There are two ways to know that your company's server has a full disk. The first is email stopping working at nine in the morning and someone calling IT support. The second is an alert, ten days before, saying that the disk is going to reach its limit — and someone freeing up space before anyone notices. The first is reactive support; the second one is proactive monitoring. The difference in cost between the two is rarely on the supplier invoice. It is in the downtime of the business.
The reactive model, and why it remains the norm
Reactive support is straightforward to understand and to buy: when something breaks, someone fixes it. It doesn't require installing anything, it doesn't generate alerts, and it doesn't cause a nuisance. That's why it remains the norm in many Portuguese SMEs — and that's why the majority of incidents we resolve in companies without a contract are incidents that had shown signs days in advance. A hard drive doesn't fill up in a second. Network equipment doesn't start failing out of the blue. A backup that fails one night, fails on the subsequent ones. Without anyone watching, the signal is missed.
What proactive monitoring measures
Monitoring is the continuous measurement of the state of each infrastructure component and comparing it with a defined threshold. When the threshold is exceeded, it triggers an alert with a severity level — warning, important, critical — which determines who is notified and by what means. In a typical SME, the following are measured:
- Staffdisk space, disk status (SMART, RAID), CPU and memory, essential services, temperature, power supplies.
- Networkavailability of switches, access points and firewall, port errors, link saturation.
- Internet connectionsmain line and redundancy, latency, operator failures.
- Backupscompleted or failed, abnormal size, execution time.
- Jobspending updates, outdated antivirus, equipment that stopped communicating.
- Deadlinescertificates, domains and licences expiring.
At DataRoad the platform is Zabbix, with Grafana dashboards, and each client views the dashboards for their own infrastructure. The full description of the service is on the page for 24/7 IT monitoring and NOC.
Reactive vs proactive: the same incident, two outcomes
Scenario 1 — file server disk filling up. Reactive: at 9.12 am on a Tuesday, the server stops accepting entries; the accounts department is unable to export data; a support ticket is opened; the technician calls at 10.30 am and frees up space by 11.00 am. Two hours of 20 people left waiting. Proactive: a warning alert sent to 80% twelve days earlier; an automatic ticket is created; space is freed up overnight; nobody even noticed.
Scenario 2 — backup failing. Reactive: nobody notices for three weeks; when the server breaks, the last good backup is 21 days old. Proactive: alert on the first night the backup fails; fixed the next day; the oldest backup that was ever missed is 24 hours old.
Scenario 3 — intermittent Wi-Fi access point. Reactive: vague complaints of “the internet is slow” for months; nobody links the issue to the hardware. Proactive: the dashboard shows the same access point dropping and reconnecting 40 times a day; it is replaced and the complaints stop.
What monitoring does not do
It is advisable to be clear. Monitoring does not stop a disk from failing without warning nor an operator from cutting the line. What it does guarantee is that the failure is detected in minutes rather than hours, that the response begins immediately and that there is a record of what happened. Nor does it replace backups or a plan for disaster recovery — they are different layers, and a serious contract includes all three.
A tool is not a service
Many companies already have “monitoring”: a tool installed years ago that sends emails to an inbox that nobody opens. That is not proactive monitoring, it is noise. What turns an alert into prevention is what happens after it: someone who sees it, a ticket that is opened, an intervention that is carried out and a log that is kept. That is what distinguishes a tool from a NOC — an operations centre with a team, escalation procedures and severity levels tailored per client. And that is why, in the contracts IT Unlimited, monitoring is not an extra: it is the foundation that allows the provider to promise unlimited support without going bankrupt on emergencies.
Where to start
If your company still operates on a reactive model, the path is short: a survey of the estate to identify what is critical, the installation of agents and probes, two to three weeks of tuning and, from then on, a monthly report on the state of the estate and recommendations. It does not require changing anything in the way the team works. It simply requires deciding that you prefer to know beforehand.
Frequently Asked Questions
Is proactive monitoring the same as having an antivirus?
No. The antivirus protects a workstation against malicious software. Monitoring measures the status of all equipment and services — disk, memory, connections, backups, certificates — and alerts you when something is out of the ordinary. They are different and complementary layers.
How long does it take to get the monitoring up and running?
The installation of the agents and probes takes a few days. The following two to three weeks are used to fine-tune thresholds and eliminate false alarms. After a month, the monitoring is stable and the first report is delivered.
Does monitoring see the contents of our files?
No. It collects technical metrics — space, load, availability, errors — and does not access the content of files, emails or databases.
Does it make sense for a company that closes at 6 pm?
Yes, the night-time alerts are logged and dealt with in the morning before the team arrives. What a company with fixed working hours might not need is 24/7 call-outs — the monitoring itself is always useful.
Want to know what your infrastructure is doing right now?
We provide a no-obligation survey. Discover the DataRoad's 24/7 IT monitoring or call 211 459 950.
