[opensuse-factory] openQA update
After having set up openQA on OpenSuse 13.2 a month or so ago, I today accepted the updates that have been on offer for a while now. I found that: - the database and .ini and .conf settings here and there had been reset (no big deal) - I began getting the following error in autoinst-log.txt: DIE Can't call method "add_console" on unblessed reference at /usr/lib/os- autoinst/backend/qemu.pm line 598. at /usr/lib/os-autoinst/backend/baseclass.pm line 48 thread 1. backend::baseclass::die_handler("Can't call method \"add_console\" on unblessed reference at /us"...) called at /usr/lib/os-autoinst/backend/qemu.pm line 598 thread 1 backend::qemu::start_qemu(backend::qemu=HASH(0x4b1d1a8)) called at /usr/lib/os-autoinst/backend/qemu.pm line 86 thread 1 backend::qemu::do_start_vm(backend::qemu=HASH(0x4b1d1a8)) called at /usr/lib/os-autoinst/backend/baseclass.pm line 221 thread 1 backend::baseclass::start_vm(backend::qemu=HASH(0x4b1d1a8), undef) called at /usr/lib/os-autoinst/backend/baseclass.pm line 43 thread 1 backend::baseclass::handle_command(backend::qemu=HASH(0x4b1d1a8), HASH(0x4e72870)) called at /usr/lib/os-autoinst/backend/baseclass.pm line 374 thread 1 backend::baseclass::check_socket(backend::qemu=HASH(0x4b1d1a8), IO::Handle=GLOB(0x407f810)) called at /usr/lib/os-autoinst/backend/qemu.pm line 817 thread 1 backend::qemu::check_socket(backend::qemu=HASH(0x4b1d1a8), IO::Handle=GLOB(0x407f810)) called at /usr/lib/os- autoinst/backend/baseclass.pm line 174 thread 1 eval {...} called at /usr/lib/os-autoinst/backend/baseclass.pm line 141 thread 1 backend::baseclass::run_capture_loop(backend::qemu=HASH(0x4b1d1a8), IO::Select=ARRAY(0x4784bb8)) called at /usr/lib/os- autoinst/backend/baseclass.pm line 96 thread 1 backend::baseclass::run(backend::qemu=HASH(0x4b1d1a8), 14, 17) called at /usr/lib/os-autoinst/backend/driver.pm line 66 thread 1 backend::driver::_run(backend::qemu=HASH(0x4b1d1a8), 14, 17) called at /usr/lib/os-autoinst/backend/driver.pm line 53 thread 1 eval {...} called at /usr/lib/os-autoinst/backend/driver.pm line 53 thread 1 Line 598 in qemu.pm starts with : my $vnc = $testapi::distri->add_console( 'sut', 'vnc-base', { hostname => 'localhost', port => 5900 + $bmwqemu::vars{VNC}}); My understanding of perl is insufficient to deal with this message. Could someone please advise if there might be a workaround for this, or would I be better off to roll back to my previous installation? Thank you for any help. Vladimir Zawalinski -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Sunday 06 December 2015 19:04:57 vzawalin1@gmail.com wrote:
After having set up openQA on OpenSuse 13.2 a month or so ago, I today accepted the updates that have been on offer for a while now. […] DIE Can't call method "add_console" on unblessed reference at /usr/lib/os- autoinst/backend/qemu.pm line 598. […] Could someone please advise if there might be a workaround for this, or would I be better off to roll back to my previous installation?
The API has changed recently. Make sure "os-autoinst", "openQA" and the tests are up-to-date. Maybe just retrying the update might help and pull in more recent version of the packages. In the case you are using openQA from the OBS repository, there is also a stable variant which might be more useful for your usecase: https://build.opensuse.org/project/show/devel:openQA:stable Regards, Oliver -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 06/12/15 21:47, Oliver Kurz wrote:
On Sunday 06 December 2015 19:04:57 vzawalin1@gmail.com wrote:
After having set up openQA on OpenSuse 13.2 a month or so ago, I today accepted the updates that have been on offer for a while now. […] DIE Can't call method "add_console" on unblessed reference at /usr/lib/os- autoinst/backend/qemu.pm line 598. […] Could someone please advise if there might be a workaround for this, or would I be better off to roll back to my previous installation?
The API has changed recently. Make sure "os-autoinst", "openQA" and the tests are up-to-date.
Maybe just retrying the update might help and pull in more recent version of the packages.
In the case you are using openQA from the OBS repository, there is also a stable variant which might be more useful for your usecase: https://build.opensuse.org/project/show/devel:openQA:stable
Regards, Oliver
What are your tests you're running? You don't seem to have a distribution set in main.pm
Greetings, Stephan
Thank you both for your suggestions. Firstly I reinstalled OpenSuse 13.2 and then installed openqa from the stable OBS repository. Same error. I reinstalled 13.2 and followed exactly the procedure I used earlier to install openqa which worked from the develop repos
obs://devel:openQA/openSUSE_13.2 openQA obs://devel:openQA:13.2/openSUSE_13.2 openQA-perl-modules
as they were about 6 weeks ago. This time, the same error again. In all cases, I started from the slitaz live micro-distribution that Ludwig Nussel used in the how-to video attached to the openqa page on the Suse web-site and followed the instructions there. The breakage I experienced when applying the openqa updates as reported above was probably a warning. I actually had this problem about a week earlier. Then I used snapper to roll back the changes and the openqa setup worked again. Now that all repos have been updated to current or development, for me there is no going back! The error seems to be a perl run-time error.
DIE Can't call method "add_console" on unblessed reference at /usr/lib/os-
autoinst/backend/qemu.pm line 598.
Now, if openqa works on other machines, as I am sure it does, then the error in my environment must have been provoked by a weakness on my machine combined with the updates in 4.1449046989.3467058-785.17 The perl error seems to be saying that
method "add_console"
( bless REF. This function tells the thingy referenced by REF that it is now an object in the CLASSNAME package)
(my knowledge of perl fails here). Could a line of perl be crafted to provide what the error is complaining about? I am stuck here and happy to provide any other diagnostics and would really welcome help to move on. Thank you and Regards Vladimir Zawalinski -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 07.12.2015 10:38, vlad wrote:
as they were about 6 weeks ago. This time, the same error again.
In all cases, I started from the slitaz live micro-distribution that Ludwig Nussel used in the how-to video attached to the openqa page on the Suse web-site and followed the instructions there.
That's the key problem. the tests given by Ludwig were too basic, they lack a distribution and we require one now. We can just use a default one within os-autoinst. Please test this for me: http://paste.opensuse.org/32739523 Greetings, Stephan -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Monday, December 07, 2015 11:06:33 AM Stephan Kulow wrote:
On 07.12.2015 10:38, vlad wrote:
as they were about 6 weeks ago. This time, the same error again.
In all cases, I started from the slitaz live micro-distribution that Ludwig Nussel used in the how-to video attached to the openqa page on the Suse web-site and followed the instructions there.
That's the key problem. the tests given by Ludwig were too basic, they lack a distribution and we require one now.
We can just use a default one within os-autoinst. Please test this for me: http://paste.opensuse.org/32739523
Greetings, Stephan
Hello Stephan, thank you for the patch. I have applied it and attached the bwmqemu.pm as I modified it. This certainly bypassed the previous problem but fell over at DIE Can't use string ("distribution") as a HASH ref while "strict refs" in use at /usr/lib/os-autoinst/backend/baseclass.pm line 91. I have attached the the entire autoinst report as badhashref. I am trying all this as a feasability exercise for automating QA of Mageia releases, but initially used slitaz as per the video to get started. Previously I had specified the distribution through the DISTRI=..... parameter on the ./client jobs post DISTRI=slitaz ... Where is the distribution specified with these new changes? Adding a default is certainly a good idea for less experienced users like myself, but following the new specifications would be much better, of course. Thank you and regards Vlad
On 08.12.2015 10:51, vzawalin1@gmail.com wrote:
Where is the distribution specified with these new changes? Adding a default is certainly a good idea for less experienced users like myself, but following the new specifications would be much better, of course.
main.pm can set a distribution to handle mageia specific things - the examples should still work unmodified. No idea what went wrong for you. Greetings, Stephan -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 06.12.2015 um 09:04 schrieb vzawalin1@gmail.com:
Could someone please advise if there might be a workaround for this, or would I be better off to roll back to my previous installation? What are your tests you're running? You don't seem to have a distribution set in main.pm
Greetings, Stephan -- Ma muaß weiterkämpfen, kämpfen bis zum Umfalln, a wenn die ganze Welt an Arsch offen hat, oder grad deswegn. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
participants (4)
-
Oliver Kurz -
Stephan Kulow -
vlad -
vzawalin1＠gmail.com