I am trying to boot without graphics card. Boot hangs whet I take it out.

Dims 06-09-2009 09:37 AM

Quote:

Originally Posted by pixellany (Post 3567807)
Once the kernel takes over and runs "init", logs are written to /var/log.

In your case, something around "starting udev" should be in the logs.

It looks like hang occures before init but after kernel starts booting. Boot don't reach starting udev message.

Am I right thinking that this is a first init message?

jamescondron 06-09-2009 10:42 AM

Quote:

Originally Posted by Dims (Post 3567998)
It seems to me that it shows data about last boot only. Since it is successful, I see no keys about hang.


Look in /var/log, you should have multiple dmesg logs. Look at the relevant one

Dims 06-09-2009 01:43 PM

I have two one: dmesg and dmesg.old. The size of them is similiar. Is it possible to store dmesg of each log?

jamescondron 06-09-2009 01:49 PM

What do you mean store? They are stored, thats what those are, files.

thorkelljarl 06-09-2009 02:14 PM

?

What about trying to edit the boot process eliminating loading the kernel module that is looking for the graphics that aren't there?

Dims 06-09-2009 03:51 PM

Quote:

Originally Posted by jamescondron (Post 3568256)
What do you mean store? They are stored, thats what those are, files.

Sorry I ment store dmesg of each boot. I booted 100 times but have only 2 files.


All times are GMT -5. The time now is 11:03 PM.