============================

Proposed title of this feature request

============================

What is the nature and description of the request?

In OpenShift 4, there is no ability to implement image pull timeouts. That is because this setting is particular to the Docker runtime, and in OCP 4 we use the CRI-O runtime, which does not have an image pull timeout flag or argument. However, there is a request for feature enhancement[1] in CRI-O upstream asking that this ability be implemented in the code: https://github.com/containerd/containerd/issues/4984 | https://github.com/kubernetes/kubernetes/pull/87641 

In cases of low bandwidth the pull of images from external repositories fails, so a solution to be able to increase the timeout period for the image pull would be desirable as per following:

https://access.redhat.com/solutions/5429831

========================================================
Why does the customer need this? (List the business requirements here)

========================================================

This issue has been reported as critical and shows blocker for the ongoing project for Orange. The issue is that customer is planning to deploy Edge platforms within customer affiliates premises, and the repository remains central with the core deployment. When they are trying to pull some images from central repos to deploy on the Edge platforms and bandwidth performance is not always as expected (low network bandwidth), the customer faces timeouts pulling the needed images

========================================================

List any affected packages or components.

========================================================

• CRI-O.