Construct task instances with ExecuteTask.make for celery executor integration tests#61311
Conversation
|
Could you please check whether we need this PR, or if the tests are already resolved? |
|
I think the current PR title might be confusing. Maybe we can update it to focus on the flaky test |
|
We generally use that title for upgrading dependencies from prek, uv on our CI. This looks like indeed related to flaky test since we haven't seen a couple and popped up again |
|
I misunderstood the rule about updating the PR title. I thought it meant this, but I see now that’s not right. |
There was a problem hiding this comment.
Even though it is in celery, what it generates as model is purely related to TaskSDK and Execution API communication, so I would also like to take @amoghrajesh on this one
@Prab-27, In the short term, as a CI team, we can increase flakiness from 3 to 5 and add a delay between runs with a delay parameter, so we can be more flexible until we have the solution. Please let me know if you can update, or I can also do it
|
@bugraoz93 Okay I'm trying to update it and get back to you |
ceb7e59 to
b908b59
Compare
|
I think your other PR should give us some more time until we are bothered more :) No rush, just wanted to clarify here too |
|
Another tests are failing , I'm updating Taskinstance class Edited : It's wrong thoughts |
|
Oops !! Tests are failing due to inmake() classmethod
Error Log : @bugraoz93 Does this fine to use |
amoghrajesh
left a comment
There was a problem hiding this comment.
Can you follow pattern similar to test_base_executor to fix the CI on this PR: https://github.com/apache/airflow/blob/main/airflow-core/tests/unit/executors/test_base_executor.py#L226-L239
b908b59 to
fa7b59a
Compare
|
Let's see ! I don't understand this - strange behavior. |
Not always — sometimes concurrency issues, network instability, or infrastructure hiccups can cause the test to be flaky. |
amoghrajesh
left a comment
There was a problem hiding this comment.
Looks good, preemptive approval assuming nits are handled.
fa7b59a to
fae96ae
Compare
ExecuteTask.make for celery executor integration tests
…integration tests (apache#61311)
…integration tests (apache#61311)
…integration tests (apache#61311)
…integration tests (apache#61311)
…integration tests (apache#61311)
…integration tests (apache#61311)
…integration tests (apache#61311)
This PR adds the required fields using
make() classmethodto the PydanticExecuteTaskclass to ensure proper validationWas generative AI tooling used to co-author this PR?
<
If generative AI tooling has been used in the process of authoring this PR, please
change below checkbox to
[X]followed by the name of the tool, uncomment the "Generated-by".-->
{pr_number}.significant.rstor{issue_number}.significant.rst, in airflow-core/newsfragments.