How does Parallel Stream work internally? Is it always better?

How It Works Internally

Source Splitting: The stream's Spliterator divides the data source into chunks.
ForkJoinPool: Each chunk is submitted as a task to ForkJoinPool.commonPool().
Parallel Execution: Worker threads process chunks concurrently.
Combining Results: Results from all chunks are merged back (e.g., via Collectors).

[1, 2, 3, 4, 5, 6, 7, 8]
        ↓ Spliterator splits
  [1,2,3,4]    [5,6,7,8]
     ↓ split      ↓ split
 [1,2] [3,4]  [5,6] [7,8]
   ↓     ↓      ↓     ↓    ← 4 worker threads
  f(1,2) f(3,4) f(5,6) f(7,8)
     ↓     ↓      ↓     ↓
      combine    combine
          ↓
       final result

The Common Pool Problem

// ALL parallel streams in your app share the same pool!
ForkJoinPool.commonPool(); // size = availableProcessors() - 1

// A slow stream blocks others:
list1.parallelStream().map(x -> slowIOCall(x));  // ⚠️ Hogs the pool
list2.parallelStream().map(x -> fastCompute(x)); // ⚠️ Starved

Workaround — Use a custom pool:

ForkJoinPool customPool = new ForkJoinPool(4);
customPool.submit(() ->
    list.parallelStream().forEach(this::process)
).get();

Is It Always Better? No!

Scenario	Sequential	Parallel	Winner
Small dataset (< 10K)	Fast	Overhead of splitting/merging	Sequential
Simple operations (identity map)	Fast	Splitting overhead > gain	Sequential
I/O-bound (REST/DB calls)	Blocking	Blocks ForkJoinPool threads	Sequential (use async instead)
Large dataset + CPU-heavy ops	Slow	Massive speedup	Parallel ✅
LinkedList source	N/A	Poor splittability	Sequential
ArrayList / Array source	N/A	Excellent splittability	Parallel ✅

Data Source Splittability

Source	Splittability	Good for Parallel?
`ArrayList`	Excellent	✅ Yes
`int[]`, `long[]`	Excellent	✅ Yes
`HashSet`	Good	✅ Yes
`TreeSet`	Good	✅ Yes
`LinkedList`	Poor	❌ No
`Stream.iterate()`	Poor	❌ No

Golden Rule

Use parallel streams only when: the dataset is large (100K+), operations are CPU-intensive, the source splits well (arrays/ArrayLists), and operations are stateless & thread-safe.