prestodb/presto · error · PrestoException

COLUMN_NOT_FOUND

COLUMN_NOT_FOUND

Error message

Column not found:

What it means

After building the SortOrder, parseSortFields verifies every sort field's sourceId exists among the schema's top-level column field IDs. A field referencing a column absent from the table schema throws PrestoException(COLUMN_NOT_FOUND, "Column not found: " + findColumnName(sourceId)). This prevents sort orders referencing non-existent columns.

Source

Thrown at presto-iceberg/src/main/java/com/facebook/presto/iceberg/SortFieldUtils.java:68

    public static SortOrder parseSortFields(Schema schema, List<String> fields)
    {
        SortOrder.Builder builder = SortOrder.builderFor(schema);
        parseSortFields(builder, fields);
        SortOrder sortOrder;
        try {
            sortOrder = builder.build();
        }
        catch (RuntimeException e) {
            throw new PrestoException(INVALID_TABLE_PROPERTY, "Invalid " + SORTED_BY_PROPERTY + " definition", e);
        }

        Set<Integer> baseColumnFieldIds = schema.columns().stream()
                .map(Types.NestedField::fieldId)
                .collect(toImmutableSet());
        for (SortField field : sortOrder.fields()) {
            if (!baseColumnFieldIds.contains(field.sourceId())) {
                throw new PrestoException(COLUMN_NOT_FOUND, "Column not found: " + schema.findColumnName(field.sourceId()));
            }
        }

        return sortOrder;
    }

    public static void parseSortFields(SortOrderBuilder<?> sortOrderBuilder, List<String> fields)
    {
        fields.forEach(field -> parseSortField(sortOrderBuilder, field));
    }

    private static void parseSortField(SortOrderBuilder<?> builder, String field)
    {
        Matcher matcher = PATTERN.matcher(field);
        if (!matcher.matches()) {
            throw new IllegalArgumentException(format("Unable to parse sort field: [%s]", field));
        }

View on GitHub (pinned to 55bb57d202)

Solutions

  1. Compare the sorted_by column names against the actual CREATE TABLE column list and fix any misspelling
  2. Use only top-level columns — nested or struct-internal fields cannot be sort sources here
  3. Run SHOW COLUMNS / DESCRIBE on the table to confirm the exact column names before re-running DDL
  4. Ensure the sort field identifier doesn't include stray whitespace or case mismatches (depending on case-sensitivity settings)

Example fix

// before
CREATE TABLE t (id BIGINT, data VARCHAR) WITH (sorted_by = ARRAY['ts ASC'])
// after
CREATE TABLE t (id BIGINT, data VARCHAR, ts TIMESTAMP) WITH (sorted_by = ARRAY['ts ASC'])
Defensive patterns

Strategy: validation

Validate before calling

// Before DDL, confirm every sorted_by column is a top-level column of the table
Set<String> columns = Set.of("id", "data", "ts"); // actual CREATE TABLE columns
for (String f : sortedBy) {
    String col = f.trim().split("\\s+")[0];
    if (!columns.contains(col)) throw new IllegalArgumentException("Column not found: " + col);
}

Try / catch

try { /* CREATE TABLE with sorted_by */ } catch (PrestoException e) {
    if (e.getErrorCode().getName().equals("COLUMN_NOT_FOUND")) {
        // message names the missing column; fix DDL and retry
    } else { throw e; }
}

Prevention

When it happens

Trigger: CREATE TABLE ... WITH (sorted_by = ARRAY['colname']) where 'colname' is not a top-level column of the created table schema — misspelled name, a nested field path, or a column defined only in a later schema.

Common situations: Typos in DDL; referencing nested struct fields (only top-level columns are supported for sorting); copying sorted_by from another table with different columns; referencing a column renamed before table creation.

Related errors

AI-assisted analysis of prestodb/presto@55bb57d202 (2026-09-04). Data as JSON: /api/errors/2d0c185378b5ec3a. Report an issue: GitHub.