Based upon the sheer number of updates being applied in our upcoming upgrade, Justin recommended that we do a POC update on a clone prior to attempting the live update. I agree complete, but have a few thoughts and questions as a follow up. Unfortunately the original ticket closed automatically, and replying does not re-open the same ticket.
My thoughts are as follows, please point me back in the right direction if I am off.
- Clone server, rename and re-IP the new clone
- Does this server need to be on a locked down subnet such that it does not attempt to communicate with the existing tentacles, and automatically upgrade them? Or does the target environment tentacles only get updated via a deployment, and not via a health check?
- Since the server thumbprint is shared between the two servers, will the tentacles listen for requests from either server? My assumption is yes, but I hate to assume.
- If the tentacles do get updated, will the tentacles still work with the original Octopus instance? Meaning if I run a test deployment via the clone to a QA environment, and the tentacles are upgraded, will the QA environment tentacles still work with the original server, or just the clone? Obviously I don’t wish to break the existing implementation.
- Run the upgrade on the clone as documented per these steps…
- Switch the clone server to Maintenance Mode
- Download latest MSI installer if not already done so.
- Run installer
- Turn off Maintenance Mode
- Attempt to access the UI via a host file entry locally
- Confirm UI is working as expected
- Attempt a deployment or two to test environments? The answers to the tentacle questions above will help with this question.
- Anything else I am not thinking of?
Thanks again for your help and quick turnaround on questions. It’s very much appreciated!