We may have a similar requirement.
We are just starting to look at the possibility of using containers for some of our workloads for production level services like Jira and Confluence that we need to run on-prem (which for various reasons we can't run with SAAS at this stage or in the public cloud).
We are just brainstorming the possibility of using something like Rancher to manage the Docker layer for the containers and persisting data through the use of volume drivers (we are a vmware shop so could in theory use; https://vmware.github.io/docker-volume-vsphere/
) so the data can survive if a container is started on a different container host. Some teams have used NFS but this typically introduces a single point of failure.
We are aware of vsphere integrated containers, but would ideally like to use something with less lockin to the vsphere hypervisor; https://www.vmware.com/products/vsphere ... iners.html
I'm busy contemplating setting up a lab for this and the implications from a backup perspective - has anyone worked with this stuff before and been able to use Veeam for backups (we don't really want to use file based backups as the volumes would contain 10000's of objects so file based backup would likely be too slow.
Any advice would be much appreciated as to the options.