Close Sun Nov 18 14:53:26 GMT 2018

Amazon Web Services SDK for Apex

The AWS SDK for Salesforce makes it easy for developers to access Amazon Web Services in their Apex code, and build robust applications and software using services like Amazon S3, Amazon EC2, etc. Clone the repository and deploy in a dev org to get started: github.com/bigassforce/aws-sdk

Amazon Simple Storage Service (S3) SDK

S3 is storage for the Internet. The Apex client acts as a proxy class for manipulating both buckets and their contents. Use the API methods to create and destroy objects and presign download URLs. The bucket name and object keys may be stored in custom objects etc.

s3-flow

Sign up then go to your AWS Console > Security Credentials > Access Keys:

String access = 'XXXXXXXXXXXXXXXXXXXX';
String secret = 'YYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYY';
AwsSdk.Connector connector = new AwsSdk.Connector(access, secret);

Creating a bucket:

AwsSdk.S3 s3 = connector.s3(region);
String name = 'thebucket';
s3.createBucket(name);

Adding an object to a bucket:

AwsSdk.S3.Bucket bucket = connector.s3(region).bucket('thebucket');
Map<String,String> headers = new Map<String,String>{'Content-Type' => 'text/plain'};
bucket.createContent('foo.txt', headers, Blob.valueOf('bar'));

Viewing an object:

AwsSdk.S3.Content content = connector.s3(region).bucket('thebucket').content('foo.txt');
HttpRequest request = content.presign();
String url = request.getEndpoint();

Amazon Elastic Cloud Compute (EC2) SDK

EC2 provides scalable computing capacity in the cloud. The Apex client calls services to launch instances, terminate instances, etc. The API responds synchronously, but bear in mind that the the instance state transitions take time.

ec2-flow

Describing running instances:

AwsSdk.Ec2 ec2 = new AwsSdk.Connector(access, secret).ec2(region);
AwsSdk.Ec2.DescribeInstancesRequest request = new AwsSdk.Ec2.DescribeInstancesRequest();
ec2.describeInstances(request);

Launching a new instance:

AwsSdk.Ec2.RunInstancesRequest request = new AwsSdk.Ec2.RunInstancesRequest();
request.imageId = 'ami-08111162'; //amazon linux machine image
ec2.runInstances(request);

Terminating an existing instance:

AwsSdk.Ec2.TerminateInstancesRequest request = new AwsSdk.Ec2.TerminateInstancesRequest();
request.InstanceId = new List<String>{'i-aaaabbbb'};
ec2.terminateInstances(request);

Salesforce design considerations

When interacting with the AWS offerings, we encounter a few variations: signature vs transport, request vs response formats, REST vs SOAP oriented, etc. To maximize developer confidence, we use strongly typed DTO for request and response where appropriate. As the API footprint increases to interact with more services, we re-use several utilities:

Where supported, region-agnostic API endpoints are used to avoid proliferation of Remote Site Settings. It would be prudent to store any access keys and secret credentials in a Protected Custom Setting.

Anon
Thanks very much for sharing your S3 sdk. It has helped me integrate successfully very fast. Really appreciate your work.


Related