Agent configuration for Step SaaS
Automation scripts are executed on Agents. Certain jobs require specific agents: Client-based automation (like browser-based testing) require UI agents. Most plugins require a specific agent type (java for most of them), we list the most relevant tasks here. Configuring this for Step SaaS in the Step Portal is an easy process and can either be done when creating a cluster or by editing an existing one.
Agent Types
Here all available Agent types are listed, including information about the available resources per running agent.
Java agent
All Java agents in Step SaaS use Java 11. The agents have different specialities:
Agent [java, ui-automation] This pool of agents is designed for java UI automation(selenium, cypress…) with one browser or client per agent. The number of agents defines the number of scripts we’ll be able to run in parallel. (1750 milliCPU, 1800 MB memory)
Agent [java, api-automation] This pool of agents is designed for java (lightweight) automation code, such as protocol based testing. One agent is allowed to execute up to 100 scripts in parallel. (1750 milliCPU, 1800 MB memory)
.NET agent
All .NET agents in Step SaaS use .NET 6. The agents have different specialities:
Agent [dotnet, ui-automation] This pool of agents is designed for dotnet UI automation with one browser or client per agent. The number of agents defines the number of scripts we’ll be able to run in parallel. (1750 milliCPU, 1800 MB memory)
Agent [dotnet, api-automation] This pool of agents is designed for dotnet (lightweight) automation code, such as protocol based testing. One agent is allowed to execute up to 100 scripts in parallel. (1750 milliCPU, 1800 MB memory)
Selecting the right agent
The following agents are needed for certain scenarios:
- JMeter: Agent [java, api-automation]
- Cypress: Agent [java, ui-automation]
- Playwright: Agent [java, ui-automation]
- Script: JS, Groovy, Java: Agent [java, ui-automation] or Agent [java, api-automation] depending on the task
- Astra: Agent [dotnet, ui-automation]
- Custom .NET: Agent [dotnet, ui-automation] or Agent [dotnet, api-automation] depending on the task
Agent tokens
As described here in more detail, Keywords are executed on so-called agent tokens. One agent can emit multiple groups of agent tokens. The token groups are configured using the parameter tokenGroups. Each token group has its own number of tokens (capacity) and attributes.
Capacity
The number of agent tokens is defined by the agent type, agents for ui-automation have one agent token per agent whereas agents for api-automation have 100 agent tokens per agent.