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
While #39 implemented support for repository_dispatch events, I don't see yet how to use it when someone makes a PR from his fork. I understand the concept of using /ok-to-test and dispatching via repository_dispatch.
However, when I print the variables in the okteto/deploy-preview job:
An ok-to-deploy action is triggered which triggers an okteto-preview-deploy action (via repository_dispatch)
The bash script in okteto/deploy-preview incorrectly determines that master branch should be deployed based on GITHUB_REF based on this line
A preview deploy is made using code from the latest master branch, not from the PR code changes
Any good suggestion to tackle this? I can make a PR, however I suppose we need to pass our custom own env var with the branch BRANCH since we don't have the PR number in the default env vars (it's dispatched)?
While #39 implemented support for
repository_dispatch
events, I don't see yet how to use it when someone makes a PR from his fork. I understand the concept of using/ok-to-test
and dispatching viarepository_dispatch
.However, when I print the variables in the
okteto/deploy-preview
job:Which contains master instead of the PR branch! And
okteto/deploy-preview
is figuring out thebranch
var to pass to the okteto deployer: https://github.com/okteto/deploy-preview/blob/main/entrypoint.sh#L34-L38 so it ends up passing the master branch to the okteto deployer.So in the end this happens:
/ok-to-deploy sha=XXX
on the PRrepository_dispatch
)okteto/deploy-preview
incorrectly determines that master branch should be deployed based onGITHUB_REF
based on this lineAny good suggestion to tackle this? I can make a PR, however I suppose we need to pass our custom own env var with the branch
BRANCH
since we don't have the PR number in the default env vars (it's dispatched)?Did you also observe this @ikuradon ?
The text was updated successfully, but these errors were encountered: