Networking & infrastructure¶
This section covers cluster networking, ingress, deployment tooling, resource tuning, autoscaling, and observability for your DataRobot installation.
The table below lists the topics included in the networking & infrastructure section:
| Topic | Description |
|---|---|
| Docker networks | Configurable networking options for Docker-based installs. |
| DataRobot ingress reference settings | Ingress requirements in Kubernetes and supported ingress controllers. |
| Deploy DataRobot with ArgoCD | Deploy DataRobot using ArgoCD. |
| Add volumes | Attach additional volumes or a ConfigMap with extra application configurations. |
| Vertical Pod Autoscaler | Configure Kubernetes VPA to automatically right-size resource requests. |
| KEDA event-driven autoscaling | Cluster-wide prerequisites and configuration for KEDA and the KEDA HTTP Add-on. |
| LRS auto-scaling with KEDA | Dynamic horizontal scaling for custom model deployments. |
| Install in a restricted network | Prerequisites for installing DataRobot in a disconnected or air-gapped environment. |
| Tune DataRobot clusters | Use configmaps to populate environment variables in cluster pods. |
| DataRobot services overview | An overview of the internal services used to build a DataRobot cluster. |
| Modify application log levels | Configure DataRobot log levels via configmap keys. |
| Send logs to AWS CloudWatch | Set up logging for CloudWatch. |
| Download the external monitoring agent | Download the installer package to run external deployments. |
| User Metrics Collector | Collect DataRobot deployment and NIM/GPU metrics. |