apache/iceberg · error · UnsupportedOperationException

Unsupported task group for columnar reads: ${partition.taskG

Error message

Unsupported task group for columnar reads: ${partition.taskGroup()}

What it means

createColumnarReader requires that every task in the input partition is a plain FileScanTask. If the task group contains other task types (e.g. combined delete/equality-delete tasks not suited to columnar reads), columnar reading cannot proceed and the factory fails fast with the task group description.

Source

Thrown at spark/v3.5/spark/src/main/java/org/apache/iceberg/spark/source/SparkColumnarReaderFactory.java:62

  @Override
  public PartitionReader<InternalRow> createReader(InputPartition inputPartition) {
    throw new UnsupportedOperationException("Row-based reads are not supported");
  }

  @Override
  public PartitionReader<ColumnarBatch> createColumnarReader(InputPartition inputPartition) {
    Preconditions.checkArgument(
        inputPartition instanceof SparkInputPartition,
        "Unknown input partition type: %s",
        inputPartition.getClass().getName());

    SparkInputPartition partition = (SparkInputPartition) inputPartition;

    if (partition.allTasksOfType(FileScanTask.class)) {
      return new BatchDataReader(partition, parquetConf, orcConf);
    } else {
      throw new UnsupportedOperationException(
          "Unsupported task group for columnar reads: " + partition.taskGroup());
    }
  }

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

View on GitHub (pinned to 86d9c8fc54)

Solutions

1. Fall back to row-based reads for these partitions or disable vectorized reads (spark.sql.iceberg.vectorized-reader.enabled=false).
2. Rewrite the table (e.g. rewrite_data_files / rewrite deletes) to remove equality deletes so all tasks are FileScanTasks.
3. Check the task group in the message to identify which scan produced the unsupported tasks.

Example fix

// before
spark.conf.set("spark.sql.iceberg.vectorized-reader.enabled", "true") // fails on equality deletes
// after
spark.conf.set("spark.sql.iceberg.vectorized-reader.enabled", "false")

// check for equality deletes before columnar read
boolean hasEqDeletes = table.currentSnapshot() != null &&
    !table.currentSnapshot().dataManifests(table.io()).stream()
        .flatMap(m -> safe(m::deleteFilesCount).stream()) ... ; // or simply check spec/format

if (!partition.allTasksOfType(FileScanTask.class)) { /* fall back to row reader */ }

try { reader = factory.createColumnarReader(p); } catch (UnsupportedOperationException e) { reader = rowReaderFor(p); }

When it happens

Trigger: Calling createColumnarReader on a SparkInputPartition whose taskGroup contains non-FileScanTask tasks, i.e. partition.allTasksOfType(FileScanTask.class) is false.

Common situations: Tables with equality deletes or mixed task types where Spark requests columnar reads (vectorized reader enabled); also custom scans producing non-file tasks.

Understand the failure class

Background: "Must be a positive integer", "Invalid value", "Unsupported": the invalid-argument-value error family, when a library rejects the value you pass — this error's family across 35 libraries.

Related errors

• Row-based reads are not supported
• Columnar reads are not supported
• Columnar reads are not supported
• Columnar reads are not supported
• ${getClass().getName()} does not support forEach

AI-assisted analysis of apache/iceberg@86d9c8fc54 (2026-09-12). Data as JSON: /api/errors/c0fa55937d1764d5. Report an issue: GitHub.