When evaluating colocation facilities for your servers, you will encounter the term "Tier" attached to a number — Tier 1, Tier 2, Tier 3, Tier 4. These tier classifications come from the Uptime Institute, a global data center advisory organization, and they describe increasing levels of infrastructure redundancy, availability, and fault tolerance. Understanding what these tiers actually mean — in concrete, operational terms rather than marketing language — is essential for making an informed colocation decision.
This guide explains each tier in plain English: what infrastructure it requires, what uptime it guarantees, how much downtime you can expect per year, and which tier makes sense for most SMBs considering colocation.
The Uptime Institute Tier Standard
The Uptime Institute's Tier Standard is the most widely recognized data center classification framework globally. It was originally published in 1995 and has been updated since. There are four tiers, each building on the requirements of the previous. A data center can self-certify as a given tier through design documentation review, or can obtain formal third-party Tier Certification from the Uptime Institute through an audit of design documents and on-site operational processes.
It is important to understand that "Tier III" in a marketing brochure does not necessarily mean the facility has received formal Uptime Institute Tier III Certification — it may only mean the facility was designed to Tier III specifications. Formal certification is a different and more rigorous process. When evaluating facilities, ask specifically whether they hold Uptime Institute Tier Certification (not just "Tier III design") and for which tier.
Tier I: Basic Infrastructure
Uptime guarantee: 99.671% — 28.8 hours of downtime per year
Tier I is the minimum baseline for a dedicated data center environment. It includes: a dedicated space for IT systems, a UPS system on the power path (to handle utility power fluctuations and brief outages), dedicated precision cooling equipment, and a standby power generator. However, all of these systems exist as single paths — there is one power distribution path to the IT equipment, one UPS, one cooling path. Maintenance of any of these components requires shutting down the IT equipment they support.
Tier I facilities are acceptable for non-critical workloads or test environments, but are inappropriate for production business systems where more than a couple of hours of downtime per month is unacceptable. With nearly 29 hours of expected downtime per year, Tier I facilities will experience multiple outages affecting your equipment during planned maintenance windows for their infrastructure components.
Tier II: Redundant Capacity Components
Uptime guarantee: 99.741% — 22 hours of downtime per year
Tier II adds redundant capacity components to the single power and cooling paths established in Tier I. This means: N+1 UPS configuration (one more UPS module than the minimum required to carry the load), additional cooling units beyond the minimum required, and additional generator capacity. However, Tier II still uses a single non-redundant distribution path — power gets from the redundant UPS systems to your equipment via one path, not multiple independent paths.
The practical implication is that while Tier II can handle the failure of a UPS module or cooling unit without downtime (because there are redundant components), planned maintenance on the distribution path — the switchgear, the PDUs, the piping — still requires taking down the IT equipment. Tier II is a meaningful improvement over Tier I but still not appropriate for systems with demanding uptime requirements.
Tier III: Concurrently Maintainable
Uptime guarantee: 99.982% — 1.6 hours of downtime per year
Tier III is where data center infrastructure becomes genuinely enterprise-grade. The defining characteristic of Tier III is concurrent maintainability: every component of the power and cooling infrastructure can be taken offline for maintenance, replacement, or testing without affecting the IT equipment it supports. This is achieved through fully redundant distribution paths — two independent power paths from utility to UPS to PDU to equipment, multiple cooling loops — so that any one path can be isolated while the alternate path continues to carry the load.
Tier III facilities have N+1 UPS, N+1 cooling capacity, dual power feeds from the utility where possible, and fully redundant mechanical distribution. Maintenance windows do not require downtime for IT equipment. With only 1.6 hours of expected downtime per year, Tier III facilities are appropriate for production business systems, and the vast majority of quality colocation facilities serving SMBs and enterprises are built to Tier III specifications.
For most SMBs: Tier III is the target. The jump from Tier II to Tier III in uptime (22 hours per year vs. 1.6 hours per year of expected downtime) is dramatic and justifies the modest cost premium. For most colocation deployments in the Southern California market, Tier III facilities are readily available at reasonable price points.
Tier IV: Fault Tolerant
Uptime guarantee: 99.995% — 0.4 hours (26 minutes) of downtime per year
Tier IV adds fault tolerance on top of Tier III's concurrent maintainability. Where Tier III can handle maintenance without downtime, Tier IV can handle an unplanned failure of any single component without causing downtime. This requires 2N redundancy (completely independent, active parallel systems for every component) rather than N+1, independent distribution paths that are physically separated, automatic transfer capabilities that do not require operator intervention, and fuel storage requirements to sustain generator operation for extended periods.
Tier IV facilities are the most expensive to build and operate. The capital cost to build Tier IV infrastructure is roughly 2–3x that of Tier III. Rack rental rates reflect this. Tier IV is the right choice for financial institutions, exchanges, healthcare systems managing critical patient data in real time, government systems, and other environments where even 26 minutes of expected annual downtime is unacceptable. For the overwhelming majority of SMBs, Tier IV is unnecessary and the cost premium is not justified by the marginal improvement in availability.
Quick Reference: Tier Summary
| Tier | Uptime | Max Downtime/Year | Key Feature |
|---|---|---|---|
| Tier I | 99.671% | 28.8 hours | Basic redundancy, single path |
| Tier II | 99.741% | 22 hours | Redundant capacity components |
| Tier III | 99.982% | 1.6 hours | Concurrent maintainability |
| Tier IV | 99.995% | 26 minutes | Full fault tolerance (2N) |
What IT Center Colocation Partners Offer
IT Center's colocation and data center solutions leverage Tier III facilities in the Southern California market. Our colo partners maintain formal Uptime Institute Tier III Certification or equivalent documentation, redundant utility power feeds, N+1 UPS with battery and generator backup, precision cooling with N+1 redundancy, 24/7 physical security with biometric access controls, and high-bandwidth internet connectivity with diverse carrier options.
For most SMB clients, the combination of a Tier III colocation environment with IT Center-managed server infrastructure and our enterprise password management backup solution provides a level of reliability and data protection that exceeds what any in-house server room can deliver — at a predictable monthly cost. The facility handles the physical infrastructure reliability; we handle the systems on top. See related reading: Colocation vs. Cloud Hosting for when colo makes sense versus cloud alternatives.
Ready to Move Your Servers to a Tier III Facility?
IT Center places and manages client server infrastructure in certified Tier III colocation facilities across Southern California. We handle racking, connectivity, monitoring, and ongoing management.
Learn About Our Data Center ServicesOr call us directly: (888) 221-0098 | [email protected]