[WIP] GitHub action to deploy OCI image for tags#1
Conversation
|
Is this ready for review? |
I think it's worth a look, but up to you. Sounds to me like getting an updated version of the Template Provider up for testing is a high priority, so I wanted to push something workable early. I'm continuing to explore scheduling and a more automated approach for merging our build process into the upstream changes. |
|
I am not fully sure why we need to maintain 'dmnd' branch. Can elaborate please? |
I'm not sure I have a 100% bulletproof answer for you, as I'm still working toward a better overall solution. The idea with the
Either way, our tags will be a different Git commit hash from the upstream, since they need to include the workflow code in order to run it. So, using Haven't been able to fully think through all the consequences yet, so thought trying to contain our changes in a |
IMPORTANT: Read the plan below and if agreeable, create a
dmndbranch and set it to the main branch. After that, this PR should be edited to merge intodmnd, notmaster.This work aims to help with deploying new versions of the
sv2-tptemplate provider by acting on new GitHub tags to run Pulumi. A few thoughts about how the DMND fork could be structured and the corresponding workflow:dmndbranch in the repository and designate it as the "main branch".stratum-mining/sv2-tpupstream directly should take special care to go tomaster.dmndbranch against that tag to capture the latest changes along with the workflow definition.demand-open-source/sv2-tprepository to trigger the workflow and deployment.There is some room to rethink how to manage the code updates from upstream (scheduling, using merges for example), but there is at least one major limitation; the workflow code must exist in the branch or tag in order to run it under GitHub Actions. I suspect some degree of trial and error over time would be beneficial before settling on too complex a solution. Hopefully what is here is helpful toward getting updated code deployed for the Template Provider.
In following iterations, we can explore: