An endpoint is an address clients use to reach a service. Its form depends on the service: a browser opens an S3 website URL, while a database client connects to an RDS address and port. A REST API also has routes, such as `POST /image`, that select an operation after a request reaches the web service. section 7.6.3 section 10.1.2 section 16.3.2
Choose the endpoint for the job:
- For an S3 static website, use the website endpoint for the bucket’s region. A custom subdomain can use a CNAME when the bucket name matches it; for HTTPS, put CloudFront in front of S3. section 7.6.3
- For RDS, retrieve the instance endpoint with `aws rds describe-db-instances --query "DBInstances[0].Endpoint"`. Use its address and port from a client that can reach the database; keeping the instance accessible only within its VPC is recommended. section 10.1.1 section 10.1.2
- For a read replica, use its separate endpoint for SQL reads. The application must support routing reads to the replica, and the replica incurs charges. section 10.6.2
Importing through an RDS endpoint
On a virtual machine with access to the database, run `mysql --host $DBAddress --user wordpress -p < wordpress-import.sql`, replacing `$DBAddress` with the endpoint’s Address. The hostname identifies where the client connects; the SQL file supplies the data to import. section 10.2
$ aws rds describe-db-instances --query "DBInstances[0].Endpoint"
{
"HostedZoneId": "Z2R2ITUGPM61AM",
"Port": 3306, #1
"Address": "wdwcoq2o8digyr.cqrxioeaavmf.us-east-1.rds.amazonaws.com" #2
}
The RDS command output shows the two connection details to look for: Address and Port. section 10.1.2
