Skip to main content

AWS integration

Seqera Platform integrates with AWS to run pipelines on AWS Batch, AWS Cloud, and Amazon EKS, and to read input/output data from S3, EFS, and FSx. This page is the entry point for the AWS-side setup.

Setup checklist

StepWhat to doPage
1Create an IAM policy with the permissions Seqera needsIAM policies
2Create the IAM user (and optionally an IAM role)Credentials
3Create an S3 bucket — and optionally EFS or FSx — for the work directoryData access
4(EKS only) Set up the Service Account IAM role and aws-authEKS additions
5(Manual Batch only) Create AWS Batch resources by handAWS Batch manual setup
6Add the credentials to Seqera and create the compute environmentAWS Batch, AWS Cloud, Amazon EKS

Choose a credential model

Seqera supports two AWS credential modes:

  • Key-based credentials — Access key + secret key with direct IAM permissions. Optionally pair with Assume role to scope down via role assumption.
  • Role-based credentials (recommended) — Role assumption only, no static keys. Seqera generates an External ID automatically.

For Seqera Cloud, role-based credentials let you allow only the Seqera Cloud access role arn:aws:iam::161471496260:role/SeqeraPlatformCloudAccessRole in your trust policy. See Role-based trust policy example.

Choose a compute environment type

Compute environmentWhen to use itIAM scope
AWS BatchProduction-scale pipelines, mature workloadsLargest IAM surface (Batch + EC2 + IAM + optional services)
AWS Cloud (preview)Studios and small/medium pipelines, simplified setupSmallest IAM surface (single role)
Amazon EKSPipelines on existing Kubernetes clustersEKS describe + Kubernetes RBAC

The IAM policy you create depends on which compute environment(s) you plan to use. See the tabbed policy in IAM policies.