On Thu, Nov 14, 2019 at 4:18 PM Nir Soffer <nsoffer@redhat.com> wrote:

On Thu, Nov 14, 2019 at 4:45 PM Dominik Holler <dholler@redhat.com> wrote:

...

Hello, how can I interpret the state of https://gerrit.ovirt.org/#/c/103739/ and what is the way to merge it? Just ask Edy to overwrite, or is there a more clean way possible how I

can solve this myself?

...

Thanks

You are using some non-standard marks:

@suite.XFAIL_SUITE_MASTER('TODO')

So I don't know what you do, but given the time out for tests that could not work, maybe you are missing run=False in the xfail:

pytest.mark.xfail(on_centos("8"), reason="...", run=False)

This is in https://gerrit.ovirt.org/#/c/103739/28/network-suite-master/testlib/suite.py...

With this the test is mark as expected failure without running it.

I am sure that the code change is fine. I wrote this mail to learn about the new gating process.

Nir

On Thu, Nov 14, 2019 at 4:36 PM Nir Soffer <nsoffer@redhat.com> wrote:

On Thu, Nov 14, 2019 at 5:24 PM Dominik Holler <dholler@redhat.com> wrote:

...

...

You are using some non-standard marks:

@suite.XFAIL_SUITE_MASTER('TODO')

So I don't know what you do, but given the time out for tests that could not work, maybe you are missing run=False in the xfail:

pytest.mark.xfail(on_centos("8"), reason="...", run=False)

This is in https://gerrit.ovirt.org/#/c/103739/28/network-suite-master/testlib/suite.py...

Strange usage of UPPERCASE for mark functions, but the xfail() looks correct.

Ack, make sense to handle decorators as functions.

...

I wrote this mail to learn about the new gating process.

Hopefully the maintainer of this code can merge the change without waiting for the gating machine.

The problem might be, that the gating fails, if any dependency in any test suite fails. In this case, it would be helpful to have some kind of feedback, why the gating failed, e.g. on which suite.