SCALE.sdm

SCALE.sdm is a platform for simulation data and process management deployed on AWS using EC2, RDS, EFS, EKS, and an Application Load Balancer. This section describes deployment, configuration, security, monitoring, and removal.

The application is deployed on Amazon EKS using Helm charts. Configuration is performed using a values.yaml file. In production environments, customers should pin container image tags, configure CPU and memory resources, define liveness and readiness probes, and enable exactly one exposure method (Ingress or Gateway API).

Regionalization of container images: All container images are hosted in region-specific Amazon ECR repositories. During deployment, images are pulled automatically from the AWS region where the EKS cluster is running. Cross-region image pulls are not required. If images are pulled from a private registry, imagePullSecrets must be configured.

Sensitive data location: Customer data including metadata, configuration, and logs is stored in Amazon RDS and Amazon EFS. No sensitive data is stored on EC2 instances. Database credentials and other secrets are managed using AWS Secrets Manager.

Encryption and cryptography: Amazon RDS and Amazon EFS are encrypted at rest using AWS KMS. All communication between components is protected using TLS. No customer-managed cryptographic keys are required. Key rotation is handled automatically by AWS.

Backup and recovery: Enable automated backups or snapshots for Amazon RDS. Amazon EFS can be backed up using AWS Backup. Recovery is performed by restoring from RDS snapshots or AWS Backup recovery points.

Deployment step: Bootstrap AWS CDK with appropriate IAM permissions.

Prepare a custom values.yaml file to configure replicaCount, image repository and tag, imagePullSecrets, service accounts, security contexts, networking, resources, autoscaling, and storage.

Install the application using helm upgrade --install with the custom values file and target namespace. Repeat for additional SCALE.sdm components if required.

Wait approx. 10 minutes for all pods and services to become ready.

Helm chart configuration overview: replicaCount controls the number of pod replicas. The image section defines repository, tag, and pull policy; fixed tags are recommended in production. imagePullSecrets are required for private registries. nameOverride and fullnameOverride allow explicit resource naming. serviceAccount controls IAM integration and token usage. podAnnotations and podLabels add metadata for monitoring and policies. podSecurityContext and securityContext define pod- and container-level hardening such as non-root execution. service defines internal or external exposure. ingress or httpRoute (Gateway API) exposes HTTP(S) endpoints and requires a compatible controller. resources define CPU and memory requests and limits. livenessProbe and readinessProbe configure health checks. autoscaling enables HPA and requires metrics-server. volumes and volumeMounts attach Secrets, ConfigMaps, or persistent volumes. nodeSelector, tolerations, and affinity control pod placement.

Using a custom values.yaml file is recommended. Multiple values files may be combined, with later files overriding earlier ones. Individual parameters can also be overridden using --set, including array values with zero-based indexes.

Uninstallation: Scale down operator workloads if applicable. Uninstall the Helm release using helm uninstall. Optionally delete the Kubernetes namespace.

Health monitoring and pricing: Monitor EC2, RDS, and EFS using the AWS Console. Pricing depends on instance sizes, storage consumption, data transfer, and backups.

Support and documentation: Technical support is available at support@scale.eu . Detailed installation documentation is available at https://www.scale.eu/sdm-manual/current/technical-manual/installation.html?sdm_token=Eic4eSeeLoh8lohKieDi