[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: various iotests failures apparently due to overly optimistic timeout
From: |
Peter Maydell |
Subject: |
Re: various iotests failures apparently due to overly optimistic timeout settings |
Date: |
Mon, 20 Jul 2020 15:21:13 +0100 |
On Mon, 20 Jul 2020 at 15:18, John Snow <jsnow@redhat.com> wrote:
>
> On 7/20/20 10:15 AM, Peter Maydell wrote:
> > On Mon, 20 Jul 2020 at 15:12, John Snow <jsnow@redhat.com> wrote:
> >>
> >> On 7/20/20 6:46 AM, Kevin Wolf wrote:
> >>> John, I think this is a result of your recent python/qemu/ changes that
> >>> make failure of graceful shutdown an error rather than just silently
> >>> falling back to SIGKILL.
> >>>
> >>> Should the default timeout be longer, should iotests override the
> >>> default, or should iotests just always kill their VM instead of trying
> >>> to shut it down gracefully?
> >
> >> Let's make it longer. Since you are seeing the failures, can you edit
> >> python/qemu/machine.py and find the default timeout of 3 seconds for
> >> shutdown() and change it to 10? 30? 60?
> >
> > Not conveniently, because this is for merge pull requests, and
> > all the iotests failures are nested inside the tests/vm BSD VM
> > setup.
>
> I meant permanently, for everything -- not as a one-off for just this case.
Can't conveniently do that either, because the merge build runs on
head-of-master plus the thing being merged, so it won't have local
edits to machine.py in it, only the version that's in master.
-- PMM