holdenk / holdenk/spark-testing-base
Rows only in test missing after switch to Kryo
- Dominant language
- Scala
- Stars
- 1.6k
- Forks
- 357
- PR merge metrics
- No merged PRs in 30d
Description
I'm using scala/sbt and recently while making the switch to kryo one of my tests start failing on a row count comparison. I can confirm when looking at the output dataframes that they are indeed missing several rows. The reason I think this could be a spark-testing-base issue is I can't replicate the issue in notebooks, locally, or in dev-env. All those places work fine. Its only when running the tests that it fails.
Unfortunately, there is no errors or anything I can find to help you guys out here. I've checked the output in several places and have found nothing out of the place. If you have any ideas or output you think can help solve things feel free to ask and I will provide it.
Thanks in advance.
Contributor guide
Research direction
Start by reproducing the failing row-count comparison with Scala/sbt after switching to Kryo, then inspect the output dataframes and compare the test run with notebooks, local runs, and the dev environment. The issue is done when the missing-row cause is identified and the affected test passes reliably.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- scala, spark
- Domain
- testing-qa
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100