feat(tests): intercept created executions through queue proxy & kill …#12804
feat(tests): intercept created executions through queue proxy & kill …#12804brian-mulier-p wants to merge 5 commits intodevelopfrom
Conversation
brian-mulier-p
force-pushed
the
fix/flaky-fix-attempt
branch
2 times, most recently
from
c599806 to
a74f4be
Compare
loicmathieu
left a comment
loicmathieu
left a comment
There was a problem hiding this comment.
Overall I would prefer we understand why things are not ended properly but as it seems overly complicated why not going into that direction...
tests/src/main/java/io/kestra/core/junit/extensions/KestraTestExtension.java
Outdated
Show resolved
Hide resolved
tests/src/main/java/io/kestra/core/junit/extensions/KestraTestExtension.java
Show resolved
Hide resolved
tests/src/main/java/io/kestra/core/junit/extensions/KestraTestExtension.java
Show resolved
Hide resolved
🐋 Docker imagedocker run --pull=always --rm -it -p 8080:8080 --user=root -v /var/run/docker.sock:/var/run/docker.sock -v /tmp:/tmp ghcr.io/kestra-io/kestra-pr:12804 server local🧪 Java Unit Tests
|
brian-mulier-p
force-pushed
the
fix/flaky-fix-attempt
branch
3 times, most recently
from
4b54afa to
16af786
Compare
|
I replied and fix all your comments + I spotted one last issue in context handling that is now fixed. |
brian-mulier-p
force-pushed
the
fix/flaky-fix-attempt
branch
from
16af786 to
32d3155
Compare
Tests report quick summary:No test report were found |
loicmathieu
left a comment
loicmathieu
left a comment
There was a problem hiding this comment.
Still not convinced by what you said about the thread local, if you're not in the test thread, you would not have access to it anyway.
I approved it, but it seems it causes more harm than good as a lot of tests are now failing
|
This kind of PR should only be merged if it proved to brings benefits: for ex when launching multiple time you see none of very few flaky tests |
brian-mulier-p
force-pushed
the
fix/flaky-fix-attempt
branch
from
32d3155 to
9900a5d
Compare
…them if running after test
Otherwise the rebuildContext was creating a new applicationContext that doesn't contain the testExecutions
brian-mulier-p
force-pushed
the
fix/flaky-fix-attempt
branch
from
9900a5d to
9148616
Compare
|
It just needs some love & polish, trust the process 👍 |
brian-mulier-p
force-pushed
the
fix/flaky-fix-attempt
branch
from
97ec7cf to
a65d9e9
Compare
|
In the end @loicmathieu I removed the need for ThreadLocal so it's easier to understand + I added a wait after sending the killing message to be 100% sure it doesn't overflow across tests 🤞 |
brian-mulier-p
force-pushed
the
fix/flaky-fix-attempt
branch
5 times, most recently
from
a6e424e to
00da357
Compare
brian-mulier-p
force-pushed
the
fix/flaky-fix-attempt
branch
2 times, most recently
from
849a872 to
7dddc3f
Compare
brian-mulier-p
force-pushed
the
fix/flaky-fix-attempt
branch
from
7dddc3f to
b3e858f
Compare
2 participants
…them if running after test