Why is parallel processing sometimes required?

• A. To ensure that processors complete in a specific order
• B. To simplify error handling
• C. Because a single flow's processors are sequential and external calls may block, so parallelism helps
• D. To avoid using concurrency controls

Answer: (C)

Parallel processing is used when a flow includes tasks that can run at the same time, especially when some steps involve external calls that might block or take a long time. If you do everything sequentially, you spend time waiting for each external operation to finish before starting the next, which slows the whole flow. By running independent tasks in parallel, you overlap waiting times and complete the overall process much faster, improving latency and throughput.

This works best when the tasks don’t depend on each other’s results or an exact order. If results are needed from a previous step, you can’t parallelize those parts without changing the flow structure. Also, parallelism introduces complexity in error handling and warrants proper synchronization; it isn’t about avoiding concurrency controls—those controls are often necessary to maintain data integrity when tasks run in parallel.

So, parallel processing is best when you have multiple independent, blocking external calls or operations, and you want to reduce total processing time by overlapping their execution.