apache/druid · error · UnsupportedColumnTypeException

Cannot handle column [%s] with type [%s]

Error message

Cannot handle column [%s] with type [%s]

What it means

FieldReaders.create dispatches on a column's ColumnType to build the matching frame field reader; array subtypes fall through to a default arm that throws UnsupportedColumnTypeException (message 'Cannot handle column [%s] with type [%s]'). This error [986] corresponds to the array-switch default: an array column of a leaf type the frame format cannot represent.

Source

Thrown at processing/src/main/java/org/apache/druid/frame/field/FieldReaders.java:78

      case COMPLEX:
        return ComplexFieldReader.createFromType(columnType);

      case ARRAY:
        switch (Preconditions.checkNotNull(columnType.getElementType().getType(), "array elementType")) {
          case STRING:
            return new StringArrayFieldReader();

          case LONG:
            return new LongArrayFieldReader();

          case FLOAT:
            return new FloatArrayFieldReader(frameType);

          case DOUBLE:
            return new DoubleArrayFieldReader(frameType);

          default:
            throw new UnsupportedColumnTypeException(columnName, columnType);

        }
        // Fall through to error for other array types

      default:
        throw new UnsupportedColumnTypeException(columnName, columnType);
    }
  }
}

View on GitHub (pinned to 9b90983fd2)

Solutions

1. Check the actual column type in the error message; avoid materializing that type into frames (cast or flatten the column first).
2. Use only supported array element types (LONG/FLOAT/DOUBLE arrays) in queries routed through the frame processor.
3. If string/complex arrays are needed, keep them out of MSQ frame shuffles or convert to a supported representation.
4. Upgrade Druid if a newer version added reader support for that array type.

Example fix

// before
SELECT arr_col FROM ...  -- arr_col is VARCHAR ARRAY, fails in frame creation
// after
SELECT ARRAY_TO_STRING(arr_col, ',') AS arr_col FROM ...  -- or ARRAY_LONG cast

if (columnType != null && columnType.getType() == ValueType.ARRAY
    && !EnumSet.of(ValueType.LONG, ValueType.FLOAT, ValueType.DOUBLE).contains(columnType.getElementType().getType())) {
  throw new IllegalStateException("array element type not frame-supported: " + columnType);
}

try { FieldReaders.create(name, t); } catch (UnsupportedColumnTypeException e) { /* coerce/cast column or fall back to non-frame path */ }

When it happens

Trigger: Calling FieldReaders.create(columnName, columnType) for an ARRAY column whose element type is unsupported by the frame readers (the switch handles LONG/FLOAT/DOUBLE arrays; other array element types hit the default).

Common situations: Attempting to write/query frames containing exotic array types (e.g. string arrays or complex arrays) that the frame processor does not support; ingest producing array columns from nested data and then materializing them as frames (MSQ shuffle).

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

• Cannot handle column [%s] with type [%s]
• Cannot handle column [%s] with unknown type
• Unsupported column type[%s]
• Aggregation [%s] does not support column [%s] of type [%s].
• endFrame[%,d] > numFrames[%,d]

AI-assisted analysis of apache/druid@9b90983fd2 (2026-09-07). Data as JSON: /api/errors/95a399eabbd91dd9. Report an issue: GitHub.