Adds a new Github Action workflow canary.yml that (currently) runs on the top of the hour, every hour, the test.yml workflow against the develop branch with the current banananation.jar model that's hardcoded in the repo.
This way, the canary e2e test run should only fail when the mission model needs to be recompiled, due to an interface change in the backend repo. This way, the UI devs can easily see if an outdated model is at fault for their e2e tests failing in unrelated PRs.
To summarize:
if the canary cron job is failing its runs, we need to update banananation.jar, and your PR changes likely aren't the cause of e2e tests failing
if the canary cron job is passing its runs, but your PR e2e tests are still failing... then your code is likely the problem
Some future work that would streamline this process even further: make a new workflow that (on failed canary run detection) automatically builds a new banananation.jar and opens a PR into the UI repo to update it.
closes https://github.com/NASA-AMMOS/aerie-ui/issues/1220
Adds a new Github Action workflow
canary.yml
that (currently) runs on the top of the hour, every hour, thetest.yml
workflow against thedevelop
branch with the currentbanananation.jar
model that's hardcoded in the repo.This way, the canary e2e test run should only fail when the mission model needs to be recompiled, due to an interface change in the backend repo. This way, the UI devs can easily see if an outdated model is at fault for their e2e tests failing in unrelated PRs.
To summarize:
banananation.jar
, and your PR changes likely aren't the cause of e2e tests failingSome future work that would streamline this process even further: make a new workflow that (on failed canary run detection) automatically builds a new
banananation.jar
and opens a PR into the UI repo to update it.