What’s the single biggest mistake that causes servers to fail before they even boot an operating system? It’s ignoring the physical environment—the room, the power, the cooling, and how you mount the hardware. For the Server+ exam, you need to know that a server’s reliability starts long before you press the power button: it starts with choosing the right floor tiles and running the right power circuits. This chapter teaches you how to plan and carry out a site installation so that the hardware survives and performs exactly as designed.
Jump to a section
A simple way to picture Server Installation and Site Preparation
Have you ever watched a professional kitchen being installed? Every single piece of equipment—ovens, fryers, fridges—must be planned before a single cable is run. A restaurant doesn’t just plug a fryer into a random wall socket. It installs dedicated high-amperage circuits, ventilates the hood above the stove, ensures the floor can hold the weight of an ice machine, and leaves space for staff to move. Mess up the power draw, and the breakers trip during the lunch rush. Forget ventilation, and the whole kitchen turns into a sauna. Now step into a data centre. A server rack is exactly that kitchen. Each server is an appliance that needs a specific voltage, a steady flow of cool intake air and hot exhaust removal, and precise physical spacing. A rack unit (U) is like the shelf space in a fridge—you can’t cram a commercial mixer into a home counter slot. Cables are the gas, water, and drainage lines—they must be routed neatly so nothing blocks airflow or becomes a trip hazard. And just as a kitchen needs a fire suppression hood, a server room needs proper grounding and fire protection. Miss one detail—like using a surge protector instead of a dedicated UPS—and your ‘kitchen’ could shut down or catch fire. The rule for both is the same: measure twice, install once.
Server installation does not begin with unboxing metal boxes. It begins with site preparation: the careful evaluation and modification of the physical space where the server (a specialised computer that provides data or services to other computers) will live. A data centre or server room must provide three essentials: stable power, controlled cooling, and physical security. If any of these is missing, the server will either fail outright, crash unpredictably, or degrade over time.
Let’s break down each area.
Power A server’s power supply unit (PSU) converts the mains alternating current (AC) into direct current (DC) the components use. A standard wall socket delivers around 120-240 volts and 10-15 amps. But a single server might draw 5-10 amps at peak, meaning you can only plug one or two into a normal circuit. For a rack of 20 servers, you need dedicated circuits—separate breakers from the building’s main panel. The exam tests two key concepts here: - Redundancy: Many servers have dual PSUs. You plug each into a different power source (two separate circuits from two UPS units) so that if one feed fails, the server keeps running on the other. - Uninterruptible Power Supply (UPS): A battery system that provides clean power for a few minutes during a blackout, allowing graceful shutdown or generator takeover. A UPS also filters surges and sags (brownouts).
Cooling Servers generate enormous heat—a single rack can produce 10-20 kilowatts of heat. Without active cooling, temperatures inside the rack rise quickly, causing components to throttle or fail. The standard approach uses computer room air conditioning (CRAC) units that blow cold air under a raised floor. The cold air enters through perforated tiles in the cold aisle—the front of the rack—and the hot exhaust from the back of the servers is drawn into the return air ducts. This is called a hot/cold aisle containment system.
Rack mounting Servers are seldom placed on shelves. They are mounted in a standard 19-inch wide equipment rack. The vertical space is measured in rack units (U), where 1U = 1.75 inches (44.45 mm). Common server heights are 1U, 2U, and 4U. Mounting requires sliding rails that screw into the rack’s square or threaded holes. The server slides in and is secured with cage nuts or clip nuts. Cable management arms (CMA) attach to the back to organise power and network cables without blocking airflow.
Grounding and bonding Every rack must be connected to a common earth ground to prevent static discharge (ESD) that can destroy electronics. The rack’s frame gets a grounding strap to the building’s grounding busbar. Additionally, technicians wear ESD wrist straps when handling components.
Physical security Servers must be protected from unauthorised access. This means lockable rack doors, server room access control (badge or biometric), and sometimes cage enclosures. The exam includes site survey factors such as:
Floor load rating: raised floors must support 500-1000 pounds per square foot.
Ceiling height: at least 8-10 feet for proper airflow above the rack.
Proximity to elevators and loading docks for heavy equipment delivery.
Why all this matters Before virtualisation and cloud computing, each application had its own physical server. Today, virtualisation consolidates many virtual machines (VMs) onto one physical host, making the host’s reliability even more critical. A single failed power distribution unit (PDU) can bring down hundreds of VMs. SK0-005 expects you to know the sequence: site survey → plan power and cooling → prepare the floor and rack → install and cable all equipment → test redundancy.
Conduct a site survey
Measure the room dimensions, door width, ceiling height, floor load rating, and locate electrical outlets and cooling vents. This determines if the space can physically and electrically support the server hardware.
Plan power and cooling infrastructure
Calculate total power draw (amps per server) and arrange for dedicated circuits, UPS capacity, and air-conditioning. Plan hot/cold aisle layout or ensure adequate airflow around the rack.
Install the equipment rack and grounding
Position the rack on the floor (or raised floor) and level it. Attach a grounding wire from the rack frame to the building’s earth ground. This prevents electrostatic discharge that could damage components.
Mount servers and cable management
Attach sliding rails to the rack at the correct U positions. Insert servers from the front, secure them, and attach cable management arms to the rear. This allows future maintenance without disconnecting cables.
Connect power and network cables
Plug each server’s PSU A into PDU A (circuit 1) and PSU B into PDU B (circuit 2). Run network cables to the patch panel using Velcro ties. Label both ends of every cable for easy troubleshooting.
Test redundancy and environment
Power on one server, then simulate a failure (unplug one PSU) to verify the other PSU takes over. Check that intake air temperature stays within manufacturer specs (typically 18-27°C).
A small business, FastCart e-commerce, is moving to a new office and must install its own server room. The IT manager, Chris, has no prior experience with physical infrastructure. Here’s how Chris applies site preparation and rack mounting step by step:
Step 1: Site survey. Chris visits the empty room. It’s a 10x12 foot space with a concrete floor and a standard drop ceiling. Using a tape measure, Chris checks the door width (36 inches) to ensure a 42U rack can be wheeled in. The electrical panel is in the basement—two dedicated circuits are available. Chris orders an electrician to install two additional 30-amp circuits for the rack.
Step 2: Ordering the rack and equipment. Chris selects an open-frame 42U rack (no sides to allow better airflow). The rack comes with mounting rails, cage nuts, and a grounding kit. Chris also orders a wall-mounted UPS rated at 3000 VA (volt-amps) and two PDUs (power distribution units) that plug into separate circuits.
Step 3: Installing the rack. Once the rack arrives, Chris places it on the concrete floor (no raised floor here). Using the included bubble level, Chris ensures the rack is perfectly vertical—if it tilts, servers may slide out or strain the rails. Then Chris attaches the ground wire from the rack’s frame to the building’s grounding rod.
Step 4: Mounting servers. Chris has three servers: two 2U database servers and one 4U storage server. The heaviest (the storage unit) goes at the bottom of the rack (position 1-4U) to keep the centre of gravity low. Chris installs the sliding rails for each server, locks the cage nuts into the rack posts, slides the server in, and screws it down. Cable management arms are attached to the back of each server.
Step 5: Cabling. Chris runs power cables from each server’s PSU to the PDUs: server PSU A to PDU A (circuit 1), PSU B to PDU B (circuit 2). The PDUs are plugged into the UPS, and the UPS is connected to the two dedicated circuits. Network cables (Cat6) are routed from each server’s NIC to a patch panel at the top of the rack. Chris uses Velcro strips to bundle cables neatly along the cable management arms.
Step 6: Testing. Chris powers on the UPS, then one server. The server POSTs (power-on self-test). Chris shuts it off and pulls the plug on PDU A. The second PSU immediately takes over—no interruption. Chris then loads a heat gun into the rack’s cold aisle to verify that the CRAC unit can maintain 20°C. Everything works.
What could go wrong? If Chris had skipped the floor load check and the rack weighed 800 pounds—the concrete floor would be fine, but if there were a raised floor, it could collapse. If Chris used a single PDU instead of two, a tripped breaker would take down all servers. If Chris mounted the heavy server at the top, the rack could tip forward when opened.
The real world is all about planning for these exact failure points.
The SK0-005 exam objective 4.1 covers ‘Plan and perform server installation including site preparation and rack mounting.’ The questions will ask you to identify the correct sequence of steps, the correct equipment for a given scenario, and the potential consequences of skipping a step. Here’s exactly what to expect.
Question types: - Scenario-based: “An administrator is installing a new server in a data centre. The server has two power supplies with separate breakers. What is this called?” Answer: redundant power. - Definition match: “Which unit of measurement is used for server height in a rack?” Answer: U (rack unit). - Best practice selection: “Which of the following should be installed at the bottom of a server rack?” Answer: the heaviest equipment. - Trap detection: “A technician wants to install a server on a standard 15-amp wall outlet. The server draws 12 amps. Is this safe?” Trap: yes, it is within 80% of the breaker rating, but the wall outlet is not a dedicated circuit—other devices may share it.
Key concepts tested: - Rack unit definition: 1U = 1.75 inches. - Hot/cold aisle: cold air enters front (intake), hot air exits rear (exhaust). - Dual power supply configuration: two PSUs, two separate circuits for N+1 redundancy. - UPS function: provides surge protection, voltage regulation, and temporary battery power. - PDU types: basic (just outlets), metered (displays load), switched (remotely controlled). - ESD precautions: wrist strap, grounding mat, connection to rack ground. - Cable management: use cable arms, Velcro ties (not zip ties that can cinch too tight), maintain airflow.
Common traps: - Confusing PDU with UPS. A PDU distributes power; a UPS provides backup and conditioning. - Mixing up rack-mount server sizes (1U, 2U) with tower servers that sit on a shelf. - Assuming all racks are 19-inch wide—some are 23-inch for telecom. - Forgetting to account for cable bend radius in structured cabling. - Thinking any extension cord is suitable for a server—must use heavy-duty PDU with proper gauge wiring.
What you must memorise: - The site survey checklist: power capacity, cooling capacity, floor load, ceiling height, door width, grounding, security access. - The order of operations: survey, plan, prepare power, prepare cooling, install rack, mount servers, cable, test. - Rack installation best practices: heavy bottom, cable management arms for sliding servers, front-to-back airflow direction.
Exam-style question pattern: “A technician is installing a rack of servers. Which of the following is the most important factor when determining where to place the heaviest server?” Answer: centre of gravity and rack stability.
Review the CompTIA Server+ official objectives and note that physical installation also includes proper labelling of power and network cables—not tested heavily but implied in best practices.
Always perform a site survey to confirm floor load, power capacity, cooling capacity, and door clearance before ordering any equipment.
Rack unit (U) equals 1.75 inches; a 2U server occupies 3.5 inches of vertical rack space.
Mount the heaviest server at the bottom of the rack to maintain a low centre of gravity and prevent tipping.
Connect each of a server’s dual power supplies to separate circuits from different UPS units for true N+1 redundancy.
Cable management arms (CMAs) ensure that sliding a server out of the rack does not disconnect power or network cables.
ESD wrist straps must be connected to the rack’s grounding point, not just clipped to a table, to safely dissipate static charge.
These come up on the exam all the time. Here's how to tell them apart.
Upright Tower Server
Sits on a shelf or floor; uses no rack space.
Cooling is internal with side vents; not designed for hot/cold aisles.
Cable management is less critical because it is not stacked with other units.
Rack-Mount Server
Fits into a standard 19-inch rack using U spacing.
Cooling relies on front-to-back airflow; must use hot/cold aisle layout.
Requires cable management arms to maintain airflow and support sliding.
Single Power Supply
One PSU; failure means server shuts down immediately.
Requires only one power cable and circuit.
Lower cost and simpler installation.
Dual (Redundant) Power Supplies
Two PSUs; one can fail and server keeps running.
Must connect each PSU to a separate circuit for true redundancy.
Increases cost and power cabling complexity.
Basic PDU
Only distributes power; no monitoring or control.
Cannot remotely turn off a port or measure load.
Cheapest option; adequate for small deployments.
Switched/Metered PDU
Monitors per-outlet power draw; can alert on high load.
Allows remote power cycling of individual outlets.
More expensive but essential for data centre management.
Cold Aisle Containment
Cold air supplied directly to server intakes; hot air returns freely.
Typically uses raised floor with perforated tiles in the cold aisle.
Worker comfort may be lower because cold aisle is colder.
Hot Aisle Containment
Hot exhaust air is contained and fed directly back to cooling unit.
Often uses overhead ducts or ceiling return plenum.
Improves cooling efficiency and can reduce overall energy use.
Mistake
Any standard wall outlet can power a server, as long as the plug fits.
Correct
Servers need dedicated circuits with the correct voltage and amperage. A shared 15-amp outlet can easily trip a breaker when other equipment draws power.
This mistake is common because people treat servers like desktop PCs, which typically run on shared circuits with low power draw.
Mistake
Rack mounting is optional—servers can just sit on a shelf or the floor.
Correct
Servers must be mounted in a rack with proper rails for airflow, cable management, and stability. Sitting on a floor blocks ventilation and can cause overheating.
Beginners often think of servers as big desktop towers and don't realise the cooling design relies on rack-mount orientation (front intake, rear exhaust).
Mistake
All server racks are the same depth, so any server fits any rack.
Correct
Rack depth varies from 24 inches to 42 inches. Server depth also varies. You must match the rail depth to the server and rack depth (adjustable rails help).
People assume 'standard rack' means uniform depth, ignoring that different vendors have different chassis depths.
Mistake
Cable management is just for neatness and has no effect on server performance.
Correct
Poor cable management blocks airflow, especially in the rear exhaust path, raising temperatures and causing thermal throttling or failure.
Since cables are 'invisible' to the operating system, beginners don't connect them to hardware reliability.
Mistake
A UPS is just a big battery that keeps the server running during a power cut.
Correct
A UPS also conditions power: it filters out surges, sags, and electrical noise that can damage components over time. Its battery runtime is typically minutes, not hours.
Movies and consumer advertising often show UPS as 'backup power for a whole house', leading to false expectations.
Reveal each answer, then mark whether you got it right. Score 60%+ to unlock the next chapter.
Only if the desk can bear the weight and the room provides proper cooling and power. However, for reliability and security, servers should always be mounted in a rack in a dedicated environment.
A rack unit (U) is a standard height measurement for equipment racks. 1U equals 1.75 inches. A server that is 2U high takes up 3.5 inches of vertical space in the rack.
Yes. A generator takes seconds to start and stabilise. A UPS provides instant, uninterrupted power during that gap, as well as surge protection and voltage regulation.
Connect a thick copper grounding wire from a bolt on the rack’s frame to the building’s main earth ground busbar. Also, ensure any metal parts of the rack are bonded together with a grounding kit.
Approximately 1 metre (3 feet) in the front and back for airflow and maintenance access. The sides may be tighter if using an open-frame rack, but still need space for cable management.
Consumer-grade power strips are not rated for the high continuous load of servers and lack surge protection calibrated for data centre equipment. You need a PDU (power distribution unit) with proper amperage and circuit breakers.
You've finished Server Installation and Site Preparation. Continue through the SK0-005 study guide to build a complete picture of the exam.
Done with this chapter?