Macrium Site Manager Deployment

Understand the Three Parts First

There are three moving parts, and mixing them up causes most deployment confusion:

The server (console) — installed on one always-on Windows PC or server. It runs a web console you open in a browser, and it holds the backup definitions, schedules and the list of computers.

The agent — a small service installed on every PC and server you want to protect. The agent does the actual imaging work on its own machine and reports back to the console.

The repository — the storage location where all backup images are written: a NAS, a dedicated backup server share, or disks attached to the Site Manager server.

The server itself does not copy your files — it commands, schedules and watches. That is why the server can be a modest machine, but the repository storage must be sized seriously.

Install the Server and Finish the First Run Wizard

Install the Site Manager server software on a machine that stays on — a real server is best, and a dedicated office PC is acceptable for small fleets. Run the installer and complete the first run wizard: choose the backup product, then either start the trial or add your licence keys.

The trial creates trial seats for a handful of workstations plus a server, valid for 30 days — enough to prove the whole system before paying. For a real office, choose the secure setup option when offered; the simple setup is meant for throwaway test labs.

Log in with a Windows domain or local administrator account and confirm the console opens in the browser. Do not skip the wizard and start adding computers first — licensing and the setup mode decided here quietly affect everything later.

Add Computers and Push the Agent

With the console ready, add the machines you want to protect from the Computers page. Click Add and pick computers from Active Directory, by name, or by browsing the network.

New computers show as Disconnected — that simply means the agent is not installed yet. Select them and choose Install Agent. The console pushes the installer using Windows File and Print Sharing, so that must be reachable on the target PC.

If the push fails on a locked-down PC, use Manual Agent Installation from the More Actions menu: download the agent installer (.msi), run it on that PC, and enter the Site Manager server address when asked. Confirm each computer flips to Connected before moving on.

Start with One Pilot PC

Roll the agent out to one pilot PC first and watch it for a few days before pushing to the whole office. A single machine proves the full chain — agent connection, repository write, schedule, alert — while any mistake is still small and easy to fix.

Create the Repository

Now tell the system where the images live. Create a repository pointing at your backup storage — a network share on a NAS or backup server, or local disks on the Site Manager server itself.

Plan the space before the first backup, not after the disk fills: a full image of every PC, plus the incrementals and the retention you set in the next step, decides how big the repository must be. Keep the repository off the same disks you are protecting — a repository that dies with the PC it backs up protects nothing.

Create the Backup Definition, Retention and Schedule

Create a backup definition: a full image of the system drives, and choose your pattern — a weekly full with daily incrementals is the workhorse for most offices.

Set retention in the same definition: how many full backups and how many days of incrementals to keep before old ones are deleted automatically. Do the disk maths here — promising thirty days of old backups on a repository that only holds ten is how backup disks silently fill and jobs start failing.

Attach a schedule and assign the definition to your computers or groups. Use tags or groups — accounts, design, servers — so different machines can follow different plans. Stagger the start times so the whole office does not image itself at the same minute and choke the network.

Run One Backup Manually and Watch the Alerts

Run one backup manually from the console to prove the chain works end to end before trusting the schedule. Then open the console's status list: every computer, one list, green or red — the failures stand out the same morning, not after a crash.

Set up the backup alerts and actually read them. A failed backup nobody sees is the same as no backup at all — the dashboard only helps if someone looks at it, so make checking it part of the morning routine.

Restore-Test One PC Before You Trust It

Whichever tool you choose, the backup is only real after you have restored from it once. From the console, recover one full PC from its image and confirm it boots and the data is really there — do this once at deployment, not for the first time during a real emergency.

The honest fork: Site Manager earns its place when there are many machines, one admin, and no time. Below roughly five PCs, standalone Macrium Reflect installed on each machine is simpler and cheaper to run — you lose the single dashboard and central scheduling, but there is almost nothing to manage. Above that count, the maths flips: one missed backup on one forgotten PC costs more than the central system.