The pre-caching feature of the TALM operator is a valuable tool for reducing the time and bandwidth required to upgrade OpenShift clusters by pre-loading container images onto the cluster nodes prior to the upgrade ops.

However, currently, this feature pre-caches all the release images for a target release upgrade (i.e. across x-, y-streams), including images that are not strictly required for the upgrade (e.g. alibaba, aws, gcp, vsphere, etc.). This results in unnecessary network bandwidth usage, especially when a large number of managed clusters are being upgraded simultaneously or at a given time.

Consequently, this feature enhancement will benefit large-scale OpenShift cluster deployments by reducing the network bandwidth required for simultaneous upgrades and improving the efficiency of the pre-caching feature.

Additionally, it will minimize the impact of upgrades on production workloads, as less time will be spent downloading unnecessary container images during the upgrade process.

We believe that this enhancement will significantly improve the efficiency and adoption of the pre-caching feature, and would contribute to positioning the TALM operator as a go-to solution when massive upgrades of OpenShift clusters are needed.

Upstream project of TALM:

• https://github.com/openshift-kni/cluster-group-upgrades-operator