On Wed, 03/16 09:09, Alex Bennée wrote: > > Fam Zheng <f...@redhat.com> writes: > > > On Fri, 03/11 16:14, Alex Bennée wrote: > >> > >> Fam Zheng <f...@redhat.com> writes: > >> > >> > The script is not named test-travis.sh so it won't run with "make > >> > docker-run", because it can take too long. > >> > > >> > Run it with "make docker-run-travis.sh@ubuntu". > >> > >> 16:08 alex@zen/x86_64 [qemu.git/review/docker-v3] >make > >> docker-run-travis.sh@ubuntu > >> ARCHIVE qemu.tgz > >> COPY RUNNER > >> RUN travis.sh in ubuntu > >> ./run: line 49: /tmp/qemu-test/src/tests/docker/travis.sh: No such file or > >> directory > > > > Will update the commit message. > > I had a bit of a further play with this while trying to help with the > recent Travis breakage.
Did this help? > I realised we need to be clearer about what this > does. It's not the same as running on travis, just a way of iterating > through the travis build matrix on whatever image you happen to be on. Right. I have no idea how to precisely replicate travis environment, and I ignored python version, packages etc for simplicity, only the command matrix was simulated. But are there any major differences you are noticing? At least we are also on Ubuntu Trusty, the same as: https://docs.travis-ci.com/user/ci-environment/ Fam > > It would be nice to have a travis image for local debug but that seems > to be harder to do than I thought. I couldn't find any such images on > the hub.
