Goal
Allow VMs connected through their secondary networks to be addressable by their domain names.
User Stories
- As a developer,
I want to address machines in the network using their domain names,
so I don't rely on IP addresses specific to given environment.
For example, something like ping -4 interfacename.vmname.namespace.vm.clustername.domainname should resolve to the relevant IP address.
I would not mind if ping -4 interfacename.podname.namespace.pods.clustername.domainname works, too.
- As a developer,
I want this feature to be opt-in per interface,
since not all interfaces are available from the outside, or meant to be used by clients.
Non-Requirements
- <List of things not included in this epic, to alleviate any doubt raised during the grooming process.>
- It is expected that the secondary network is already available to the client
Notes
- An implementation may require the cluster-admin to request a NS entry in the organizational DNS, delegating name resolution of vms.clustername.domainname to the API IP address of the cluster, where CoreDNS would be running.
- We could also consider doing it for internal secondary network
- This should work for L3 (unlike services)
Q&A
- Addressable from outside the cluster? - Yes
- From inside the cluster? - This is secondary, nice to have
- Will be pod network connected? - Not a must
- Cluster DNS or outside? - Customer does not care
Owners
| Role |
Contact |
| PM |
TBD |
| Documentation Owner |
TBD |
| Delivery Owner |
(See assignee) |
| Quality Engineer |
(See QE Assignee) |
Done Checklist
Epic
Normal
