Deploy Valohai workers to your environment – Valohai Help

Valohai can be deployed in your environment, either in the cloud or on-premise. The platform can also be deployed in multi-cloud setups and in hybrid setups with both cloud and on-premise resources.

When using Valohai:

Your data stays in your environments (object storage, databases, and data warehouses).
The machine learning jobs are performed on your own virtual machines.
Models are deployed to your own object stores and registries.

Cloud	Architecture	Installation
AWS	Architecture	CloudFormation / AWS Terraform / Deploy to AWS manually
GCP	Architecture	GCP Terraform / Deploy to GCP manually
Azure	Architecture	Deploy to Azure

In short, Valohai allows you to spend less time on infrastructure management and manual experiment tracking and let your data scientists concentrate on building custom models.

Valohai components

The Valohai application doesn’t directly access your workers but instead writes requests and updates to the job queue machine, from where your own workers find jobs that are scheduled for them.Valohai comprises of two layers, the Application Layer and the Compute & Data Layer.

Application layer

Compute & Data Layer

Hosted by Valohai

Hosted in your cloud/on-premises

The web application
A PostgreSQL database for storing user and execution metadata information (who ran what, and when)
A scaling service responsible for managing virtual machine instances
A deployment image builder for building and publishing images for online-inference
The API services that enable you to:
- Launch, view, and manage jobs, pipelines, and deployments
- Access execution results, files, and metrics
- Launch batch processing jobs
- …more Valohai APIs

A virtual machine that manages the machine learning job queue
Object storage(s) to store job logs, snapshots, and generated files (e.g. models, dataset snapshots)
Autoscaled virtual machines to run machine learning jobs
Other optional services, like:
- Your existing Kubernetes cluster, to enable teams to deploy models for real-time inference
- Private Docker Registry
- Other data sources (databases, data warehouses, etc.)
- A Spark cluster

Permissions and access

Valohai will need a set of permissions in your cloud environment to autoscale your virtual machines and store logs, snapshots, and generated files in object storage.

First, you provision the core Valohai services in your cloud and register your organization in Valohai. Then you can set up your projects and connect them with your Git repositories, object storage, and private Docker registries.

See also

See the installation pages for AWS, GCP, and Azure for detailed guides and installation templates.

Service	Access	Configuration
Core Services	Access to manage virtual machines and read/write to one object storage.	Deploy to AWS Deploy to GCP Deploy to Azure
Object Storage	Read or Read/Write access to existing object storage (AWS S3, GCP Cloud Storage, Azure Blob Storage).	Configured in-app: Add a cloud data store
Git repository	SSH or Deploy key with read-only access to select repositories.	Configured in-app: Connect to a Git repository
Docker Registry	Permissions to pull from your Docker repository.	Configured in-app: Access a private Docker repository
Kubernetes Cluster	A Kubernetes service account that can manage pods and other resources in a defined namespace. Permissions to read/write to your private Docker registry.	Configuration guide

Valohai components

Permissions and access

Related articles