Open jduo opened 1 year ago
@jduo is there a script that we could use to reproduce this error?
I've attached a test case that shows this problem. https://github.com/jduo/arrow/blob/master/java/dataset/src/test/java/org/apache/arrow/dataset/file/TestDatasetWriterMap.java
@jduo I can reproduce this error. I will look into this.
take
Thanks @vibhatha . Just wondering do you still need the example code that's been posted?
@jduo all good, I created branch and testing further. But would you mind if I use this code as the base for testing and further investigation.
Yes you are welcome to use the example code as you see fit.
Thank you @jduo
any update on this issue? I run into the same error recently. Thanks.
Describe the bug, including details regarding any error messages, version, and platform.
Trying to write a map column to a parquet file using DatasetFileWriter results in the following error:
Runtime java.lang.IllegalArgumentException: not all nodes and buffers were consumed. nodes: [ArrowFieldNode [length=2, nullCount=1], ArrowFieldNode [length=0, nullCount=0], ArrowFieldNode [length=0, nullCount=0]] buffers: [ArrowBuf[509], address:140461672595576, capacity:0, ArrowBuf[510], address:140461672595576, capacity:1, ArrowBuf[511], address:140461672595584, capacity:8, ArrowBuf[512], address:140461672595592, capacity:0, ArrowBuf[513], address:140461672595592, capacity:0, ArrowBuf[514], address:140461672595592, capacity:0] at org.apache.arrow.vector.VectorLoader.load(VectorLoader.java:89) at org.apache.arrow.vector.ipc.ArrowReader.loadRecordBatch(ArrowReader.java:220) at org.apache.arrow.vector.ipc.ArrowStreamReader.loadNextBatch(ArrowStreamReader.java:161) at org.apache.arrow.c.ArrayStreamExporter$ExportedArrayStreamPrivateData.getNext(ArrayStreamExporter.java:72)
This happens when the data-to-be-written comes from an ArrowFileReader or ArrowStreamReader. This problem may affect more file formats than parquet (and might actually be related to the DatasetFileWriter in general).
Component(s)
Java, Parquet