We have integrations with third-party systems that unfortunately make use of a very old-school, file-based protocol. Instead of an API or HTTP-based protocol where you make a request to an endpoint and you get a response, the way this thing works is that you send a file, and they send you a file back. To make all of that easier and manage the SFTP part of it, we are using the SFTP Gateway for AWS. When a file comes in as a response or as a request, we need to respond to that, or when we send a request, we need to be able to pass the response as soon as it comes back, so we made use of a managed SFTP Gateway from AWS. They handle all the setup. They handle setting up the SFTP directory and the SSH keys. We specify the config, but we don't manage it. We get access to the files through AWS's storage system, which is S3, so all of the directories and files are backed by S3. In terms of our code and our processing, we only deal with S3 files. We don't have to deal with the files that are sitting on a desk somewhere on some SFTP server, which makes it a lot easier for us and for our code. Also, when new files appear in the SFTP server, we are able to automatically respond to them or handle them. AWS has triggers on these buckets. We can set a trigger on a bucket saying that, "If there's a new file, run this Lambda or this function to pass it automatically." We don't have to manage watching the directories or files when they appear.
We have integrations with third-party systems that unfortunately make use of a very old-school, file-based protocol. Instead of an API or HTTP-based protocol where you make a request to an endpoint and you get a response, the way this thing works is that you send a file, and they send you a file back. To make all of that easier and manage the SFTP part of it, we are using the SFTP Gateway for AWS. When a file comes in as a response or as a request, we need to respond to that, or when we send a request, we need to be able to pass the response as soon as it comes back, so we made use of a managed SFTP Gateway from AWS. They handle all the setup. They handle setting up the SFTP directory and the SSH keys. We specify the config, but we don't manage it. We get access to the files through AWS's storage system, which is S3, so all of the directories and files are backed by S3. In terms of our code and our processing, we only deal with S3 files. We don't have to deal with the files that are sitting on a desk somewhere on some SFTP server, which makes it a lot easier for us and for our code. Also, when new files appear in the SFTP server, we are able to automatically respond to them or handle them. AWS has triggers on these buckets. We can set a trigger on a bucket saying that, "If there's a new file, run this Lambda or this function to pass it automatically." We don't have to manage watching the directories or files when they appear.
We use SFTP for code changes, and then deploy to AWS. We are customers of SFTP.