Ensure we trigger building new OBS packages first, then wait plenty of
time until the binary packages are available (and run jobs in the
meantime that don't need them), and only after that we run the jobs
that need the binary packages.
Otherwise TTCN3 jobs may test the packages from the previous day, and
some jobs may fail completely due to packages not being completely built
yet. For example, yesterday the new Osmocom CNI releases were tagged,
which means the :latest packages also need to be rebuilt (-> building
all OBS packages takes longer). The osmocom-release-manuals and
-tarballs jobs failed, because the new binary packages were not
available yet when they ran.
Change all timers to the format "H 20 * * *" to have a deterministic
hour and semi-random minute based on the job name.
Change-Id: Ib68f9a78bae27a63706a8c95715bf6a244b7bf1d
Now that we made a new release including the new .spec.in files, from
which the rpms are built, we have the rpms in the repository and can
enable the repo install test for latest.
Change-Id: I5da2b895d636b348e5aa0539a23fe4d99e8644ae
Add the slave axis again, so the jobs aren't stuck forever. I had
assumed that without the axis, it would run on any node, but that's not
how it works. Add a label for this job, with several nodes attached,
like we do it for TTCN-3, master-builds, gerrit-verifications etc.
Related: https://jenkins.osmocom.org/jenkins/label/repo-install-test/
Related: OS#4567
Fixes: fcf669 ("repo-install-test: run on all build slaves again")
Change-Id: I276ab47f76a0f4db542ca99825ebb019236b4d27
It was not possible to reproduce the weird rpm errors on
admin2-deb9build, which had lead to limiting the build slaves to
build2-deb9build-ansible. Enable building on all again.
Closes: OS#4567
Change-Id: I82ef1f0c581de8ee826adedd9ecde6b4adaa36ba
Give OBS more time to build the repository, before verifying that we can
install all packages from the repository. Apparently, OBS publishes the
repository in WIP state, before all packages for a distribution have
been built.
This leads to problems in the "nightly" and "next" repositories. In
contrary to "latest", we do not bump soname versions when doing ABI
changes, so we require the user to have all installed Osmocom packages
built from the same timestamp. With recent changes in the OBS scripts,
we enforce this by having all packages built from the same timestamp
depend on the exact version of a dummy package with that timestamp as
version.
The repo-install-test installs all packages from the binary repository,
and so it fails, as it should, if the repository is in an inconsistent
state with some packages built today and some packages built yesterday.
Related: OS#4733
Change-Id: I8df9b449d6213b5dca6fd9bf5c06b5c96d468f66
We use centos8 instead of centos in all docker-playground.git setups and
the infrastructure there expects that kind of naming.
Related: OS#4888
Change-Id: Idfbb2c4fc1ca10741406c8ab8930dabe8ce632ee
Rename osmocom-debian-install.yml to repo-install-test.yml to get debian
out of the name. Extend it with a new distro parameter and update the
description. Adjust the shell section to run the script from its new
location (in osmo-ci.git, not docker-playground.git). Turn it into a
matrix job, so we can have two parameters (distro, feed) for each job.
Related: OS#4563
Change-Id: I777098f19d75f7efbd68b837ccdcd83309429c39
Oliver Smith
Harald Welte
Pau Espin Pedrol