What causes "Unsupported type: " in apache/iceberg?

Change the write schema to a supported physical type (INT32/INT64/FLOAT/DOUBLE/FIXED_LEN_BYTE_ARRAY/BINARY as appropriate).

apache/iceberg · error · java.lang.UnsupportedOperationException

Unsupported type:

Error message

Unsupported type:

What it means

The same primitive(...) factory in FlinkParquetWriters falls back to a switch on the Parquet physical type (for columns without a handled annotation). Only the supported physical types get writers; any other physical type reaches the default branch and throws UnsupportedOperationException naming the primitive.

Source

Thrown at flink/v2.3/flink/src/main/java/org/apache/iceberg/flink/data/FlinkParquetWriters.java:213

        }
      }

      switch (primitive.getPrimitiveTypeName()) {
        case FIXED_LEN_BYTE_ARRAY:
        case BINARY:
          return byteArrays(desc);
        case BOOLEAN:
          return ParquetValueWriters.booleans(desc);
        case INT32:
          return ints(fType, desc);
        case INT64:
          return ParquetValueWriters.longs(desc);
        case FLOAT:
          return ParquetValueWriters.floats(desc);
        case DOUBLE:
          return ParquetValueWriters.doubles(desc);
        default:
          throw new UnsupportedOperationException("Unsupported type: " + primitive);
      }
    }
  }

  private static class LogicalTypeWriterBuilder
      implements LogicalTypeAnnotationVisitor<ParquetValueWriter<?>> {
    private final LogicalType flinkType;
    private final ColumnDescriptor desc;

    private LogicalTypeWriterBuilder(LogicalType flinkType, ColumnDescriptor desc) {
      this.flinkType = flinkType;
      this.desc = desc;
    }

    @Override
    public Optional<ParquetValueWriter<?>> visit(StringLogicalTypeAnnotation strings) {
      return Optional.of(strings(desc));
    }

View on GitHub (pinned to 86d9c8fc54)

Solutions

Change the write schema to a supported physical type (INT32/INT64/FLOAT/DOUBLE/FIXED_LEN_BYTE_ARRAY/BINARY as appropriate).
For timestamps, use INT64 with TIMESTAMP logical annotation instead of INT96.
Add a mapping case in the switch if a new physical type needs support, and upstream the change.

Example fix

// before
// Column: ts INT96
// after
// Column: ts INT64 (TIMESTAMP(MICROS)) via TimestampType in the Iceberg schema

Defensive patterns

Strategy: validation

Validate before calling

// java
Set<PrimitiveTypeName> ok = EnumSet.of(INT32, INT64, FLOAT, DOUBLE, BOOLEAN, BINARY, FIXED_LEN_BYTE_ARRAY);
if (!ok.contains(primitive.getPrimitiveTypeName())) {
  throw new IllegalArgumentException("Unsupported physical type: " + primitive.getPrimitiveTypeName());
}

Try / catch

// java
try {
  ParquetValueWriter<?> w = FlinkParquetWriters.createWriter(...);
} catch (UnsupportedOperationException e) {
  // reject schema or remap the column before writing
}

Prevention

Never write INT96 columns; always use INT64 with a timestamp annotation.
Derive the Parquet write schema from the Iceberg schema, not from input files.
Validate all physical types during schema planning, not at write time.

When it happens

Trigger: Creating a Parquet value writer for a column whose unannotated primitive type is not in the supported set (e.g. INT96, or another physical type the builder does not map).

Common situations: Writing into a schema derived from legacy files with INT96 timestamps; programmatic schema construction using unusual Parquet physical types; engine-written schemas with types Iceberg's Flink writer does not handle.

Understand the failure class

Background: UnsupportedOperationException and "is not supported" errors: when a library deliberately refuses a call — this error's family across 30 libraries.

Related errors

AI-assisted analysis of apache/iceberg@86d9c8fc54 (2026-09-12). Data as JSON: /api/errors/d7cfc135e6a5c5a2. Report an issue: GitHub.