Clusterra documentation

Clusterra is a managed Slurm service for biotech and life-sciences teams that need HPC without hiring an HPC admin. You bring an AWS account; we run the scheduler, the autoscaler, the job console, and an AI agent that speaks GROMACS, AMBER, and Nextflow.

Getting started

Connect an AWS account, provision a cluster, run your first job.

Architecture

How the central control plane and your cluster talk over VPC peering.

User guide

Submit jobs, stream logs, manage files, understand autoscaling and quotas.

API reference

Full OpenAPI spec for the cluster-api service at api-use1.clusterra.cloud.

What Clusterra gives you

Managed Slurm control plane. We run slurmctld, slurmdbd, and slurmrestd for you in our central cluster. You never SSH anywhere to patch it.
Your AWS account, your data. Compute nodes run in your VPC under your IAM, paid for by your AWS bill. The central plane connects via a cross-account role and VPC peering.
Autoscaling that understands Slurm. Karpenter in your cluster scales slurmd workers to match the pending queue, then drains them when the queue empties.
A console built for running jobs. Submit from a form or srun, watch stdout stream live from EFS, see cost per job, cancel in a click.
An AI agent that knows your stack. Ask it about Slurm options, GROMACS flags, or Nextflow params and it can answer with access to your cluster state.

Where things live

Service	URL
Console	console.clusterra.cloud
API (prod, us-east-1)	`api-use1.clusterra.cloud`
Docs	/docs/
Marketing	clusterra.cloud

Status: Beta. The product is in active development; API shapes may change between minor versions. Design-partner customers get a heads-up on breaking changes.