Frequently Asked Questions

A server rack is a standardised 19-inch frame that holds IT equipment in units of 1.75 inches, written as 1U. Height is quoted in U (24U, 42U, 45U, 48U), width in millimetres at the outside of the cabinet (600 mm or 800 mm), and depth by the mounting-rail span (1000 mm to 1200 mm). The 800 mm width buys side channels for cable management and PDUs; the extra depth buys room behind deep servers for power and network tails. Anything mounted in it uses the same rail spacing regardless of manufacturer, which is why a mixed-brand build works.

A UPS is a device that keeps a load powered when mains voltage fails or drifts outside tolerance, using stored energy in batteries until either the mains returns or a generator takes over. In a data centre it does two separate jobs: it bridges the gap before the generator picks up, typically five to fifteen seconds, and it conditions the incoming supply so that sags, surges and frequency drift never reach the equipment. Capacity is quoted in kVA and kW; the kW figure is the one that matters, because modern server power supplies run close to unity power factor. Runtime is a battery decision, not a UPS decision.

Network attached storage is a storage system that presents shared folders over the existing Ethernet network using file protocols such as SMB and NFS. Clients see directories and files, not raw disks, so permissions, snapshots and quotas are managed by the NAS itself. It suits file shares, media, backups and departmental data where several machines need the same set of files. Because it rides on standard networking, it is the cheapest storage tier to deploy, and the network becomes its performance ceiling.

NAS serves files over Ethernet; a storage area network serves blocks over a dedicated fabric, usually Fibre Channel or iSCSI. A SAN volume appears to a server as a local disk, so the server formats it and owns the filesystem, which is what databases and virtualisation clusters need for shared, low-latency access. NAS is simpler and cheaper because it uses the network you already have; SAN is faster and more deterministic because the fabric is separate from user traffic. Many sites run both and place workloads by requirement rather than by preference.

An ATS is a switch that monitors two power sources and moves the load to the healthy one automatically when the primary fails. In a rack it is a 1U device with two inlets and a bank of outlets, sized between 10 A and 32 A, and it exists so that single-corded equipment can still be fed from two independent supplies. At building scale the same function is a panel-mounted switch rated from tens of amps up to 3200 A, sitting between the mains and a generator. Transfer time is the specification that matters: rack units switch in milliseconds, panel switches in seconds.

Runtime depends on how much of the UPS capacity you actually draw, not on the kVA on the label. A UPS loaded to half its rating runs considerably more than twice as long as one loaded to its full rating, because battery discharge is non-linear. Typical rack installations are specified for five to fifteen minutes at full load, which is enough to bridge to a generator or to shut systems down cleanly. If you need an hour, you are buying a battery cabinet, and the floor loading and room ventilation become part of the specification.

A rack server is a server built to a fixed U height so that it mounts on the rails of a 19-inch cabinet, most commonly 1U, 2U or 4U. Height decides what fits inside: 1U takes low-profile cards and small fans, 2U takes full-height GPUs and more drive bays, 4U takes the largest expansion and cooling. Compared with a tower it is denser, easier to cable in bulk and designed for front-to-back airflow that matches the cabinet. Current models are built on Intel Xeon or AMD EPYC platforms.

A rack PDU is the strip inside a cabinet that takes one or more feeds and distributes them to the outlets that equipment plugs into. It is specified by input current (16 A, 32 A and 63 A are the usual points), by phase (single or three), and by outlet mix, typically C13 and C19. Format is either 0U, mounted vertically in the side channel and taking no rack height, or 1U mounted horizontally. Every rack has at least one; resilient racks have two, on separate feeds.

A basic PDU only distributes power. A metered PDU adds current measurement, either for the whole strip or per outlet, so you can see how loaded a rack really is before you add another server. A monitored PDU puts that data on the network with SNMP or a web interface and raises alarms on thresholds. A switched PDU adds relays so individual outlets can be turned off and on remotely, which is how you power-cycle a frozen machine in a site you cannot reach.

0U means the PDU mounts vertically in the cabinet’s rear side channel and consumes no usable rack height. A 42U cabinet with two 0U PDUs still offers 42U for equipment, which is why the format dominates in dense installations. The trade-off is that 0U strips are long, so the cabinet must have the matching mounting brackets and enough rear depth to keep the outlets clear of server tails. A 1U PDU is easier to retrofit but costs you a slot.

Standard cabinets are 600 mm or 800 mm wide, 1000 mm to 1200 mm deep, and between 24U and 52U tall, with 42U and 47U the most common in data centres. Internal usable depth is always less than the external figure, because doors, rail offset and cable management take space at both ends. The 19-inch mounting width is fixed by the standard and does not change with the cabinet width; the extra 200 mm on an 800 mm cabinet becomes side channels for cabling and vertical PDUs. Weight rating matters as much as size, typically 1000 kg to 1500 kg static.

An online UPS converts incoming AC to DC and back to AC continuously, so the load is always fed from the inverter and never directly from the mains. Because the output is regenerated rather than filtered, voltage and frequency are held steady no matter what the incoming supply does, and there is no transfer time at all when the mains fails. It costs more and runs slightly less efficiently than line-interactive designs, which is why it is used where the load justifies it. Data centre, medical and industrial loads are normally specified online.

A modular UPS is built from identical power modules in a common frame, so capacity is added by inserting modules rather than by replacing the unit. It lets a site buy the capacity it needs now and grow to the frame’s maximum later, and it makes redundancy cheap: N+1 is one extra module, not a second UPS. Modules are hot-swappable in most current designs, so a failed module is replaced without dropping the load. The frame rating, not the module count, is what caps the installation.

A three-phase UPS takes a three-phase supply and, in most data centre designs, delivers three-phase output to the distribution board. The practical threshold is load: single-phase makes sense to roughly 20 kVA, and above that the current on one phase becomes impractical to distribute and balance. Three-phase also matches how buildings and generators are supplied, so the transfer arrangement stays simple. Below the threshold a single-phase unit is cheaper and easier to install.

A static transfer switch uses semiconductors rather than contacts, so it moves the load between two synchronised sources in a few milliseconds, fast enough that connected equipment never notices. A conventional ATS uses mechanical contacts and takes longer, which is acceptable when a UPS or generator covers the gap. STS units are used downstream of two UPS systems to give single-corded equipment genuine dual-path power. The two are not interchangeable: the choice follows how much interruption the load will tolerate.

Size the switch on the full-load current of the supply it carries, not on the generator’s nameplate kVA, and then check it against the utility breaker upstream. Motor and UPS loads draw significant inrush, so the switch must be rated for the starting current as well as the running current. Also confirm the number of poles: a four-pole switch that breaks the neutral is required where the generator neutral is separately earthed. Panel-type switches are available from tens of amps to 3200 A, so the constraint is specification accuracy, not availability.

An industrial UPS is built for wider temperature ranges, higher fault currents and more electrical noise than an IT-room unit, and it is usually specified for a service life measured in decades rather than years. Enclosures are heavier, boards are conformally coated, and DC-bus architectures are common because the plant already runs on a battery system. Serviceability is designed in: components are replaceable in place rather than as a sealed assembly. The trade-off is size, weight and cost against IT-grade equivalents at the same kVA.

Add the actual power draw of everything in the rack, not the nameplate rating of the power supplies, because a server rarely draws more than a fraction of its supply rating. Add headroom for growth, then check the result against the UPS kW rating rather than the kVA rating. Decide the runtime separately, because that is what fixes the battery count and the cabinet footprint. Finally confirm the input feed: a 32 A single-phase circuit will not carry a UPS you have sized to 30 kW.

An all-flash array uses solid-state media throughout, so latency is consistently low and performance does not depend on where the data sits. A hybrid array mixes flash and spinning disks and moves hot data to the flash tier, which makes capacity much cheaper per terabyte but leaves performance dependent on how well the working set fits the tier. All-flash suits databases, virtual desktops and anything latency-sensitive; hybrid suits archives, backups and large sequential workloads. Data reduction narrows the price gap considerably on the right dataset.

A typical enterprise rack is provisioned at 3 kW to 7 kW, a dense virtualisation rack at 10 kW to 15 kW, and an AI or HPC rack well beyond that. The limit is almost never the cabinet: it is the feed available to the rack and the cooling available to remove the heat, since every watt drawn becomes a watt to be extracted. Above roughly 15 kW per rack, air containment stops being optional, and beyond that liquid cooling enters the conversation. Size the PDU, the UPS and the cooling to the same number.

No. DATACENTERPORT is an independent supplier. We source through authorised distribution channels and we say so plainly rather than implying a factory appointment we do not hold. What you get instead is a comparison across manufacturers and one point of contact for a mixed-brand order.

Within 48 working hours for a defined requirement: product group, capacity and delivery country. If the requirement needs a sizing study or a non-standard configuration, we tell you inside those 48 hours how long the full quote will take, rather than letting it drift.

Iraq, the UAE, Saudi Arabia, Qatar and Azerbaijan are the regular destinations, and the team’s export history covers more than 30 countries on five continents. We can quote elsewhere; the honest constraint is freight cost and lead time, not willingness.

EXW, FCA, CPT and DAP are the usual ones, with FOB and CIF for sea freight. Tell us which term your procurement process expects and the quote comes back on that basis, with the cost split visible rather than buried.

Terms are agreed per order. Advance payment, letter of credit and cash against documents are all workable; which one applies depends on the destination, the order value and whether we have traded before. Bank details are only ever sent on a signed proforma, never in a plain e-mail.

Stocked rack accessories, PDUs and small ATS units typically ship in days. Cabinets and standard three-phase UPS are usually weeks. Large modular UPS, high-end storage and configured servers are manufactured to order and can run to a couple of months. The quote carries the current figure, not a generic one.

No fixed minimum. A single rack ATS is a real order. Below a certain value freight starts to dominate the total, and we will say so and suggest consolidating with a later shipment if that saves you money.

The manufacturer’s standard warranty applies, typically one to three years depending on the product line, and extensions are quotable. On-site response times vary by country. Tell us the delivery country and we confirm what is actually offered there rather than what the global datasheet says.

Supply is our core business. Cabinet assembly, UPS commissioning and site installation are handled case by case; if you want them in scope, say so with the requirement and they are quoted as separate line items rather than assumed.

Because in this category a published figure never survives contact with the configuration. One UPS family has a different order code and a different price for each voltage, battery and outlet combination. A number on a web page would only be wrong by the time you read it.