I'm running MATE 22.04 with 32GB ram with an NVIDIA graphics card. I use Virt-manager to run various machines.
The host and or guest machines crash intermittently. Usually the host or the guest will crash or just reboot.
I've seen two sets of errors, on the host machine and on the virtual machine.
On the host I've seen errors such as "Ubuntu has experienced an internal error" or "Virtual Machine Manager has stopped responding".
On the guest machine, also MATE 22.04(doesn't seem to matter what distro it is either), it will just freeze, sometimes its the host, sometimes its just the guest. I'm usually(mostly) using Firefox and Thunderbird. Nothing that I would consider very demanding especially on a machine with 32GBs of mem. When I run top, resources used are under 1%.
There is about a week, maybe two, where everything runs "tip-top" and is very smooth and a pleasure to use. Then I get an update that breaks something, not sure what, and the cycle starts all over again. Rather frustrating when trying to get something done.
I've been looking through the syslog in /var/log but haven't found a smoking gun so to speak. Is this the correct file to look in? Based on the errors I've received I suspect its Virtual Machine Manager but I don't like to assume. Is there a log I can post that someone can take a look at the next time it happens to at least get me looking in the right direction?
(Usual disclaimer: please note that I'm just another Forum user here. I'm NOT a Ubuntu or Ubuntu MATE developer).
I don't use "virt-manager" in Ubuntu / Ubuntu MATE, but let me try to help a bit: regarding your virtualization "host machine" - and considering that you've mentioned that you got (popup) messages saying "Ubuntu has experienced an internal error" - I would also look at the /var/crash directory.
Do you have any files with extension .crash in that /var/crash folder? If so, what are their filenames? I also suggest that you open those ".crash" files (with a text editor) and see their content. Usually, the more informative lines (at least for me) in those .crash files are the lines that begin with:
ProblemType: (usually that line says ProblemType: Crash)
ExecutablePath: (which has the full path for the program that crashed)
ProcCmdline: (which has the line that was was used to start that program)
Title: (unfortunately, in some ".crash" files that "Title:" line does not exist)
Package: (the name and version of the package of the program that crashed)
If you're interested, you may also want to take a look at the "StacktraceTop:", "Stacktrace:", "ThreadStacktrace:", "XsessionErrors:" and "JournalErrors:" sections also in that ".crash" file, but please be aware that some of those sections may contain a LOT of information and that some of that information, in some cases, you may NOT want to disclose in a public forum.
You may also like to check, inside your Ubuntu "virtual machines", the content of that /var/crash directory.
Hello,
IMHO, you have pointed out the most probable reason of crashes already.
Well, an update changes some binaries on disk. Processes which are loaded into memory from previous versions of their respective binaries have to be restarted. Without restart such the processes may misbehave and crash.
To test this hypothesis I'd propose to explicitly restart your system after each update.
Thanks for the tip. Looking in /var/crash shows _usr_bin_virt-manager.1000.crash and that has a file titled "virt-manager crashed with SIGSEGV. When I search that I find a bug report. Thanks for the help!
