Re: Fedora Rawhide-20161209.n.0 compose check report
On Fri, 2016-12-09 at 15:43 +0000, Fedora compose checker wrote:
Missing expected images:
Kde live x86_64
Kde live i386
Failed openQA tests: 11/90 (x86_64), 5/16 (i386), 1/2 (arm)
New failures (same test did not fail in Rawhide-20161208.n.0):
ID: 50861 Test: x86_64 Workstation-boot-iso install_default
URL: https://openqa.fedoraproject.org/tests/50861
ID: 50862 Test: x86_64 Workstation-boot-iso install_default@uefi
URL: https://openqa.fedoraproject.org/tests/50862
ID: 50864 Test: i386 Workstation-boot-iso install_default
URL: https://openqa.fedoraproject.org/tests/50864
Dependency issue between libvirt and xen;
http://koji.fedoraproject.org/koji/buildinfo?buildID=823764 fixes it.
ID: 50873 Test: x86_64 Server-dvd-iso install_default_upload
URL: https://openqa.fedoraproject.org/tests/50873
ID: 50889 Test: i386 Server-dvd-iso install_default
URL: https://openqa.fedoraproject.org/tests/50889
ID: 50941 Test: x86_64 universal install_kickstart_firewall_configured
URL: https://openqa.fedoraproject.org/tests/50941
These all timed out. But now I take a closer look, there's actually a
bug here: these aren't just taking too long and timing out, the
installer is actually *hanging* during the install process. They get to
post-install quite fast, then the spinner stops animating at some
point during post-install and they just sit there at a stuck screen
until the timeout is reached. We don't even get logs, because the
system does seem to be hung: when openQA tries to switch to a tty to
gather logs, it fails.
I'll have to run a few installs manually locally and see if I can
reproduce this.
ID: 50953 Test: i386 universal upgrade_2_desktop_32bit
URL: https://openqa.fedoraproject.org/tests/50953
This is another case of a problem that's shown up since upgrading the
openQAÂ boxes to Fedora 25: qemu is crashing with some SPICE errors. I
asked the SPICE folks about these and they say they probably indicate
some kind of guest error, but this never happened when the openQA boxes
were on Fedora 24. I'll file a bug on something.
15:23:50.4842 32146 QEMU:
15:23:50.4843 32146 QEMU: (process:32151): Spice-WARNING **:
display-channel.h:295:validate_surface: canvas address is 0x561113e0be10 for 0 (and is
NULL)
15:23:50.4844 32146 QEMU:
15:23:50.4844 32146 QEMU:
15:23:50.4844 32146 QEMU: (process:32151): Spice-WARNING **:
display-channel.h:296:validate_surface: failed on 0
15:23:50.4844 32146 QEMU:
15:23:50.4844 32146 QEMU: (process:32151): Spice-CRITICAL **:
display-channel.c:1666:display_channel_update: condition `validate_surface(display,
surface_id)' failed
15:23:50.5387 32146 waitpid for 32151 returned 0
(that's the qemu process dying right at the start of the test).
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net