microsoft/semantic-kernel · error · VectorStoreOperationException

Sequence repetition in filter expressions exceeds the…

Error message

Sequence repetition in filter expressions exceeds the maximum allowed size.

What it means

Thrown by _safe_repeat (in_memory.py:367-370) when `seq * n` would exceed the cap: len(seq) > max_filter_sequence_repeat_size // n. Default cap is 1024. This prevents memory-exhaustion via repetition in filter expressions; the bound depends on actual runtime value lengths so it cannot be fully checked at parse time.

Solutions

Reduce the repetition count or the source sequence length.
Raise max_filter_sequence_repeat_size on the collection instance if the repetition is legitimate.
Avoid repetition in filters entirely; precompute and compare against a stored field instead.

Example fix

# before
collection.max_filter_sequence_repeat_size = 1024  # default
VectorSearchOptions(filter="lambda x: x.name * 10000 == x.target")
# after
collection.max_filter_sequence_repeat_size = 50_000
# or avoid repetition:
VectorSearchOptions(filter="lambda x: x.name == x.target[:len(x.name)]")

Defensive patterns

Strategy: validation

Try / catch

try:
    results = await collection.search(search_type=SearchType.VECTOR, options=opts)
except VectorStoreOperationException as e:
    logger.warning("filter rejected: %s", e.__cause__ or e)
    results = None

Prevention

Do not build large repeated sequences inside filters.
Tune max_filter_sequence_repeat_size deliberately, not as a workaround.

When it happens

Trigger: A filter like `lambda x: x.name * 1000` where name is a non-empty string, or `lambda x: x.tags * 500` with a sizable list, producing a result larger than 1024 elements.

Common situations: Building large padded strings/lists inside a filter; legitimate but oversized repetition needs.

Related errors

AI-assisted analysis of microsoft/semantic-kernel@c028a0c7dc (2026-08-13). Data as JSON: /api/errors/e7f4f9060620406f. Report an issue: GitHub.

Appendix: source

Thrown at python/semantic_kernel/connectors/in_memory.py:368

    def _safe_mult(self, left: Any, right: Any) -> Any:
        """Safely evaluate multiplication."""
        if isinstance(left, (int, float)) and isinstance(right, (int, float)):
            return left * right
        if isinstance(left, int) and isinstance(right, (str, list, tuple)):
            return self._safe_repeat(right, left)
        if isinstance(right, int) and isinstance(left, (str, list, tuple)):
            return self._safe_repeat(left, right)
        raise VectorStoreOperationException(
            "Multiplication in filter expressions is only allowed for numeric values and bounded sequence repetition."
        )

    def _safe_repeat(self, value: str | list[Any] | tuple[Any, ...], repeat_count: int) -> Any:
        """Safely repeat a sequence."""
        if repeat_count <= 0 or len(value) == 0:
            return value * repeat_count
        if len(value) > self._max_sequence_repeat_size // repeat_count:
            raise VectorStoreOperationException(
                "Sequence repetition in filter expressions exceeds the maximum allowed size."
            )
        return value * repeat_count

    def _safe_numeric_operation(
        self,
        operator_node: ast.AST,
        left: Any,
        right: Any,
        operation: Callable[[float | int, float | int], Any],
    ) -> Any:
        """Safely evaluate a numeric binary operation."""
        if not isinstance(left, (int, float)) or not isinstance(right, (int, float)):
            raise VectorStoreOperationException(
                f"Operator '{type(operator_node).__name__}' is only allowed for numeric values in filter expressions."
            )
        return operation(left, right)

View on GitHub (pinned to c028a0c7dc)