Uptime Institute Annual Outage Analysis 2021 (3rd annual) - press release

Overview

This document is the press release for the third annual outage analysis that Uptime Institute published in April 2021. It offers the public summary of the findings rather than the full report. The release aims to show the industry how common data centre and IT service outages are and which causes stand out. Its value for designers and operators is that it gives a benchmark built from real operator experience rather than from assumed risk.

The release reports that most operators had at least one outage in the past three years. Three quarters experienced an IT service outage, and three in ten had one with significant impact. One in six operators reported a serious or severe outage in the same period. These proportions show that a high-availability target depends on the whole chain of equipment, processes and providers, not only on the choice of individual equipment.

Third-party providers form a separate finding: 56% of those using third-party data services had a moderate or serious outage caused by the provider. According to the release, networking and configuration issues are rising, while power failures are becoming somewhat less dominant than before. Power outages have historically been caused by failures in UPSs, transfer switches and generators, which remain the main hardware contributors named.

The human-factor figures come from the 2021 resiliency survey. Of operators, 42% had an outage caused by human error in the past three years. Of those, 57% cited staff failing to follow procedures or execute them correctly, and 44% cited incorrect procedures. Because this is a press release, it gives no methodological detail, sample size or sector breakdown, so the results are best read as directional and not as direct numerical inputs to a specific facility decision.

Key points

1. Three quarters of operators had an IT service outage in the past three years; three in ten had one with significant impact.
2. One in six operators reported a serious or severe outage in the past three years.
3. 56% of those using third-party data services had a moderate or serious outage caused by the provider.
4. Power outages have historically been caused by failures in UPSs, transfer switches and generators.
5. Networking and configuration issues are rising, while power failures are becoming somewhat less dominant.
6. 42% of operators had a human-error outage in the past three years.
7. Of human-error outages, 57% cited staff execution and 44% cited incorrect procedures.
8. The document is a press release, not the full report, and gives no method or breakdown detail.

Using it in design

• When building a failure-mode catalogue, list the UPS, ATS/STS and generator as the main power-outage contributors.
• Add procedure-quality warnings to operations checklists to address incorrect procedures and execution errors.
• Treat dependence on third-party hosting or data services as a separate outage source in risk analysis.
• In redundancy decisions, consider network and configuration outages as well as the power chain.

Who should read it

Data centre designers, operations managers and decision makers carrying out risk assessments.

Publication details

Authors

Uptime Institute

Publisher

Uptime Institute

Year

2021

Document language

English

Source type

Industry body

Access

Free to read

Licence

Copyright retained; citation only — the publisher’s terms apply to the original document

Citation