|
<< Click to Display Table of Contents >> Navigation: Optimizer API > Choosing the Optimization Method |
The API and service provides two options for optimization, asynchronous and synchronous. It may become an important design decision as to which method to use. Some of the advantages and disadvantages are described here:
Asynchronuos |
|
Synchronous |
POST is made to service, response is job token. Results are collected later or pushed via webhook
|
|
POST is made to service, response is optimization results |
Marginally slower than synchronous due to the small latency of having to poll or wait for webhook response
|
|
Faster than asynchronous as the results are returned the moment the optimization completes |
Better for larger optimization problems that may take a long time to complete
|
|
Better for very small fast optimizations that complete in near real-time |
Able to poll or request the service for progress during the optimization
|
|
Call blocks until completed - no progress can be reported |
Enables much simpler parallel processing even from a monolithic client
|
|
One call per client process / thread |
Best for: Palletization, Truck / Railcar loading
|
|
Best for: Cartonization |
No risk of timeout on the client connection
|
|
Risk of timeout on the client connection |
|
|
|
|
|
|
|
|
|
|
|
|
You should also discuss your requirements with MagicLogic.