Product guide

Agent Deployment

Generate, distribute, and verify organisation-specific agent packages for Windows, macOS, and Linux.

What Agent Deployment Is#

Agent Deployment is where a device becomes a managed endpoint. It produces an installer that already knows which organisation it belongs to, so enrolment does not require manual configuration on the device.

The page also serves as the reference for deployment: supported operating systems, installation privileges, and the verification step that confirms the endpoint is reporting.

Agent deployment screen — Sensor Deployment Gateway with pre-configured agent binaries for Windows (MSI/EXE), macOS (PKG), and Linux.

Agent Packages#

Operating system	Installation behaviour
Windows	Installs as a service, supports silent installation for mass deployment, and works on Windows 7 and later.
macOS	Installs as a system service with a configuration profile for the required permissions.
Linux	Available as a package for Debian and Ubuntu, and for CentOS and RHEL based distributions.

One package per organisation

The package contains enrolment information for your organisation only. Do not reuse a package generated for another environment; the endpoint would enrol into the wrong tenant.

Deployment Steps#

1
Generate the package
Open Agent Deployment in the platform and select the operating system. The package is generated for your organisation, so it already contains the enrolment information.
2
Distribute the package
Deliver the package with your existing software distribution tool, a login script, or manually. Silent installation flags are provided for mass deployment.
3
Install with the required privileges
The agent runs with system privileges so it can read system data and enforce policy. The installing account must have administrative rights on the device.
4
Confirm the first check-in
The endpoint appears in the fleet after the agent completes its first check-in. If it does not appear, verify connectivity and the installation log.

Enrollment and Verification#

Verify the endpoint appears in the fleet view within a short time of installation.
Confirm the agent is healthy in the endpoint detail before you apply policies to it.
Assign the endpoint to a group and apply an agent profile so its configuration is consistent with the rest of the fleet.
If a device never appears, check that it can reach the appliance and review the installation log on the device.

Frequently Asked Questions#

Does the agent run as a service?

Yes. It runs as a system service so it can collect data and enforce policy whether or not a user is signed in. Because of this, the platform reads the current operating system user separately from the service account.

How much resource does the agent use?

The agent is designed to stay within a small resource budget on the endpoint. Data collection is throttled so that a busy endpoint does not become slow because of monitoring.

What happens when the agent cannot reach the appliance?

The agent stores its data locally and delivers it when the connection returns, sending it gradually so it does not saturate the link. Local data is kept for a limited period and is pruned afterwards.

Can I deploy the agent silently?

Yes. Silent installation is supported so the package can be pushed by a software distribution tool without user interaction.