-
Feature Request
-
Resolution: Unresolved
-
Undefined
-
None
-
6.16.z, 6.17.z, 6.18.z
-
None
-
False
-
sat-rocket
-
None
-
None
-
None
-
None
Q1 - What is the exact nature of the problem trying to be solved with this request?
A1 - The orchestrator when creating/registering/changing a host was silently failing because even though we were handling DNS externally, we were able to select a capsule for a subnet, so the orchestrator would attempt to send omapi updates to the capsule instead of the external DNS server, then when it got no response, it would simply stop the rest of the steps without an error. We discovered this because our DHCP records were consistently not reflecting our network environment being managed by foreman/satellite. We eventually noticed that the orchestrations tasks were suddenly stopping after attempting to update DNS about a record instead of finishing off DHCP, etc. Once we dove deeper we noticed messages that were indicating that it was trying to control a capsule DNS instead of an external DNS. Finally we noted that we have accidentally selected the "Reverse DNS Capsule" in the Subnet, and the "DNS Capsule" in the Domain, instead of leaving them blank. Apparently the orchestration or configurator that's running under the hood allowed that UI to take precedence over the settings we used in the `satellite-installer` command.
Q2 - List the business requirements.
A2 - I have no idea what this means.
Q3 - Do you have specific timeline dependencies?
A3 - We fixed it. We lost almost a month of development time in other projects trying to track down this issue though.
Q4 - Can you please explain us the what are the functional requirements? If test cases can be provided this would be even more ideal.
A4 - Configure the `satellite/capsule-installer` command to interact with an external DNS instead of the internal `named` service of the capsule. Then when configuring the Domain and Subnet for that particular capsule, use the dropdown to populate the capsule anyway. Then monitor the logs while registering or editing/saving hosts and you'll see no obvious failures, but if you dig down enough and watch the orchestrator tasks, you'll see that it suddenly just quits at the DNS stuff.
Q5 - Would you be able to assist in testing this functionality if implemented?
A5 - No, we have our Satellite environment running again and we have other deliverables to prioritize.