On Tue, Mar 31, 2020 at 02:21:00PM +0200, Andrea Bolognani wrote: > On Mon, 2020-03-30 at 18:12 +0100, Daniel P. Berrangé wrote: > > On Mon, Mar 30, 2020 at 07:05:26PM +0200, Andrea Bolognani wrote: > > > My only concern is that our ci/Makefile scaffolding will bitrot now > > > that it will no longer be exercised directly through CI... Perhaps we > > > could leave a single Linux job enabled on Travis CI for the sole > > > purpose of preventing that? > > > > True, but I'd rather like to eliminate duplicate failures being reported > > by Travis that we're already getting from GitLab, and thus stop loooking > > at Travis as much as is possible, so if it fails we know it is macOS > > related. > > Since we're going to be forced to keep the macOS job on Travis, we > will still ultimately have to look in two places... Additionally, > while right now CentOS CI gives you an overview of the CI status of > all jobs, from libvirt all the way to virt-manager, once all projects > have been moved to GitLab CI we're going to lose that. > > Could we create a sort of high-level CI dashboard that provides an > at-a-glance status? It doesn't have to be very detailed, it could be > basically just a single colored dot per project to signal its overall > status, along with links pointing to the appropriate pages for when > additional details are necessary. That's doable with a combination of GitLab CI + Pages I expect. A script would fetch the job status from all the relevant places and render a static HTML page. We publish that to gitlab Pages, and setup a scheduled CI pipeline to refresh once an hour or so. Regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
