Before a job or step can run against a server, that server has to be registered here first -- it doesn't need to be part of the fleet, just listed. This is a deliberate security boundary, not busywork: without it, anyone who can edit a job could point a step at any reachable SQL Server, registered or not. Registering a server here is what makes it eligible to be targeted at all.

A Target Server is any SQL Server instance registered here -- Minion Agent's own catalog of "everywhere work can happen," independent of where Minion Agent itself is installed. This is what lets one installation orchestrate jobs across an entire fleet of machines from a single console, unlike native SQL Server Agent, which only ever runs a job on the box it lives on.

Register one with a friendly Name (how it shows up everywhere else in the app), the real connection Host (the SQL Server Data Source, e.g. SQLDEV01\INST1), and an optional Description. Once it exists, a job or step can target it directly, inherit it from its job, or -- with Availability Group Settings -- resolve to whichever server currently holds the AG role it wants instead of a fixed one. It's also the thing Credentials and Permissions are scoped against: a credential connects to a Target Server, and a Permissions entry says who's allowed to run work against it.

Deactivating one

The grid's ACTIVE column toggles Deactivate/Activate. Deactivating blocks any job or step from running against that server -- an ad-hoc run is refused immediately, and a real job run targeting it refuses to start at all. Nothing is deleted; reactivating allows runs again right away.

New here? Start with the Quick Start Guide.