New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Sluggishness/unresponsiveness in domU and dom0 when VM runs out of free memory #3780

Closed
0spinboson opened this Issue Apr 3, 2018 · 7 comments

Comments

Projects
None yet
3 participants
@0spinboson

Qubes OS version:

R4.0


Steps to reproduce the behavior:

Assign a given amount of memory to a VM (say 4GB), make the VM run out of memory (youtube videos, interactive/heavy websites), get sluggishness / stutter both in domU and dom0.

Expected behavior:

Graceful handling of this, no sluggishness/stutter bleeding into dom0.

Actual behavior:

VM becomes sluggish in a way that I never ran into on R3.2 / on R4 roughly prior to when the Spectre/MD stuff 'happened'. More importantly, dom0 also becomes intermittently unresponsive, which should just not happen. Video playback also

General notes:

Ryzen 5 1600, 32GB ram. dom0 kernel 4.15.6


Related issues:

@0spinboson

This comment has been minimized.

Show comment
Hide comment
@0spinboson

0spinboson Apr 4, 2018

A related symptom: after about 10m of video playback in a VM, the entire system freezes (including cursor). only fix being a hard reboot.

A related symptom: after about 10m of video playback in a VM, the entire system freezes (including cursor). only fix being a hard reboot.

@lunarthegrey

This comment has been minimized.

Show comment
Hide comment
@lunarthegrey

lunarthegrey Apr 5, 2018

What is your system load like when you check in dom0 (when this happens)?

What is your system load like when you check in dom0 (when this happens)?

@0spinboson

This comment has been minimized.

Show comment
Hide comment
@0spinboson

0spinboson Apr 5, 2018

basically idle in dom0, not very busy inside the VM either; the VMs that have the issues only have 4 or 8 threads (out of 12) assigned, and they're never all that busy when this occurs; total memory used also way below the maximum (generally at least 8gb unreserved).

0spinboson commented Apr 5, 2018

basically idle in dom0, not very busy inside the VM either; the VMs that have the issues only have 4 or 8 threads (out of 12) assigned, and they're never all that busy when this occurs; total memory used also way below the maximum (generally at least 8gb unreserved).

@lunarthegrey

This comment has been minimized.

Show comment
Hide comment
@lunarthegrey

lunarthegrey Apr 7, 2018

Have you checked logs like /var/log/messages, syslog on the VM? Anything in /var/log/xen/console/hypervisor.log in dom0?

Have you checked logs like /var/log/messages, syslog on the VM? Anything in /var/log/xen/console/hypervisor.log in dom0?

@0spinboson

This comment has been minimized.

Show comment
Hide comment
@0spinboson

0spinboson Apr 8, 2018

I did. I don't see anything there that screams impending doom to me, though.

I did. I don't see anything there that screams impending doom to me, though.

@lunarthegrey

This comment has been minimized.

Show comment
Hide comment
@lunarthegrey

lunarthegrey Apr 8, 2018

Maybe @marmarek would know how to debug this better.

Maybe @marmarek would know how to debug this better.

@0spinboson 0spinboson closed this Apr 16, 2018

@0spinboson

This comment has been minimized.

Show comment
Hide comment
@0spinboson

0spinboson Apr 16, 2018

seems to have been due to hardware acting up.

seems to have been due to hardware acting up.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment