Systems Management

In an office without regular maintenance, a failure rarely comes out of nowhere. The server disk has been reporting reallocated sectors for weeks, the UPS battery is nearly spent, and half the computers have not installed their updates because nobody restarted them. Nobody notices until one morning the invoicing software stops.

The other common problem is knowledge that lives in one person's head: the router password, where the backup is, which licence expires when. When that person goes on holiday, every small thing turns into a stoppage.

We install a monitoring agent on every device that reports disk health, temperatures, free space and the services that have to be running. Alerts are tuned to flag real problems, because alerts that ring all the time soon stop being read.

• Windows and application updates go out in two waves: first to a few machines, then to the rest once they check out. Microsoft releases its regular updates on the second Tuesday of each month, so the schedule is predictable.
• Router, switch and NAS firmware is updated in an agreed slot outside working hours.
• Backups are checked by restoring a test file, not just by a green tick in a report.
• Antivirus status is tracked from one place, so a machine with protection switched off or outdated definitions shows up straight away.
• RAID is kept for what it is: protection against one disk failing, not a substitute for backup.

You get an equipment list with serial numbers, purchase dates and warranty and licence expiry, a network diagram, and one place where access credentials are recorded. Once a month a short report arrives: what was patched, what raised an alert, and what we suggest replacing before it fails. If you ever move to another provider, all the documentation stays with you.

For a small office or a home network the same approach can be scaled down: disk and backup monitoring on a few machines, without the monthly report.