[VL] Pandas UDFs fail with "Cannot retain a released vector" on partitions with more than one batch
Maintainers usually reply within 3 days
Assessment
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Newbie friendliness
- 22/100
- Issue type
- Bug
- Clarity
- Mostly clear
- Activity status
- Active
- Domain
- data-engineering
Research direction
The failure is in ColumnarArrowEvalPythonExec.scala, where adjustRefCnt retains vectors that RefCountedArrowColumnVector.java has already released after the first joined batch. Read how the Python runner's reader reuses the Arrow stream root's vectors, then reproduce with a pandas UDF over a 10,000-row Parquet partition. Done means the second batch no longer hits the released-vector error; PR #13707 is already open against this, so check it first.
Written by the indexing model from the issue text.
Description
Since #13280, a scalar pandas UDF that runs in ColumnarArrowEvalPythonExec fails when a partition sends more than one batch to Python, which happens once a partition has more than spark.gluten.sql.columnar.maxBatchSize rows (4096 by default):
java.lang.IllegalStateException: Cannot retain a released vector
at org.apache.gluten.vectorized.RefCountedArrowColumnVector.retain(RefCountedArrowColumnVector.java:49)
at org.apache.spark.api.python.ColumnarArrowEvalPythonExec.adjustRefCnt(ColumnarArrowEvalPythonExec.scala:519)
The Python runner's reader returns every output batch over the same RefCountedArrowColumnVectors, which wrap the vectors of the Arrow stream reader's root. Closing the first joined batch releases them, and the second batch then retains a released vector.
It reproduces on v1.8.0-rc1 with the default configuration: a pandas UDF over a single 10,000-row Parquet partition fails.
This issue was written with the assistance of AI (Claude Opus).
- Dominant language
- Scala
- Stars
- 1.6k
- Forks
- 679
- Avg merge
- 4d 16h
- Merged PRs (30d)
- 84
Getting set up
Starts the project's dev container in your browser, under your own GitHub account.
- No Dockerfile or Docker Compose file
- Has a pull request template
- Read the contributing guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from apache/gluten
-
[CORE] Selecting a cost model the backend has no coster for fails with ClassNotFoundExceptionPossibly taken @LuciferYang claimed this 1 day ago. Open
Difficulty 2/5 1-3 hours Newbie friendliness 66/100
Maintainers usually reply within 3 days
-
bug unit testing velox backend
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
apache/gluten#13463 · 1 comment ·
Maintainers usually reply within 3 days
-
Difficulty 1/5 Under an hour Newbie friendliness 82/100
Maintainers usually reply within 3 days
-
[VL] madvise(WILLNEED) call fails in MmapFileStream because of wrong calculation of fetching lengthMay be free again A pull request for this issue was closed without being merged. Openbug triage
Difficulty 2/5 1-3 hours Newbie friendliness 82/100
Maintainers usually reply within 3 days
-
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
Maintainers usually reply within 3 days
Similar issues
-
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
rchillyard/TableParser#206 ·
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
Maintainers usually reply within 2 days
-
? - Needs Triage bug
Difficulty 2/5 1-3 hours Newbie friendliness 66/100
NVIDIA/cudf-spark#17063 ·
Maintainers usually reply within 2 days
-
bug 🐞 User issue 🚨
Difficulty 2/5 1-3 hours Newbie friendliness 76/100
valkey-io/valkey-glide#7813 · 2 comments ·
Maintainers usually reply within 1 day