-
-
Notifications
You must be signed in to change notification settings - Fork 48
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
audiovm - Argument list too long - no sound for new qubes #8966
Comments
I wonder if |
When I had this bug I ran I augmented the watch quota per qubes to 300 with |
I reached 190. When I start a new dispvm, the number increase by 6. The number doesn't go down when the dispvm shutdown. Also, maybe the default value is a bit low, there are probably others things than starting vm that increase the number of watches. in the context of an audiovm, if the number is only increased when a VM is started, that limit the number of VM with working audio to 128/6=~21 (i agree, that already a big number of vm). However I am using a custom version of qvm-start-daemon.py (https://github.com/neowutran/qubes-core-admin-client/blob/f00ae868e444244419dae70453bcc3b88ffcf872/qubesadmin/tools/qvm_start_daemon.py ), I maybe created the issue myself, I don't think it is case but I will verify and report here when I have some time. update: |
If I start a vm, the counter increase by 6. If I start a vm, and shutdown the said vm without actually using audio, the counter have increased by 6. With my version of qvm_start_daemon.py, "sys-audio" start with a watches count of 10, meaning that with the default watches counts limit it can handle maximum ( 10 + x*6 < 128 => ) 19 vm not actually using audio before blocking any new vm from using audio |
Can you check process list in the audiovm? Perhaps some (pacat-simple-vchan) remain running after VM shutdown? |
It was indeed the case, and manually killing pacat-simple-vchan decrease the watches counter |
I tested to add that in qvm_start_daemon.py # def on_domain_stopped(self, vm, _event, **_kwargs):
# ...........
if xid != -1:
self.cleanup_guid(xid)
self.cleanup_pacat_process(xid)
if stubdom_xid != -1:
self.cleanup_guid(stubdom_xid)
self.cleanup_pacat_process(stubdom_xid)
def cleanup_pacat_process(self, xid):
try:
with open(self.pacat_pidfile(xid)) as f:
pid = int(f.readline())
os.kill(pid, signal.SIGTERM)
print(f"Sent SIGTERM signal to pacat-simple-vchan process {pid}")
except OSError:
print(f"Failed to send SIGTERM signal for the pacat-simple-vchan with xid of {xid}") seems to work, maybe there is a better way to do that |
Automated announcement from builder-github The package
|
Automated announcement from builder-github The package
|
Automated announcement from builder-github The component
|
Automated announcement from builder-github The component
|
Automated announcement from builder-github The component
|
Automated announcement from builder-github The package
|
Automated announcement from builder-github The component
|
As a note, how it work means that there is a maximum of ~19 qubes that can be running simultaneously without using audio (but with a audiovm configured). After that number, when a new qube is created, sys-audio will stop working. should probably be mentioned in the future official sys-audio documentation + a note mentioning that it is interesting to unset the audiovm parameter for qubes that will never need audio (all the sys-* vm for example) |
Automated announcement from builder-github The component
|
Please open a separate issue or a doc PR. |
Automated announcement from builder-github The package
|
Automated announcement from builder-github The package
|
Automated announcement from builder-github The component
|
Automated announcement from builder-github The component
|
Automated announcement from builder-github The component
|
Automated announcement from builder-github The component
|
Qubes OS release
R4.2
Brief summary
For audiovm, after some time "qvm-start-daemon.py" print this kind of errors:
Consequence is that new qubes don't have a working audio
Steps to reproduce
No reliable step to reproduce yet.
Starting and stopping qubes maybe have an effect
The text was updated successfully, but these errors were encountered: