iDRAC vs iLO: Remote Server Management Compared
Every PowerEdge has iDRAC, every ProLiant has iLO — and the brand matters far less than most buyers...
A UPS is the one piece of infrastructure that's judged entirely on a day it's never been tested. It sits in the rack for three years doing nothing visible, and then the power drops — and either your servers shut down cleanly, or they don't.
Two mistakes account for most UPS failures in practice. The first is sizing against the VA number on the box, which is not the number that limits your load. The second is treating the UPS as a one-time purchase rather than a device with a consumable part — because the batteries inside will fail, and in Qatar's climate they'll fail sooner than the datasheet suggests.
This guide covers the VA versus watts distinction properly, how to size and calculate realistic runtime, why battery life is a thermal problem, and what to look for when buying.
VA vs Watts: The Number That Actually Limits You
Every UPS carries two ratings, and they mean different things:
VA (Volt-Amperes)
Apparent power — the headline number in the product name. It's the maximum the inverter can supply before overloading, combining real and reactive power.
Watts
Real power — the actual usable capacity that does work. This is the figure to size against, and it's typically 60–90% of the VA rating.
The ratio between them is the power factor. A 1500VA UPS with a 0.6 power factor delivers only 900W of real power — so a 1000W load will overload it despite the reassuring "1500" on the front panel. Modern units are better: many current server-grade UPS models run a power factor of 0.9 to 1.0, where VA and watts are close or identical.
The rule: check both limits, and the more restrictive one wins. Your total load in watts must be under the UPS watt rating, and your load converted to VA must be under the VA rating. Sizing on VA alone is how people end up with an overloaded UPS.
One helpful nuance for server rooms specifically: modern server power supplies use active power factor correction, giving them a power factor around 0.9–0.99. That means for server hardware, watts and VA are close together — the large gap between the two ratings matters most with older equipment and cheaper consumer UPS units.
How to Size a UPS
Four steps, and the first one is where most people go wrong.
1. Measure real draw, don't use nameplate ratings
A server with a 750W power supply might draw 280W under normal load. Sizing to the PSU rating leads to buying two or three times more UPS than needed — expensive, and as covered in our power supply guide, oversizing has its own downsides. Read actual consumption from iDRAC or iLO, which report real-time power draw, or use a plug-in meter over a few days of normal operation.
2. Add everything the UPS will protect
Servers, storage, switches, firewall, and — easily forgotten — any PoE switch, since its power budget is feeding cameras and access points that also stay up. A 24-port PoE switch delivering 200W to cameras is a 200W+ load on your UPS, not the switch's idle draw.
3. Add 25–30% headroom
For inrush current at startup, future expansion, and the simple fact that a UPS running at its limit has almost no runtime. A measured 650W load means sizing for roughly 810–845W minimum.
4. Check both ratings on the candidate unit
Confirm the watt rating exceeds your headroom figure, and that the VA rating exceeds your load divided by its power factor. Then pick the nearest standard size above that.
Worked example: two servers at 300W each plus a 150W PoE switch and a 50W firewall is 800W measured. Add 25% and you need 1000W of real capacity. A 1500VA/900W unit is too small; a 2200VA/1980W or 3000VA/2700W unit gives comfortable room — and considerably better runtime, since it'll be running at well under half load.
Runtime: Why It's Shorter Than You Think
The single most common surprise: a UPS at full load typically delivers only a few minutes. Runtime is not linear with load — halving the load more than doubles the runtime, because batteries discharge less efficiently at high current draw.
Indicative figures for a typical 1500VA/900W unit show the shape of the curve clearly:
| Load | % of Capacity | Approximate Runtime |
|---|---|---|
| 900W | 100% | ~3 minutes |
| 450W | 50% | ~8 minutes |
| 270W | 30% | ~16 minutes |
| 90W | 10% | ~55 minutes |
Figures vary by model, but the pattern holds universally. Always check the manufacturer's runtime curve for the specific unit rather than assuming.
Decide what runtime is actually for. Most business UPS deployments don't need to ride out a long outage — they need enough time to either bridge a generator starting up, or trigger a clean, automated shutdown of every server. Five to ten minutes at your real load is usually sufficient for the second, and far cheaper than chasing an hour of runtime.
That second point matters more than runtime length: a UPS that keeps servers alive but doesn't shut them down gracefully before the battery dies has only delayed the unclean shutdown. Install the vendor's shutdown agent (or configure NUT on Linux), connect it via USB or network card, and test that servers actually power down when the battery reaches a threshold.
UPS Types: Which Topology for Servers
| Type | How It Works | Suitable For |
|---|---|---|
| Standby (offline) | Passes mains through, switches to battery on failure | Desktops only — not servers |
| Line-interactive | Adds automatic voltage regulation, correcting sags and surges without using battery | Most small business server rooms — the practical default |
| Online (double-conversion) | Continuously converts AC→DC→AC, giving perfectly clean output and zero transfer time | Critical systems, poor mains quality, larger deployments |
One specification worth confirming regardless of topology: pure sine wave output. Server power supplies with active PFC can behave unpredictably on the stepped or simulated sine wave output that cheaper units produce — including refusing to transfer to battery at all. For anything with a server in it, pure sine wave is not optional.
Battery Replacement: The Part People Forget
The batteries inside a UPS are a consumable. Sealed lead-acid (VRLA) batteries — still the most common in this class — have a typical service life of three to five years under good conditions. Lithium-ion units cost more upfront but last considerably longer, often eight to ten years, which changes the total cost calculation over a decade.
The failure mode is what catches people out: batteries don't stop working suddenly, they lose capacity gradually. A UPS with four-year-old batteries may still report "online" and pass a self-test, while actually delivering two minutes of runtime instead of fifteen. Nothing appears wrong until the outage arrives.
- Diary the replacement date — treat it as scheduled maintenance from the install date, not something to react to
- Run periodic runtime tests — a self-test confirms the battery exists; an actual load test confirms how long it lasts
- Prefer hot-swappable battery cartridges — most rack-mount UPS units allow battery replacement without powering down protected equipment, which turns a disruptive job into a five-minute one
- Replace the full set together — mixing old and new cells in a string shortens the life of the new ones
Heat and Battery Life: The Qatar Factor
This is the section that matters most locally, and it's rarely mentioned in generic UPS guides. Temperature is the primary determinant of lead-acid battery life. The widely used industry rule of thumb: for roughly every 10°C above the rated 25°C, VRLA battery life halves.
The implication for Qatar is direct. A UPS rated for a five-year battery life at 25°C, installed in a comms cupboard that sits at 35°C through the summer, may realistically deliver around half that. In an unventilated cabinet running hotter still, less again.
- Install the UPS in the conditioned space, not the hottest corner of the rack. Heat rises — mounting a UPS at the bottom of a rack is both better for weight distribution and cooler.
- Shorten your replacement interval — plan on three years rather than five for VRLA batteries in a warm room, and verify with runtime testing rather than assuming
- Consider lithium-ion for hot installations — the higher upfront cost is easier to justify when heat is aggressively shortening VRLA life, and lithium tolerates elevated temperatures considerably better
- Remember the UPS adds heat too — it's another device dissipating energy into the room, which feeds back into the cooling load discussed in our Qatar cooling guide
Practical summary for Qatar: buy a UPS with more headroom than you strictly need, install it in the coolest part of a conditioned space, and treat battery replacement as a three-year cycle rather than five unless the room is genuinely well cooled.
What to Look For When Buying
- Rack-mount vs tower — rack-mount fits the same cabinet as the servers and usually offers hot-swap batteries; tower suits an office server without a rack
- Network management card — lets the UPS be monitored remotely and trigger shutdown across multiple servers over the network rather than a single USB link
- Outlet count and type — check you have enough, and that they match your equipment (C13/C14 for most servers)
- Battery availability — confirm replacement cartridges are obtainable locally before buying; an orphaned model with no local battery supply becomes disposable
- Redundant PSU feeds — as noted in our PSU guide, feeding both server power supplies from the same UPS protects against PSU failure but not UPS failure; larger deployments use two UPS units on separate circuits
Unlike servers, a UPS is one component we'd generally advise buying new. The chassis and electronics may be fine second-hand, but the batteries are the part that matters and their remaining life is difficult to verify — and battery age is precisely what determines whether the unit does its job.
Frequently Asked Questions
What is the difference between VA and watts on a UPS?
VA is apparent power — the headline rating. Watts is real power, the usable capacity that does work, typically 60–90% of the VA figure depending on the unit's power factor. Size against watts, and check that your load in VA also stays within the VA rating. The more restrictive of the two limits applies.
How long should a UPS run my servers?
For most businesses, long enough to shut down cleanly — typically five to ten minutes at real load — or long enough to bridge a generator starting. Extended runtime is expensive and rarely the right goal. What matters more is that automated shutdown is configured and tested.
How often should UPS batteries be replaced?
Sealed lead-acid batteries typically last three to five years at a rated 25°C, but life roughly halves for every 10°C above that. In Qatar's warmer installations, plan on a three-year cycle and verify with periodic runtime testing. Lithium-ion units last considerably longer and tolerate heat better.
Do I need a pure sine wave UPS for servers?
Yes. Server power supplies use active power factor correction, which can behave unpredictably on the simulated or stepped sine wave output of cheaper units — sometimes refusing to transfer to battery at all. Pure sine wave output is essential for server hardware.
Should I buy a line-interactive or online UPS?
Line-interactive is the practical default for most small business server rooms, adding voltage regulation without using battery. Online double-conversion suits critical systems, sites with poor mains quality, or larger deployments where zero transfer time matters.
Where can I buy a UPS for servers in Qatar?
ServerDove can advise on UPS sizing for your server, storage, and networking load and supply suitable units with delivery in Doha and nationwide. Send us your equipment list and we'll calculate the required capacity and realistic runtime.
Final Thoughts
Three things separate a UPS that works from one that disappoints. Size against watts using measured draw rather than nameplate ratings. Aim for enough runtime to shut down cleanly rather than chasing hours, and actually configure and test that shutdown. And treat the batteries as a scheduled consumable, not a permanent part of the device.
In Qatar, that third point carries more weight than anywhere temperate. Heat is what kills UPS batteries, and a unit installed in a warm cupboard will quietly lose most of its runtime years before anyone checks.
Need help sizing a UPS? Send us your server, storage, and network equipment list and we'll calculate the capacity and realistic runtime you need — with delivery across Qatar and local support. Request a Quote · Chat on WhatsApp
