Skip to main content

How to Connect an AWS S3 Bucket

Give applications running on the Signaloid Cloud Compute Engine read and write access to a bucket you own in AWS.

Connecting a bucket takes two sides. Register it on the Signaloid Cloud Developer Platform, then grant the Signaloid AWS account access to it from your own AWS console. The dialog walks you through both.

You need the 12-digit AWS Account ID that owns the bucket, the bucket name, and permission to edit the bucket policy.

Connect the Bucket​

  1. On the Data and Sensors page, click New Data Source in the Data Sources section and pick S3 Bucket from the dropdown.
  2. Fill in the AWS Account ID and bucket name, then select the permissions your application needs. Read is selected by default and Write is not. See the Data Sources reference for what each field accepts.
  3. Click Copy Suggested IAM Policy and paste the policy into your bucket's policy editor in the AWS console, which the dialog links to directly. Without this step the Signaloid AWS account cannot reach the bucket and the connection fails.
  4. Back in the dialog, click Connect.

The bucket then appears as a card in the Data Sources section, and in the Mount Data Source picker everywhere you attach data to a run.

Figure 1. Connecting an AWS S3 bucket to your Signaloid Cloud Developer Platform account.

Your account tier caps how many buckets you can have connected at once. At the limit, the create card switches to Upgrade.

Read and write the Bucket from your code​

The bucket behaves like any other data source once connected. Mount it at a location of your choice from the Code Playground toolbar or from a connected repository, then open paths underneath that location. With the bucket mounted at mountAWSBucket:

FILE *  fp = fopen("mountAWSBucket/myInputFile.csv", "r");

Writing only works if you granted the Write permission when you connected the bucket. You can change the permissions later from Edit Configuration on the bucket card, but the account ID and bucket name are locked once the source exists.

Disconnect the Bucket​

Click Disconnect () on the bucket card. What happens next depends on whether anything else depends on the bucket:

  • If no Data Drive mounts it, the bucket disconnects once you confirm.
  • If one or more drives mount it, the dialog lists those drives by name and disables Disconnect until you remove the bucket from each of them. This is deliberate. A drive that mounts a source removed underneath it would break.

Disconnecting cannot be undone, but it frees a slot against your quota and does not delete any objects in the bucket itself.

Figure 2. Disconnecting an AWS S3 bucket from your Signaloid Cloud Developer Platform account, with the Disconnect flow blocking the action while the bucket is mounted on a Data Drive.
Last updated on