fix: handling infinite loop #812
Conversation
|
@harshilgajera-crest can you switch base to develop, so we first get beta release and run some testing in CI? |
mkolasinski-splunk
requested review from
artemrys,
mbruzda-splunk and
kdoroszko-splunk
| echo $SPLUNK_VERSION | ||
| docker compose -f "docker-compose-ci.yml" build | ||
| SPLUNK_PASSWORD=Chang3d! docker compose -f docker-compose-ci.yml up -d splunk |
There was a problem hiding this comment.
We are not actually removing this,
This line SPLUNK_PASSWORD=Chang3d! docker compose -f docker-compose-ci.yml up --abort-on-container-exit
would do the same thing, it would start all the services, earlier we did in individually which seems to be not required.
|
|
||
| # Here we are not interested in the failures or errors, | ||
| # we are basically checking that we get results and test execution does not get stuck | ||
| assert result.parseoutcomes().get("passed") > 0 |
There was a problem hiding this comment.
Is there a way to verify that the exception is raised by one of the workers?
There was a problem hiding this comment.
I don't think so, because that is handled internally via xdist. But everything would run on either of the workers.
There was a problem hiding this comment.
What about checking the logs if those are available somewhere?
My point is - if you don't make a fix - will this test pass anyways?
There was a problem hiding this comment.
@artemrys
Okay, So I actually tested it on main branch without making the fix and test kept on running.
Let me make the same in github actions, I will provide a link here for the actions run which should go on for 6hrs.
Here is the link with test change in main branch without fix: https://github.com/splunk/pytest-splunk-addon/actions/runs/8753677154/job/24023897918
This will go on forever till runner timeout as it doesnot have fix.
| export SPLUNK_APP_PACKAGE=./tests/e2e/addons/TA_fiction | ||
| export SPLUNK_ADDON=TA_fiction | ||
| export SPLUNK_APP_ID=TA_fiction | ||
| export SPLUNK_APP_PACKAGE=./tests/e2e/addons/TA_fiction_indextime |
There was a problem hiding this comment.
@harshilgajera-crest
What is the reason of this change?
There was a problem hiding this comment.
Actually we were installing TA_fiction addon because our splunk docker file requires SPLUNK_APP_PACKAGE param. It was not necessary for the test as it just searched a query in _internal, but for the test that I added it requires this TA_fiction_indextime addon as we are ingesting its data(sc4s data).
Hence I updated to TA_fiction_indextime instead of TA_fiction.
|
🎉 This PR is included in version 5.2.6-beta.1 🎉 The release is available on GitHub release Your semantic-release bot 📦🚀 |
|
🎉 This PR is included in version 5.2.6 🎉 The release is available on GitHub release Your semantic-release bot 📦🚀 |
There was a bug in PSA where if in splunk_ingest_data first worker errors out or throws exception then it does not write to a file, now other workers would be waiting indefinitely for the file to be present, So the execution would continue in infinitly.
To handle this scenario, we can handle the exception thrown by the first worker and make sure that in any case it writes to a file, so that other workers don't wait indefinitly