I just found out that my Ubuntu 18.04 machine crashed 20 hrs ago and switched off.

How can I find out why? Is there an event logger?

Thank you for your help!

  • It could also be your hardware that shut itself down (not the OS). If cpu gets above a certain temp it can slow down, but if it reaches its max threshold (same with some other pieces of hardware) it just forces the psu to instantly shutdown (to protect itself from destruction). The OS doesn't know about this as hardware is just off. You won't find any logs in your OS if this is the case, but machine logs (if a good device that keeps it) will report the raising temperature & near threshold (but not the breach of threshold as it just turns off).
    – guiverc
    Commented Jan 21, 2019 at 8:39
  • kernel and syslog where both empty. I installed smartctl afterwards but there was nothing either. A few weeks I replaced the SSD from mSATA m1 to a SATA SSD. Never had any issues with the M1 SSD.
    – ssssstut
    Commented Jan 21, 2019 at 20:27

2 Answers 2


You can use dmesg command from terminal, it may show some useful information of the kernel from device drivers.

  • thank you, I couldn't find any suspicious information, however I'm a linux newbie
    – ssssstut
    Commented Jan 21, 2019 at 20:38

Consider this message a 'kick-starter' for a collaborative effor to resolve these kinds of mystery problems. It is not really an answer -- Yet. (See process).

**notion / seed **

It begins with this idea (I had)...

 mkdir ~/trace-system
 cd    ~/trace-system
 tail  -f /var/log/syslog | tee syslog.feed.log


My Ubuntu 18.10 system has been stable since I updated to 18.04 LTS.

Recently it developed a syndrome where I can boot and work on the screen for a good while -- Today it ran all day long. Only to resume the cycle of freeze/reboot just when I wanted to USE the blessed thing.

But there is no error or message or log entry that I've detected or find even interesting.

I set-up the above tail command in term on the hope that when the system freezes-up, the last thing few things in the file will be pointers or clues to the problem and/or cause.

So far that approach has yielded little information. True it does record the final last words of Ubuntu's syslog before the crash, not that interesting.

I feel that the procedure is on-track, it is time to add to the discussion and see how / if this may work with more appropriate "target" logs or commands -- Such as the dmesg suggestion posted earlier.

Looking forward to your additions/comments.


  • this post is a "seed" for an idea I had to capture these kinds of problems.
  • Others are invited to edit this post with methods that enhance the concept or get similar result.
  • If you can only comment on the ideas, that's OK. I will test your comment, if it proves workable or helps I'll update this post myself.
    • I can up-vote useful ideas we add to the body of knowledge.

Also, it is OK for you (others) to improve the process. I reserve the right to keep things trim.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .