Server Management in Rwanda

What is server management, and who needs it

It’s a recurring service covering the administrative and technical upkeep of your physical and virtual servers, together with the virtualisation hosts they run on. It suits organisations whose servers were installed and then left alone, those whose IT person has moved on, and those with several servers across sites that are each looked after differently.

The problem it solves

Servers decay quietly when nobody owns them. Updates are skipped, disks fill, certificates expire, backups fail without anyone reading the report, former staff keep administrator rights, and changes made in a hurry are never written down. Eventually a failure or attack exposes all of it at once. A regular routine catches each issue while it’s small, and keeps the server in the state it was delivered in.

The monthly routine

Patching. Operating system and application updates on a schedule, tested on less critical servers first, applied in an agreed window, with a check that services restarted correctly.

Backups. Job results reviewed, failures fixed, and a test restore performed regularly, not assumed.

Capacity. Disk space, memory and processor trends checked, so upgrades are planned before they’re urgent.

Access. Administrator and service accounts reviewed, with leavers and unused accounts removed.

Health. Hardware warnings, event logs and security alerts reviewed, working with system monitoring.

Lifecycle. Operating system and hardware support dates tracked, so replacement is budgeted rather than forced.

Documentation. Every change recorded with the reason, who made it and how to reverse it.

Out-of-support servers

Unsupported systems are among the most common risks we find. Windows Server 2012 and 2012 R2 stopped receiving security updates in October 2023, and the paid extended updates for them ended in October 2026. Windows Server 2016 reaches the end of extended support on 12 January 2027. Any server on these versions needs an upgrade or replacement plan, and in the meantime should be isolated on the network and closely watched.

Linux servers have the same issue under a different name: each distribution release has a published end-of-life date, after which security updates stop. The same rule applies whatever the operating system. Know the date, plan the upgrade before it arrives, and never let a server that holds important data or faces the internet run past it.

What the work involves

1. Onboard: document every server, its role, support status, accounts and backups.
2. Fix urgent gaps: missing critical updates, failing backups, exposed remote access and shared passwords.
3. Agree maintenance windows and who approves changes.
4. Run the routine monthly, with urgent patches applied when needed.
5. Report each month on updates, backups, capacity, access and risks.
6. Plan upgrades and replacements a year ahead.

What’s included and what’s not

Typical timelines and what drives cost

Onboarding typically takes one to two days per server, depending on how well it’s documented. After that, management is a recurring fee driven by the number of physical and virtual servers, their complexity, the maintenance window hours, response times, and whether out-of-hours patching or on-site visits are included.

Equipment typically used

Remote management tools reached through a secured connection, update management for Windows and Linux, the virtualisation platform’s management console, backup software reports, the servers’ out-of-band management interfaces, and a documentation system holding configurations and change records.

Common mistakes we see

Patching done only when someone remembers. Backups reported as successful but never test-restored. Former staff still holding administrator rights. Certificates expiring and breaking applications. Out-of-support servers left running with no plan. And changes made without a record, so nobody can explain how the server got into its current state.