diff --git a/query-engine/core/src/interpreter/query_interpreters/nested_read.rs b/query-engine/core/src/interpreter/query_interpreters/nested_read.rs index 7e47b7f545a4..8573e2853c45 100644 --- a/query-engine/core/src/interpreter/query_interpreters/nested_read.rs +++ b/query-engine/core/src/interpreter/query_interpreters/nested_read.rs @@ -190,7 +190,6 @@ pub async fn one2m( // If we're fetching related records from a single parent, then we can apply normal pagination instead of in-memory processing. // However, we can't just apply a LIMIT/OFFSET for multiple parents as we need N related records PER parent. // We could use ROW_NUMBER() but it requires further refactoring so we're still using in-memory processing for now. - let processor = if uniq_selections.len() == 1 && !query_args.requires_inmemory_processing() { None } else {