Why Promise.allSettled Beats Promise.all for Robust Async Handling

In the world of JavaScript asynchronous programming, Promise.all has long been the primary tool for concurrency control, but it has a clear limitation: if any task fails, the entire process stops and the results of other tasks cannot be obtained. A more powerful API is now available for developers to handle concurrent tasks.

Advantages of Promise.allSettled

Promise.allSettled is a concurrency‑control API provided by JavaScript that lets you wait for all promises to complete (whether fulfilled or rejected) and obtain each promise's final status and result. Unlike Promise.all, Promise.allSettled does not short‑circuit on a single rejection; it continues waiting for all promises to finish.

Basic usage

Comparison with Promise.all

Promise.all rejects as soon as any promise is rejected, which may be desirable in some scenarios, but in many cases we want to obtain the results of all tasks even if some fail.

Practical application scenarios

Batch data processing

When you need to process a large number of independent data items and each processing step may fail, Promise.allSettled is the ideal choice.

API request aggregation

When you need to send requests to multiple endpoints and aggregate all results, Promise.allSettled ensures you obtain all available data even if some requests fail.

Concurrent task management

When you have multiple independent tasks that need to run in parallel and the success or failure of one should not affect the others, Promise.allSettled is a safer choice.

Other concurrency control APIs

Besides Promise.allSettled, JavaScript also provides other concurrency control APIs:

Promise.all : use when all promises must succeed.

Promise.race : use when you need the result of the fastest promise.

Promise.any : use when you need the first fulfilled promise.

Promise.allSettled gives JavaScript developers more powerful concurrency control. When handling multiple asynchronous tasks, it ensures you can retrieve the final state of every task, regardless of success or failure, making error handling and result processing more flexible and reliable.