When to use Modeler vs. Integration Server?

good morning people

i need to know what is the sequence of integrated components that we follow while developing a webMethods project,i.e at first we start with IS or modeler or developer… etc.I need a clear review on it because people overthere have worked on various real-time projects.so, i hope i may get a solution for this query.

With Regards,

I recommend using Modeler only when the process involved is stateful and long-running. Long running could be anything from 5 minutes to 5 months.

This is just a rule of thumb. Sometimes it may make sense to use Modeler if the elapsed time for the process is 2 minutes, but in general if there is a user waiting for the result of a service or “process” then don’t use Modeler, use Developer instead.

The other question to ask is "does it make sense to resubmit a transaction that fails somewhere in the middle of the service or process. A corallary to this question is "will my organization ever assign someone to monitor this “process” and resubmit failed transactions. If the answer is “no, we’ll just report an error to the user or we’ll just resubmit a new transaction” then the service is not a good candidate for Modeler.

Every step transition in Modeler results in messages being published to Broker and rows in database tables being updated or inserted. If the entire “process” is only going to take a few seconds, implementing it in Modeler / PRT will add significant overhead which in most cases will not be worth the value added.