Architecture options

Now that you know what NeoLoad is, you need to decide where to deploy it. NeoLoad is flexible: you can run it entirely in the cloud, entirely on your own infrastructure, or as a mix of both.

This topic helps you choose the deployment model that fits your requirements. For the technical details you need to prepare before deployment, see deployment considerations.

Core components

Before you choose a deployment model, it helps to understand the four components that make up a NeoLoad environment:

  • Controller: Orchestrates the test. It manages the scenario, coordinates load generators, and collects results. Every test run requires exactly one controller.

  • Load generator: Simulates virtual users and generates traffic against your application. A test can use one or many load generators across different zones.

  • Monitoring agent, optional: Collects server-side metrics like CPU, memory, and disk from your application infrastructure during the test. This helps you correlate user-facing response times with server-side behavior.

  • NeoLoad Web: The centralized platform where you manage tests, run them, and analyze results. It's the hub your team uses every day.

Deployment models

NeoLoad supports three deployment models. The following table compares them:

Aspect Cloud / SaaS On-premise Hybrid
Who manages NeoLoad Web Tricentis. You. Tricentis.
Where load generators run Tricentis cloud. Your servers. Both.
Where NeoLoad stores results Tricentis cloud. Your database. Tricentis cloud.
Network requirements Outbound HTTPS to Tricentis cloud. Internal only. Internal + outbound HTTPS.
How it scales Elastic. Add cloud load generators on demand. Fixed. Limited by your hardware. Elastic. Burst into cloud when needed.
What you manage Controller only. Everything. On-premise controllers and load generators.

Choose the right deployment

Use the following criteria to decide which model fits your situation. If you're not sure, start with cloud. It's the fastest way to get started and you can always add on-premise components later.

Choose cloud if

Cloud is the right choice when you want to start quickly and your application is accessible from the internet:

  • You want minimal infrastructure work. Tricentis manages NeoLoad Web and the load generators.

  • Your application under test is accessible from the internet, or you can allow Tricentis IP ranges through your firewall.

  • You need elastic scaling. Provision load generators for a big test, then release them when you're done.

  • You're comfortable with test results stored in the Tricentis cloud.

Choose on-premise if

On-premise is the right choice when compliance or network constraints require everything to stay within your infrastructure:

  • Compliance or data residency rules require all test data and traffic to stay within your network.

  • Your application is in an air-gapped environment with no internet access.

  • You need complete control over the infrastructure, including the NeoLoad Web server and database.

Choose hybrid if

Hybrid is the right choice when you need on-premise load generation but want cloud-based analytics:

  • Your application isn't accessible from the internet, but you want centralized, cloud-hosted results and dashboards.

  • You want to burst into cloud load generators for large-scale tests while you keep on-premise capacity for day-to-day work.

  • Your application has both internal and external components that need load generation from different networks.

Architecture requirements by model

Each deployment model has different requirements for subscriptions, infrastructure, network, and access. The following sections give you the high-level picture. For specific hardware, network, and sizing details, see deployment considerations.

Cloud requirements

A cloud deployment requires the following:

  • A NeoLoad Web SaaS tenant, provisioned by Tricentis.

  • A server for the controller. You manage this server.

  • Outbound HTTPS on port 443 from the controller to the Tricentis cloud.

  • An API access token for the controller to authenticate with NeoLoad Web.

On-premise requirements

An on-premise deployment requires the following:

  • Servers for the controller, load generators, and optionally the monitoring agent.

  • A server for NeoLoad Web with a database backend.

  • All components must be able to reach each other: the controller, load generators, monitoring agent, and NeoLoad Web.

  • Load generators must reach the application under test.

Hybrid requirements

A hybrid deployment requires the following:

  • A NeoLoad Web SaaS tenant, provisioned by Tricentis.

  • Servers for the controller and on-premise load generators. You manage these servers.

  • Outbound HTTPS from on-premise components to the Tricentis cloud.

  • Cloud load generators, managed by Tricentis, are available for burst capacity.

Roles in the deployment process

Deployment is a collaborative effort. The following roles typically participate:

  • Enterprise architect: Evaluates the deployment models and decides which one fits the organization.

  • Technology expert: Provisions infrastructure, configures network access, and makes sure system requirements are met.

  • Testing expert: Installs and configures the testing components, including controllers, load generators, and monitoring agents, and creates the first tests.

  • Tricentis cloud operations, for cloud and hybrid deployments: Provisions the NeoLoad Web SaaS tenant, configures entitlements, and confirms readiness.

Share the relevant sections of this guide with each role so everyone knows their responsibilities.

What's next

Now that you've chosen a deployment model, here's where to go: