Why backups fail quietly
Backup jobs stop for ordinary reasons. A password changes and the backup can no longer log in. A new drive is added to the server but not to the job. A subscription lapses when the credit card on file expires. A laptop stays switched off for two weeks during leave. A Windows update changes a setting. None of these makes a noise. The office carries on working normally, and the backups simply stop growing.
The danger is that a backup report saying 'last success: 147 days ago' only gets read when someone needs a restore. Monitoring exists to bring that information forward, so the gap is found on a quiet morning rather than in the middle of a crisis. It turns a backup from something you hope is working into something you know is working, which is a very different position to be in when a client or insurer asks.
What should be monitored
At a minimum, watch for jobs that failed, jobs that completed with warnings, scheduled jobs that did not run at all, and devices or users that have not been backed up within an expected period. Also watch for sudden changes in backup size. A huge jump can mean ransomware encrypting files, which creates a lot of apparently new data, while a sharp drop can mean a folder or database has fallen out of the backup.
For cloud backups of Microsoft 365 or Google Workspace, watch for new users or shared drives that are not yet protected, and for the backup app losing its permissions to the tenant. For local appliances, watch storage capacity and hardware health, because a full or failing backup device stops protecting anything.
Who receives the alerts
Alerts sent to a generic inbox that nobody reads are worse than useless, because they give the impression that monitoring is happening. Send backup alerts to a named person, or better, to an IT service desk with a process for acting on them. Make sure the alerts reach someone who can actually fix the problem, or who knows who to call.
Avoid sending every alert to the business owner unless they want them. Too many messages, especially noisy warnings about minor issues, quickly teach people to ignore the lot. Tune alerts so that failures and missed backups stand out, and minor warnings are reviewed in a daily or weekly summary. Revisit the settings after a month or two of real alerts and adjust anything that triggers often without needing action.
What a good daily check looks like
A practical daily check takes a few minutes. Open the backup dashboard and confirm that every server, device and cloud account completed a backup in the last day. Investigate any red or amber items and note what was done. Check that any devices offline for more than a few days are expected, for example a staff member on leave. Glance at storage usage and trends.
Weekly, review the list of protected items against your user and device lists to catch anything new that was missed. Monthly, do a test restore. Quarterly, review retention settings and alert recipients. Writing these routines down means they keep happening when the usual person is away. A simple checklist in a shared document, ticked off with a date, is enough.
Silence should mean success
The aim is a system where no news genuinely means good news. That requires alerts for missed jobs as well as failed ones, because a job that never started produces no error message on its own. Some monitoring tools send a regular heartbeat; if it does not arrive, that is treated as a failure. This catches cases where the backup server itself has gone offline.
If your backup only emails when something works, consider changing it. Successful-job emails pile up and get filtered away, and their absence is easy to miss. It is better to receive nothing when all is well and a clear, unmissable alert when it is not. Test the alerting itself occasionally by pausing a non-critical job and confirming the right person hears about it.
Using monitoring as evidence
Monitoring records also show insurers, auditors and clients that backups are actively managed. A monthly report showing job success, test restores performed and issues resolved gives a clear picture of how well protected the business is. It also helps when choosing whether to keep or change your backup setup, because you can see how often things go wrong and why.
If an outside provider looks after your backups, ask for that report as part of the service. You are entitled to know, in plain terms, whether your backups worked last month and what was done about any failures. A provider confident in its service will be happy to show you. If the answer is vague, treat that as a signal to look more closely at how your backups are run.
In short
- Most backup failures are silent; daily checks catch them early.
- Monitor missed jobs, warnings, stale devices and big size changes.
- Send alerts to a named person or service desk that acts on them.
- Ask for a monthly report if someone else runs your backups.