Installation Overview
Installing the DynamoAI platform follows three stages, in order:
- Environment Dependencies: verify everything your environment must provide, per cloud, before the deployment window starts.
- Deploy: install the platform with Helm.
- Deployment Validation: run the acceptance checks that prove the platform is operational.
Before starting, read the Platform Architecture page for the install model, and Resource Guidelines to size compute for your expected workload. Container Image Delivery covers registry and node preparation for the platform's multi-gigabyte worker images, which is the most common cause of a stalled first deployment.
Environment Dependencies
3 items
Deploy
Helm installation guide for the DynamoAI platform, covering the dynamoai-base, dynamoai-dynamoguard, and dynamoai-dynamoeval charts.
Deployment Validation
Acceptance checks that prove a DynamoAI deployment is operational, per feature, with target results and interactive checklists.
Container Image Delivery
Prepare a registry, node capacity, and pre-pull strategy for DynamoAI container images before a deployment or version upgrade.
Workstation Tools
The engineer driving the deployment needs these command-line tools:
| Tool | Used For |
|---|---|
| Helm | Installing and managing the platform charts |
| kubectl | Interacting with the Kubernetes cluster |
| Docker | Container image management and mirroring |
| Cloud CLI (AWS, Azure, gcloud, or oc for OpenShift) | Provisioning or configuring cloud infrastructure directly |
| Terraform or equivalent | Only if infrastructure (cluster, IAM, queues) is provisioned as code |