On 6/19/19 3:49 PM, Ken Dreyer wrote:
On Wed, Jun 19, 2019 at 12:03 PM Sage Weil <sweil@redhat.com> wrote:
I'm not sure I read this paragraph correctly. Are you saying we should use quay? Or that garbage collection is hard in general? Because I think the single oc adm prune command above is all that we need... we'll be publishing a zillion ci-built containers and trim them after 2 weeks. What's missing?
Yes, if I were doing this over, I would spend the time trying to make it all work with quay.io instead of running our own registry.
Running that oc prune command in cron is just the tip of the iceberg trying to operationalize this. For example with OpenShift 3.6, there was a bug with the way OpenShift (or docker-distribution, or whatever) writes incoming files to NFS, so "docker push" just fails. There is no workaround, and the only fix is to upgrade to newer versions of OpenShift 3.
We're already struggling to keep our Jenkins masters updated for security patches, and a public OpenShift/Kubernetes infrastructure will compound the problem.
I'm all for *not* running yet-another-service. But are quay.io going to be okay with us creating containers for every sha1? Christina and I did a rough estimate and it'd be about 500GB worth of containers at one time. Should we ask quay ahead of time if this is okay? I'd rather get a definitive answer ahead of time instead of set everything up, have them get mad, and then we end up needing to run our own registry anyway.