how to connect and load on premise sql server data to aws redshift through glue for etl operations?

Question

I have been tasked to connect to the on premise sql server (which is installed in my personal laptop) to AWS. The task is very simple and straight forward.

Basically, we do ETL operation on aws using glue, with different data sources and different data targets.

For example -> loading data from s3 -> Running glue crawler to get the schema -> and run glue job to load data into target like redshift or RDS or s3.

I have installed RDS SQL SERVER on aws and ran the crawler and job to load data from source to redshift. This works fine.

But manager asked me to connect to on-premise to aws and run glue job to load data into the target.

I came across AWS DIRECT CONNECT service to do this job.

My confusion is,

1. Is this the right approach?
2. How about connecting multiple on premise to AWS?
3. What exactly the AWS DIRECT CONNECT is?