Async batch processing in Rest API - Step-by-Step Execution

Concept Flow - Async batch processing

Receive batch request

↓

Split batch into tasks

↓

Start async processing of tasks

↓

Tasks run in parallel

↓

Collect task results

↓

Send combined response

The server receives a batch request, splits it into tasks, processes them asynchronously in parallel, collects results, and sends back a combined response.

Execution Sample

Rest API

POST /batch
Receive batch tasks
Start async tasks
Wait for all to finish
Return combined results

This code handles a batch request by processing tasks asynchronously and returning all results together.

Execution Table

Step	Action	Task ID	Status	Result	Notes
1	Receive batch request	-	Received	-	Batch with 3 tasks received
2	Split batch	-	Split	-	Tasks 1, 2, 3 identified
3	Start async task	1	Running	-	Task 1 started
4	Start async task	2	Running	-	Task 2 started
5	Start async task	3	Running	-	Task 3 started
6	Task 1 completes	1	Completed	Result A	Task 1 finished successfully
7	Task 2 completes	2	Completed	Result B	Task 2 finished successfully
8	Task 3 completes	3	Completed	Result C	Task 3 finished successfully
9	Collect results	-	Collected	Results A,B,C	All task results gathered
10	Send response	-	Sent	Results A,B,C	Combined results sent to client
11	Exit	-	-	-	All tasks done, response sent

💡 All async tasks completed and combined response sent to client

Variable Tracker

Variable	Start	After Step 2	After Step 5	After Step 8	Final
batch_tasks	[]	[1,2,3]	[1:running,2:running,3:running]	[1:done,2:done,3:done]	[1:done,2:done,3:done]
results	{}	{}	{}	{1:'Result A',2:'Result B',3:'Result C'}	{1:'Result A',2:'Result B',3:'Result C'}
response_sent	false	false	false	false	true

Key Moments - 3 Insights

Why do tasks run in parallel instead of one after another?

Because the server starts all tasks asynchronously at step 3-5, they run at the same time, speeding up total processing (see execution_table rows 3-5).

How does the server know when all tasks are done?

It waits until all tasks report completion (rows 6-8), then collects results at step 9 before sending the response.

What happens if one task takes longer than others?

The server waits for all tasks to finish before responding, so slower tasks delay the combined response (see variable_tracker showing all tasks done only at final).

Visual Quiz - 3 Questions

Look at the execution table, at which step does the server start processing task 2?

A Step 2

B Step 6

C Step 4

D Step 9

💡 Hint

Check the 'Action' and 'Task ID' columns in execution_table rows 3-5.

According to variable_tracker, what is the status of batch_tasks after step 5?

A [1:done,2:done,3:done]

B [1:running,2:running,3:running]

C [1:waiting,2:waiting,3:waiting]

D []

💡 Hint

Look at the 'batch_tasks' row under 'After Step 5' in variable_tracker.

If task 3 never completes, what step would the server never reach?

A Step 9

B Step 6

C Step 8

D Step 2

💡 Hint

See execution_table rows 6-9 for when results are collected after all tasks complete.

Concept Snapshot

Async batch processing:
- Receive batch request with multiple tasks
- Start all tasks asynchronously in parallel
- Wait for all tasks to complete
- Collect all results
- Send combined response
Key: parallel task execution speeds up processing

Full Transcript

Async batch processing means the server gets a batch of tasks and runs them all at the same time without waiting for each to finish before starting the next. The flow starts by receiving the batch, splitting it into individual tasks, and then starting each task asynchronously. Each task runs independently and finishes at its own time. The server waits until all tasks are done, collects their results, and sends back one combined response. This method is faster than running tasks one by one because it uses parallel processing. The execution table shows each step from receiving the batch to sending the response. The variable tracker shows how the list of tasks and results change over time. Key points include understanding parallel execution, waiting for all tasks to finish, and combining results before responding.