Parallel Pipeline Bottleneck¶

Rule details


Rule ID	`parallel-pipeline-bottleneck`
Severity	WARNING — likely performance problem
Confidence	MEDIUM — likely correct, some context-dependent
Category	Concurrency
Complexity	O(n) contended → O(n) lock-free

Description¶

Detects parallelStream().forEach() where the callback mutates shared state. Mutating a shared collection from parallel threads negates the benefits of parallelism and risks data corruption — ArrayList, HashMap, and most standard collections are not thread-safe.

Typical¶

List<String> results = new ArrayList<>();
items.parallelStream().forEach(item ->
    results.add(item.toUpperCase())  // shared mutable state!
);

Multiple threads call results.add() concurrently on a non-thread-safe ArrayList. This can silently lose elements, throw ConcurrentModificationException, or corrupt internal state.

After¶

List<String> results = items.parallelStream()
    .map(item -> item.toUpperCase())
    .collect(Collectors.toList());

collect() handles thread-safe accumulation internally. No shared mutable state needed.

What it catches¶

parallelStream().forEach() with mutation calls (add, put, remove, addAll, clear, etc.) on variables defined outside the lambda
Mutations on the iterated element itself are fine and are not flagged

Suggestion¶

Use .collect() or .reduce() instead of mutating shared state in .forEach()