Uploaded image for project: 'Project Quay'
  1. Project Quay
  2. PROJQUAY-326

Develop Embedded Quay ('QuayCore') for OpenShift

XMLWordPrintable

    • Icon: Epic Epic
    • Resolution: Won't Do
    • Icon: Minor Minor
    • Alongside OpenShift 4.5
    • None
    • quay
    • None
    • Embedded Quay (QuayCore)
    • To Do

      *Goal: *Create an embedded version of Quay that can replace OpenShift's Internal Registry

      Problem:
      Duplicate registries exist under OpenShift organization; an opportunity exists to centralize on a single registry solution for OpenShift.

      Why is this important:
      Removes dependency on Docker distribution for Internal Registry. Leverages existing Red Hat product and expertise.

      Dependencies (internal and external):

      OpenShift Internal Registry

      Prioritized epics + deliverables (in scope / not in scope):

      As a user I can build and pull images using Quay exactly as is done with the Internal Registry.
      As a user I have access to some of Quay's value-add features (e.g. Time Machine).
      As a user I can see Quay UI from within OpenShift console

      *Estimate (XS, S, M, L, XL, XXL): *
      XL

      *Previous Work: *
      None

      Open questions:
      What sort of migration is expected for images from OIR into QuayCore? Could we run side-by-side for a while?
      Unclear what sort of Web UI integration needs to be done here?
      What sort of database solution is acceptable here (sqlite?) to ensure resilience?
      How much effort will it be to retrofit Quay to leverage Kube Auth/RBAC model?
      What Red Hat Quay features should be preserved in QuayCore that doesn't undercut Red Hat Quay's value proposition?

              Unassigned Unassigned
              bdettelb@redhat.com Bill Dettelback
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

                Created:
                Updated:
                Resolved: