Hi svx23 I've sort seen an issue similar to this, even though the servers were in the same zone, I had horrible latency problems.  There ended up being extra hops between the servers the contractor infrastructure guys had to figure out, something to do with the DR arrangement from memory.  In my case though it was a traditional Microsoft SQL Server VM on one machine and a Microsoft RDS access server in another.  However there are bigger questions to check in this arrangement:

1. How are the clients accessing Sage 300 - where are the clients executing from?  Is there a Microsoft RDS server in the same zone?  Or are you deploying the Sage 300 Web screens to an IIS server in the same zone?  Or something else?
2. Using the Activity Monitor, capture an expensive query and execute it directly in the Management Studio on the Amazon RDS database - how does it perform vs the same activity in the Sage 300 client?
3. What tools can you use do run a tracert to the Amazon RDS environment - can you check the hops?  Is there a way you can test latency and throughput?  Make sure you're running this test from where the client is executing and not from the EC2 application server.

Remember this deployment is very much unsupported so will do my best to suggest options but it could be a tricky one.

Cheers

Good luck on AWS EC2.  I would tell the client if the database isn't on the same server as Sage, you're going to be in a lot of hurt.

 svx23 Have you got to the bottom of this issue please as I have a client who has reported the same issue with same setup as yours?