QVM Manager restart fails periodically

[cubit]

Hello

I am testing the new "restart" feature in the Qubes VM Manager and noticing that it is a bit buggy.   There is one piece of the puzzle however that I can not figure out allowing me to replicate it consistently or narrow down why it fail some times

The first time I  select a VM and press "restart"  after an indeterminate time frame the restart will fail with the error below.   If I try and restart any number of other vm immediately after they will all restart ok.

error:

  ---
  line: assert not vm.is_running()
  func: start_vm
  line no.: 1158
  file: /usr/lib64/python2.7/site-packages/qubesmanager/main.py
  ----
  line: self.caller.start_vm(vm)
  func: check_if_vm_has_shutdown
  line no.: 288
  file: /usr/lib64/python2.7/site-packages/qubesmanager/main.py

cubit

[Andrew David Wong]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Thanks for the report. Tracking:

https://github.com/QubesOS/qubes-issues/issues/2362

- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----
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=rXJm
-----END PGP SIGNATURE-----

[cubit]

4. Oct 2016 12:31 by a...@qubes-os.org:

Thanks for the report. Tracking:

https://github.com/QubesOS/qubes-issues/issues/2362

Some unscientific testing.

Trying a restart after 15 minutes of not doing a restart and the function works as expected

Trying a restart after 50 minutes of not doing a restart and the function fails the first time with the error as mentioned previously.

[Jeremy Rand]

cubit:

I can confirm that I encountered this today. It was with a VM that had
been running for several days. I've encountered the same error (or one
similar to it) in the past when trying to shut down a VM that takes a
long time -- Qubes Manager asks me if I want to kill it, and if I say
yes, sometimes there's some apparent race condition and the error comes
up (I guess because the VM finished shutting down right before Qubes
Manager tried to kill it).

Cheers,
-Jeremy

[cubit]

4. Oct 2016 16:04 by jerem...@airmail.cc:

I can confirm that I encountered this today. It was with a VM that had
been running for several days. I've encountered the same error (or one
similar to it) in the past when trying to shut down a VM that takes a
long time -- Qubes Manager asks me if I want to kill it, and if I say
yes, sometimes there's some apparent race condition and the error comes
up (I guess because the VM finished shutting down right before Qubes
Manager tried to kill it).

For mich I do not see any request to kill the AppVM and the uptime of the VM is very low. for example 50 minutes when I was testing the function to try and get it to error.     

[raah...@gmail.com]

happened to me a couple times, when I tried to restart sys-usb and shut down sys-net at the same time. Sys-usb would fail to start ask me to kill it, saying there is a possible bug in qubes manager.

Now I just wait for sys-net to fully shutdown, wait a sec and then restart sys-usb and haven't had the issue.

[Jeremy Rand]

raah...@gmail.com:

Yes -- like a lot of race conditions, it happens more often when the
system is under heavy load. I see this behavior more often when
starting/stopping multiple VM's at once, or when the system is otherwise
under heavy CPU or I/O load.

Cheers,
-Jeremy