What a proper on-site IT audit actually checks.

Testing the failures that stay quiet

None of this is a walk-around ticking 'looks fine'. Each check below targets something that fails silently and takes a business down at the worst possible time, and every one earns its place on the list.

UPS batteries, tested, not assumed

A battery backup (UPS) with a dead battery is worse than none: it lies to you. The unit shows a green light and reports 'online' right up until the power blips, then it drops instantly and takes the server down hard. UPS batteries wear out in a few years and give almost no warning. We load-test them: pull mains power and confirm the battery actually holds the gear up for its rated time, rather than trusting the front panel. 'Assumed good' is the number one thing we find wrong.

Which breaker feeds what

When a circuit trips at 2pm on a Friday, you don't want to be guessing which breaker killed the server. Part of the audit is mapping the switchboard: which circuit feeds the server and comms rack, which feeds the workstations, and, the one everyone forgets, which feeds the server-room air conditioner. If the aircon and the server share a circuit, one trip takes out both and the room cooks. We label it so the next person isn't guessing in the dark.

Thermal paste on PC refreshes

When a machine is already open for a refresh or a clean-out, we replace the thermal paste between the CPU and its cooler. Old paste dries out, the chip runs hotter, and a hot chip throttles itself slow and dies younger. It's a five-minute job while the lid is off, and it buys years. Skipping it is why some 'perfectly good' PCs feel gutless a few years in.

A spare power supply on the shelf

Power supplies are one of the most predictable failures, and the signs usually come first: a noisy fan, random reboots, bulging capacitors. Where we can see one coming, we leave a compatible spare power supply on-site so the fix is a ten-minute swap, not a day waiting on a part. Predicting and pre-staging beats reacting every time.

Backups, is there actually free space?

A backup that ran green for months quietly stops the day the destination fills up. We check the free space on every backup target, and that the last several jobs actually completed, because 'the backups are running' and 'the backups are working' are two different sentences. A full backup drive is a silent time bomb. Proving you can actually restore is a separate job, covered on our backup and recovery page.

The 100Mb port hiding on a gigabit network

Here is a quiet one: a single port negotiated down to 100Mb on a network that should be running at gigabit (1000Mb). A cracked cable, a cheap patch lead, a tired switch port or a dodgy wall socket drops one machine, or a whole run, to a tenth of the speed, and nobody reports it because 'the internet still works'. We check the link speed on switch ports and key machines and hunt down the cause. It is often the reason one desk swears everything is slow, and it belongs to the wider networks and Wi-Fi picture.

Are the critical services actually running?

Last, we confirm the things the business depends on are genuinely up, not just that the server is switched on. The backup service, the antivirus, the line-of-business application's services, the database. Services sometimes fail to restart after an update or a reboot and sit there stopped while everything looks fine from the outside. We check them by name, on the box.

The short version: a proper audit tests the quiet failures: load-test the UPS battery instead of trusting the light, map which breaker feeds the server and the server-room aircon, refresh thermal paste on any PC that is already open, pre-stage a spare power supply for machines showing their age, confirm backups have free space and are completing, hunt the 100Mb port dragging down a gigabit network, and verify the critical services are genuinely running.

General guidance only, a real audit is tailored to your site and equipment on the day. This is what we look for, not a substitute for someone testing your gear.