Scaleway 2: Scaleway vs AWS - Core Services Compared
If you have built on AWS, you already know the service names. When you start looking at an alternative, the first practical question is usually simple: what replaces what? This post answers that by covering the core services that appear in most AWS architectures.
Scaleway does not mirror every AWS service, and no provider really does. But for standard production workloads, the key equivalents are there, and in a number of cases the Scaleway option is simpler, cheaper, or both.
Compute: EC2 → Scaleway Virtual Instances
AWS EC2 sits at the centre of most cloud architectures. Scaleway’s equivalent is Virtual Instances, which provide on demand compute across all three regions: Paris, Amsterdam, and Warsaw.
Scaleway offers a range of instance types, including General Purpose (GP1), Cost Optimised (PLAY2), CPU Optimised (PRO2), RAM Optimised (HM), and GPU instances. For most web application and API workloads, the PLAY2 and PRO2 ranges cover much of what you would typically run on EC2 m5 or c5 families.
One useful difference is how billing works. Scaleway charges by the hour, and when you stop an instance, the compute billing stops too. On AWS, a stopped EC2 instance can still generate charges for EBS storage. On Scaleway, storage billing is separate and clearly defined.
S3 → Scaleway Object Storage
Scaleway Object Storage is S3 compatible, which means it works with the same API, the same SDKs, and the same tools, including AWS CLI, s3cmd, and rclone. For most applications, moving from S3 to Scaleway Object Storage is straightforward. In many cases, it is just a matter of changing the endpoint URL and credentials.
Scaleway Object Storage is available in Paris, Amsterdam, and Warsaw. It supports versioning, lifecycle policies, and multipart uploads. There is no minimum object size fee, and there are no data retrieval fees.
Egress fees are where the comparison starts to matter more. AWS charges for data transferred out of S3. Scaleway does not charge egress fees within its own network, so data moving between Object Storage and a Scaleway Virtual Instance or Kubernetes cluster does not create an extra cost.
Managed Databases: RDS → Scaleway Managed Databases
Scaleway Managed Database provides PostgreSQL and MySQL, with MySQL compatibility also covering MariaDB. Both are offered as fully managed services, with automatic backups, point in time recovery, and read replicas.
The setup experience is broadly similar to RDS. You choose an instance size, storage allocation, region, and version, and Scaleway takes care of the underlying infrastructure, operating system patching, and minor version upgrades.
For PostgreSQL workloads that generate lots of short lived connections, connection pooling through PgBouncer is available. That is a common pattern in web applications and API servers.
Kubernetes: EKS → Scaleway Kapsule
Scaleway Kapsule is its managed Kubernetes service. It runs standard upstream Kubernetes, with no proprietary extensions and no real lock in beyond the Scaleway node pools themselves.
One meaningful difference compared with AWS EKS is the cost of the Kubernetes control plane. On Scaleway, it is free. AWS charges around €0.10 per hour for each EKS cluster, and that adds up quickly across production, staging, and development environments. With Scaleway, you only pay for the worker nodes you actually run.
Kapsule also integrates directly with Scaleway Load Balancers and Object Storage for persistent volume claims. Ingress works through standard Kubernetes controllers, so Nginx, Traefik, and others are all supported.
Messaging: Kinesis / SQS / SNS → Scaleway Messaging
Scaleway offers four messaging services that are already production ready: NATS for high throughput pub/sub, Queues as an SQS compatible managed queue service, Topics and Events as an SNS compatible fan out service, and Apache Kafka as a fully managed option for event streaming workloads.
For teams moving away from Kinesis, Kafka on Scaleway is the closest match, with the same protocol and the same broader ecosystem. For SQS and SNS workloads, the Queues and Topics services use compatible APIs, which makes migration simpler for applications already using AWS SDK messaging clients.
Monitoring: CloudWatch → Scaleway Cockpit
Scaleway Cockpit provides metrics, logs, and alerting. It is built on Grafana, Prometheus, and Loki, which is the same open source stack many AWS customers already use after outgrowing CloudWatch.
Basic Cockpit usage is free. You can send custom metrics from your own applications alongside the infrastructure metrics that Scaleway generates automatically. Dashboards use standard Grafana, so any existing dashboard configuration can move across without needing changes.
DNS: Route 53 → Scaleway Domains and DNS
Scaleway Domains and DNS handles both domain registration and DNS management. It supports all the standard record types, including ALIAS records for zone apex setups. The API is straightforward and works well with Terraform.
For teams using Route 53 for health check based routing or more complex traffic policies, Scaleway DNS is simpler. Whether that feels like a limitation or a welcome simplification really depends on what you actually need.
Serverless: Lambda → Scaleway Serverless Functions
Scaleway Serverless Functions supports Node.js, Python, Go, PHP, and Rust runtimes. It follows the usual function as a service model, with billing per invocation and automatic scaling down to zero.
For event driven workloads, Scaleway Serverless Containers lets you run any containerised application on a serverless platform. Depending on the use case, it is closest to AWS Lambda with container image support or AWS App Runner.
What Is Not There
Scaleway does not match the full breadth of AWS’s catalogue, but no provider really does. Services without a direct Scaleway equivalent include AWS Cognito for identity management, AWS Step Functions for workflow orchestration, and AWS SageMaker for machine learning. For identity management, Keycloak is a widely used open source alternative that can be self-hosted on Scaleway Virtual Instances. If your architecture relies heavily on any of these, it is worth factoring in the extra migration effort or the cost of self hosting alternatives.
For most web application, API, and data processing workloads, the Scaleway service set covers what you are likely to need. The next post in this series looks at the CLOUD Act risk in more detail, and Post 4 covers the cost comparison.
Ready to Have a Conversation?
If you have read this series and recognised your organisation in it, the next step is simple. Get in touch. We will have a chat, ask a few questions, and give you an honest view of what is involved. No charge. No obligation.
Back to: Migrate from AWS to Stackit Series
