On Fri, Oct 18, 2024 at 04:25:07PM +0100, Peter Maydell wrote:
> On Fri, 18 Oct 2024 at 10:01, Daniel P. Berrangé <berra...@redhat.com> wrote:
> > Test reliability is a different thing. If a particular test is
> > flaky, it needs to either be fixed or disabled. Splitting into
> > a fast & slow grouping doesn't address reliability, just hides
> > the problem from view.
>
> On the subject of 'flaky', here's another low-repeatability
> intermittent with migration-test that I just ran into in
> 'make vm-build-openbsd':
> ――――――――――――――――――――――――――――――――――――― ✀ ―――――――――――――――――――――――――――――――――――――
> stderr:
> warning: fd: migration to a file is deprecated. Use file: instead.
> warning: fd: migration to a file is deprecated. Use file: instead.
> **
> ERROR:../src/tests/qtest/migration-helpers.c:322:check_migration_status:
> assertion failed (current_status != "failed"): ("failed" != "failed")
> qemu-system-ppc64: Failed to connect to '127.0.0.1:24109': Address
> already in use
> qemu-system-ppc64: Failed to peek at channel
>
> (test program exited with status code -6)
> ――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――
>
> Probably
> https://lore.kernel.org/qemu-devel/cafeaca8p9bkcrn9effxtphe+5w-_8zhte_52spzlus-+zpf...@mail.gmail.com/
I think I've finally found the root cause of this bug. Setting SO_REUSEADDR
on client sockets on OpenBSD causes re-use of ports in TIME_WAIT when
auto-assigning a local bind address for connections. I've sent a patch to
remove this, since it is essentially pointless todo this AFAIK.
With regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
