The agent push button makes it tempting to select all 40 computers and click Install. Do not. Every office has that one PC — ancient, half-broken, firewalled oddly — and finding it with a fleet-wide failure is a long day.
Agents are how Site Manager touches the real world: a small service on each PC that takes imaging orders from the console. The technology is simple. The rollout discipline is what separates a calm deployment from 40 simultaneous question marks. This guide uses the pattern that works everywhere: one pilot, a short watch, then waves — with the exact checks to run at each stage.
Your pilot should be boringly normal: a standard office desktop, on the domain, wired to the LAN, used by someone cooperative. Not the boss's machine, not the server, not the ancient accounts PC with the weird software. The pilot's job is to prove the whole chain — push install, connection to the console, backup to the repository, schedule firing on time — on the easiest possible target. If the chain fails here, it fails everywhere, and you have annoyed exactly one person.
Some PCs will refuse the push: sharing turned off by policy, a firewall profile set to public, a PC on a different subnet. Do not fight each one for an hour.
.msi.Keep a note of which PCs needed the manual route — they are usually the ones with non-standard firewall or policy settings, and that list is gold during later troubleshooting.
Let the pilot run its real schedule for at least three nights. Check: did every scheduled backup succeed, how big are the incrementals, how long does a backup take, does the user notice anything during the run? Those numbers feed your schedule staggering and your repository maths. Then roll out in waves of five to ten machines, checking the console each morning before the next wave. Servers come last, one at a time, each with its restore path thought through — they are licensed differently and matter more.
Server install, agent push, backup definitions and schedules — the whole Macrium Site Manager deployment, live and step by step.
▶️ Watch Full VideoIn most cases the agent installs and connects without disturbing the user, and backups can run the same day. Plan the rollout during normal hours, tell people a backup service is being installed, and save the server agents for a maintenance window — servers deserve the cautious treatment even when a reboot is unlikely.
Name resolution and reachability, in that order. The agent must be able to resolve and reach the Site Manager server by the address given at install, and the server's agent communication port must not be blocked by a firewall between them. If the PC was installed manually, re-check the server address typed during setup — one wrong character gives you a permanently Disconnected machine.
If the network path allows the sharing traffic and the agent can reach the server afterwards, yes — but think about where the backup data flows. An agent across a WAN writes its images across that WAN all night. For branch offices, a local repository per site is the pattern that survives contact with reality.