You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently we use a static assets stage in our docker image build to generate static assets required for addons-server. These are copied into the final production stage and pushed along with the image. During deployments, we pull the image, extract these files and upload them to GCP bucket for use in production.
This is several round trips and a lot of image bloat for no real benefit. We build our image during the ci.yml when we publish a new release. Here is where we build the image and that is where we can upload the assets.
Then the deployment job can access the static assets independently of the image, or even remove the need to uplaod at all if we upload directly. Need to coordinate with @bqbn on the structure of our bucket and what is possible in terms of uploading assets. Two possible options:
upload assets to github asset storage (accessible by url embedded in the workflow run json) and let the deployment job download and reupload to gcp
upload assets directly to gcp and deployment does nothing with assets
Acceptance Criteria
The content you are editing has changed. Please copy your edits and refresh the page.
If we plan to upload static assets to GCS buckets at build time, I'm wondering if it's feasible to upload them to a subfolder unique to each build. For instance, when we publish a release like YYYY.MM.DD-N, the static assets could be uploaded to a subfolder named YYYY.MM.DD-N. However, the dev environment should be handled differently. In the dev environment, the static assets could either be uploaded to a common subfolder or to one that represents its short commit number.
Description
Currently we use a static assets stage in our docker image build to generate static assets required for addons-server. These are copied into the final production stage and pushed along with the image. During deployments, we pull the image, extract these files and upload them to GCP bucket for use in production.
This is several round trips and a lot of image bloat for no real benefit. We build our image during the
ci.yml
when we publish a new release. Here is where we build the image and that is where we can upload the assets.Then the deployment job can access the static assets independently of the image, or even remove the need to uplaod at all if we upload directly. Need to coordinate with @bqbn on the structure of our bucket and what is possible in terms of uploading assets. Two possible options:
Acceptance Criteria
Milestones/checkpoints
Checks
┆Issue is synchronized with this Jira Task
The text was updated successfully, but these errors were encountered: