What causes "Columnar reads are not supported" in apache/iceberg?

Use a factory that supports columnar reads (e.g. the SparkBatchQueryReader/SparkScan columnar path) for columnar output

apache/iceberg · error · UnsupportedOperationException

Columnar reads are not supported

Error message

Columnar reads are not supported

What it means

SparkRowReaderFactory never supports columnar reads; createColumnarReader unconditionally throws UnsupportedOperationException and supportColumnarReads returns false. The row reader factory exists only for row-based (Record) consumption such as changelog and position-delete reads.

Source

Thrown at spark/v4.0/spark/src/main/java/org/apache/iceberg/spark/source/SparkRowReaderFactory.java:61

    if (partition.allTasksOfType(FileScanTask.class)) {
      return new RowDataReader(partition);

    } else if (partition.allTasksOfType(ChangelogScanTask.class)) {
      return new ChangelogRowReader(partition);

    } else if (partition.allTasksOfType(PositionDeletesScanTask.class)) {
      return new PositionDeletesRowReader(partition);

    } else {
      throw new UnsupportedOperationException(
          "Unsupported task group for row-based reads: " + partition.taskGroup());
    }
  }

  @Override
  public PartitionReader<ColumnarBatch> createColumnarReader(InputPartition inputPartition) {
    throw new UnsupportedOperationException("Columnar reads are not supported");
  }

  @Override
  public boolean supportColumnarReads(InputPartition inputPartition) {
    return false;
  }
}

View on GitHub (pinned to 86d9c8fc54)

Solutions

Use a factory that supports columnar reads (e.g. the SparkBatchQueryReader/SparkScan columnar path) for columnar output
Ensure the scan reports supportColumnarReads=false so Spark does not request columnar readers from this factory
Wrap results as rows instead of ColumnarBatch if row-only readers are required

Example fix

// before
return new SparkRowReaderFactory(); // used for columnar reads
// after
return supportsColumnar ? new SparkColumnarReaderFactory() : new SparkRowReaderFactory();

Defensive patterns

Strategy: validation

Validate before calling

if (readerFactory instanceof SparkRowReaderFactory && wantColumnar) {
  throw new IllegalArgumentException("Row reader factory cannot serve columnar reads");
}

Prevention

Always return false from supportColumnarReads when advertising the row reader factory
Select the reader factory based on required output type (rows vs ColumnarBatch)
Never mix reader factories across scan implementations

When it happens

Trigger: Any call to SparkRowReaderFactory.createColumnarReader, e.g. Spark requesting a columnar batch reader because supportColumnarReads was consulted for a different factory.

Common situations: Custom SparkScan wiring the row reader factory where Spark demands columnar batches; misconfigured batch scan implementations in custom sources built on Iceberg readers.

Related errors

AI-assisted analysis of apache/iceberg@86d9c8fc54 (2026-09-12). Data as JSON: /api/errors/f713b2cc4151a290. Report an issue: GitHub.