Many small and mid-sized businesses buy their first server when a problem forces the issue: the accounting software has slowed down, files are scattered across computers or backups are irregular. The right server solves these problems for years; the wrong one is quickly outgrown or costs more than it should. This guide walks through what to consider, step by step.
1. What will the server do?
The first question is not hardware but purpose. In small businesses, servers are usually used for:
- File sharing and permissions: Department folders and user access rights.
- Directory services (Active Directory): Central management of user accounts and computers.
- Business applications: ERP, accounting, CRM and databases.
- Virtualisation: Running several virtual servers on one physical machine.
- Backup: Scheduled backups of computers and other servers.
If you only need file sharing and backup, a NAS is often the more economical choice. If you run ERP, databases or virtualisation, you need a proper server.
2. Tower or rack?
| Tower server | Rack server | |
|---|---|---|
| Location | Under a desk, corner of the office | Server room, rack cabinet |
| Noise | Quieter | Louder |
| Expansion | Limited | Easy to add servers to the cabinet |
| Best for | Small single-server offices | Growing, multi-server setups |
Without a cabinet and with just one server, a tower makes sense. If you have a server room or plan more servers within two or three years, starting with a rack server avoids space and cabling problems later.
3. Processor and memory
In small offices a single-processor server is enough for most tasks. As virtualisation and database workloads grow, core count and memory capacity become more important. As a general starting point, aim for:
- 32 GB of memory for a file and directory server,
- 64 GB or more for ERP or databases,
- for virtualisation, the combined needs of all virtual servers plus at least 25% headroom.
Servers should always use ECC memory, which detects and corrects memory errors to prevent data corruption.
4. Drives and RAID
Drives are the most critical part of a server. A RAID configuration is essential so that a single drive failure does not mean data loss:
- RAID 1: Two mirrored drives. Simple and reliable.
- RAID 5: At least three drives; survives one drive failure and uses capacity efficiently.
- RAID 10: At least four drives; both speed and resilience, ideal for databases.
SSDs for the operating system and databases, with high-capacity SAS/SATA drives for large file archives, strike a good balance. Remember: RAID is not a backup. You still need a separate backup plan.
5. Power and continuity
Power cuts are one of the most common risks for servers. A UPS (uninterruptible power supply) keeps the server running during an outage and allows a clean shutdown if needed. For critical systems, choose servers with redundant power supplies.
6. Warranty and management
A server sits at the heart of business continuity. Check the scope and length of the manufacturer’s warranty. Remote management tools such as Dell iDRAC, HPE iLO and Lenovo XClarity make it easier to monitor server health and spot problems early.
Quick checklist
- Define the purpose and number of users.
- Choose tower or rack based on where it will live.
- Plan memory above your current needs and use ECC.
- Decide on the RAID configuration and plan backups separately.
- Include a UPS in the budget from the start.
- Check the warranty length and remote management features.
Send us these details and we will quote a suitable Dell, HPE or Lenovo configuration for your business, together with alternatives.