In this article
An infrastructure stops scaling when every new employee, office or application calls for a patch instead of fitting into a design. Companies grow in leaps — ten people in a quarter, a new office, another piece of software — and the infrastructure grows by additions: a switch hanging off the previous one, an extra disk, a consumer-grade access point. It works for a while. Then the symptoms below appear — the ones DataRoad most often finds when taking over the management of companies that have grown faster than their network.
Symptom 1: a single internet line that everything depends on
Email on Microsoft 365, business software in the cloud, IP phones: if the only connection drops, the company stops and waits for the operator. Acceptable when everything sat in the rack; unacceptable when everything is outside.
A second connection from another operator, or over mobile, with automatic failover on the firewall costs little compared with a morning of downtime. That is what we implemented at FreshDesign, with load balancing and failover — see the FreshDesign project. More context in internet redundancy for business.
Symptom 2: a rack nobody dares to open
Unlabelled cables, switches from several brands daisy-chained, the operator’s router acting as the firewall and the unspoken rule “don’t touch it”. Every fault takes hours to track down.
It is the most visible symptom and the cheapest to fix relative to the downtime it causes: reorganise, document, replace the daisy chain with stacked managed switches and segment into VLANs. See when the network needs modernising and our business network restructuring service.
Symptom 3: Wi-Fi that gives way when the room fills up
It works in the morning; after 10 a.m. it degrades and video calls drop in crowded meetings. It is not the internet: it is a network designed for 15 people serving 50 people and 120 devices. Consumer access points, badly placed and without central management, do not scale, however many you add.
The fix starts with a coverage study and ends with a managed wireless network, with roaming and a separate guest network. Read why business Wi-Fi fails and how to work out the access points you need.
Symptom 4: a server always on the edge
Disk at 90%, memory exhausted by late morning, business software slow when everyone is working. More disk is bought and the problem returns six months later, because the server was sized for the company of five years ago.
The right decision is between replacing, virtualising or migrating to the cloud, with the company of three years from now in mind — analysis in new server, virtualisation or cloud. With Windows Server 2016, the decision has a date: January 2027.
DataRoad IT Unlimited
Talk to DataRoad. We assess your operation and propose the right plan.
Symptom 5: onboarding that takes days
Email on day two, laptop on day three, access to business software the following week. Without a standard workstation image or role-based profiles, everything is done by hand. With 15 people it is tolerable; with 60 it means weeks of productivity lost every year.
The solution: a standard workstation image, accounts created from profiles and a one-page onboarding procedure, as in the 7 internal cybersecurity policies.
The common root: no inventory
If it takes more than a day to find out how many workstations, servers, access points and licences there are, the infrastructure is being managed reactively. Before investing, take an inventory — the IT estate audit checklist can be done in an afternoon.
Principles for a design that scales
- Network and Wi-Fi for twice the size The cost difference is small, the difference in lifespan is large.
- Servers: decide before buying First what stays on premises and what goes to the cloud.
- Document everything Every device and connection installed.
- Monitor before the limit Alerts that arrive before the problem.
For IT Unlimited clients, the NOC (with 24×7 coverage on Premium 365) continuously measures disks, memory, network and Wi-Fi, and the replacement plan is reviewed annually with the company. That is how we supported Velv by OnRising from 90 to 200 workstations without changing the architecture.
Frequently asked questions
How many years should infrastructure be sized for?
Network and cabling for 10 years or more, because the building work is the expensive part. Wi-Fi and servers for 4 to 5 years; workstations in 4- to 5-year cycles.
Does moving everything to the cloud solve the problem?
It solves server sizing, but increases dependence on the local network and the internet. Without reliable Wi-Fi and a redundant connection, you swap one problem for another.
How much does it cost to modernise the network of a 50-person office?
It depends on the existing cabling and what can be reused. Only an on-site survey gives a firm figure; it almost always comes in below a year of faults and downtime.
If all five symptoms are present, where do you start?
Redundant internet and the rack: they are the quickest and cheapest interventions and the ones that most reduce downtime. Server and Wi-Fi follow, with a proper design.
Next step: a phased growth plan
We carry out the survey, tell you in writing what is at its limit and propose a phased plan with costs broken down — a no-obligation proposal. More than 10 workstations · support contract · remote, on-site and telephone support. Discover DataRoad’s managed IT services or call +351 211 459 950*.
*Call to a Portuguese landline.
