Conversation
|
dont merge yet, we messed up image pulling |
|
This seems kinda roundabout to me. You are basically saying find me a worker and create a container, then you expose back the worker just so that we can look up the container all over again. Why don't we just add a It will probably also help to separate the \cc @ddadlani |
|
@pivotal-jwinters I think the goal is to get to what you were saying. I agree its kind of weird right now. But this wasn't meant to be that big refactor but rather way to get things correct again and ease some pain points for users. This gives them something while we can work on longer term solutions after. |
|
@wagdav FYI |
|
@wagdav this change exposed a larger issue where checks error every now and then. Another recent change in how check sessions are handled also makes it difficult to patch this issue. We are still working on a fix for that error, after which we will merge this into master |
Co-authored-by: Krishna Mannem <[email protected]>
|
@wagdav we're gunna workaround the issue, this should be going in considering the tests pass. |
Signed-off-by: Divya Dadlani <[email protected]> Co-authored-by: Krishna Mannem <[email protected]>
|
@ddadlani fixed the tests, we accidentally commit changes to the docker-compose. its ready for review now. |
Signed-off-by: Divya Dadlani <[email protected]> Co-authored-by: Krishna Mannem <[email protected]>
| - -c | ||
| - | | ||
| exit 2 | ||
| - in_parallel: |
There was a problem hiding this comment.
@ddadlani instead of making a new topgun test cause its heavy, I decided to just update the existing one to now assert that in_parallel tasks get distributed evenly.
| - -c | ||
| - | | ||
| exit 2 | ||
| - in_parallel: |
|
fixes #3301 |
|
@aledeganopix4d yes it should have. I think I had to comment that before I merged this PR, though. Thanks for pointing it out! |
Signed-off-by: Divya Dadlani <[email protected]>
manually revert changes from PR #3902
Signed-off-by: Denise Yu [email protected]
Co-authored-by: Krishna Mannem [email protected]