fix: add max_items limit to fetch_all_edges to prevent unbounded memory growth

[JiayuuWang]

Bug

The fetch_all_nodes function has a max_items parameter (default 2000) to prevent fetching too many items, but fetch_all_edges was missing this safeguard. This could cause memory issues with graphs that have a large number of edges.

Proposed Fix

Add _MAX_EDGES = 5000 as the default limit for edges, matching the pattern already established for nodes.

File

• backend/app/utils/zep_paging.py

Code Change

# Add constant
_MAX_EDGES = 5000

# Update function signature
def fetch_all_edges(
    client: Zep,
    graph_id: str,
    page_size: int = _DEFAULT_PAGE_SIZE,
    max_items: int = _MAX_EDGES,  # Add this parameter
    ...
):
    # Add check in the loop
    if len(all_edges) >= max_items:
        all_edges = all_edges[:max_items]
        logger.warning(f"Edge count reached limit ({max_items}), stopping pagination...")
        break