Earlier, we mentioned how we should manage image updates. In that instance, we were focused on security updates to avoid bugs and exploits in production. Similarly, we can apply this concept to application fixes and releases.
Base images should be updated in critical image components and these changes do not happen very frequently. Usually, the application releases are weekly or even daily (or hourly, depending on numerous factors, such as business requirements and critical fixes).
Depending on how many containers are running based on a specific image, a new image release can be a big change. These changes can be done in a couple of minutes or they can take you an hour. However, the procedure using containers is very quick; let the orchestrator do its job. Kubernetes and Swarm will provide automated image updates and rollback and we will be able to manage how this deployment should be done, how many containers will update their images in parallel, how much...