apache/flink · error · IllegalArgumentException

The record must not be null.

Error message

The record must not be null.

What it means

FloatPrimitiveArraySerializer writes the array length then each float. serialize() rejects null because a null array has no length to emit and primitive-array type information is non-nullable by contract.

Source

Thrown at flink-core/src/main/java/org/apache/flink/api/common/typeutils/base/array/FloatPrimitiveArraySerializer.java:71

        float[] copy = new float[from.length];
        System.arraycopy(from, 0, copy, 0, from.length);
        return copy;
    }

    @Override
    public float[] copy(float[] from, float[] reuse) {
        return copy(from);
    }

    @Override
    public int getLength() {
        return -1;
    }

    @Override
    public void serialize(float[] record, DataOutputView target) throws IOException {
        if (record == null) {
            throw new IllegalArgumentException("The record must not be null.");
        }

        final int len = record.length;
        target.writeInt(len);
        for (int i = 0; i < len; i++) {
            target.writeFloat(record[i]);
        }
    }

    @Override
    public float[] deserialize(DataInputView source) throws IOException {
        final int len = source.readInt();
        float[] result = new float[len];

        for (int i = 0; i < len; i++) {
            result[i] = source.readFloat();
        }

View on GitHub (pinned to 2f3c205e92)

Solutions

  1. Initialize float[] fields to an empty array instead of null before the sink/state.
  2. Filter out records with null arrays upstream.
  3. If nullability is legitimate, switch to an object/nullable type (e.g., Float[] or a Row with null handling).
  4. Add a null check in your mapper to default null arrays to empty.

Example fix

// before: out.values may be null -> serializer.serialize(values) throws
// after: data.map(r -> { if (r.values == null) r.values = new float[0]; return r; })
Defensive patterns

Strategy: validation

Validate before calling

// Validate before serialize
float[] safe = record == null ? new float[0] : record;
serializer.serialize(safe, target);

Try / catch

try {
    serializer.serialize(record, target);
} catch (IllegalArgumentException e) {
    if (e.getMessage().equals("The record must not be null.")) {
        record = new float[0];
        serializer.serialize(record, target);
    } else throw e;
}

Prevention

When it happens

Trigger: Calling serialize(null, target) on the float[] serializer — a field typed float[] that resolved to null at runtime.

Common situations: A numeric/vector field typed float[] that is null due to a null source value; a UDF returning null where a float array is expected; a nullable SQL column mapped to a primitive float[].

Related errors


AI-assisted analysis of apache/flink@2f3c205e92 (2026-08-14). Data as JSON: /api/errors/13dcd69ea96641aa. Report an issue: GitHub.