UII KEYNOTE REPORT 131M • MARCH 2024

RESILIENCY, RISK AND SECURITY, FINANCE AND STRATEGY,

IT ARCHITECTURES, SYSTEMS, NETWORKS

Annual outage
analysis 2024
The causes and impacts of IT and data center outages

Avoiding digital infrastructure failures remains paramount for data center owners
and operators. This report analyzes recent Uptime Institute data on IT and data center
outage trends: their causes, costs and consequences.

Authors
Douglas Donnellan, Research Analyst
Andy Lawrence, Executive Director of Research

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 1

KEYNOTE REPORT: Annual outage analysis 2024

Key findings
• Data relating to outages should be treated • Power issues are consistently the most
skeptically. All methodologies used to common cause of serious and severe data
track the frequency, severity and costs of center outages. However, network-related
outages are subject to uncertainty, partly issues are the largest single cause of IT
because of a lack of transparency and service outages.
reliable reporting mechanisms.
• Four in five respondents to the 2023
• Uptime research suggests that overall Uptime Institute data center survey say
outage frequency and severity continue to that their most recent serious outage
decline. However, cyber-related incidents could have been prevented with better
— which are responsible for many of the management, processes and configuration.
most severe outages — are on the rise, This suggests that, as in previous years,
causing extensive and serious disruption. there is an opportunity to reduce outages
through training and process review.
• Outages are costly. More than half (54%)
of the respondents to the 2023 Uptime • Uptime data suggests that each year there
Institute data center survey say their are, on average, 10 to 20 high-profile IT
most recent significant, serious or severe outages or data center events globally
outage cost more than $100,000, with 16% that cause serious or severe financial
saying that their most recent outage cost loss, business and customer disruption,
more than $1 million. reputational loss and, in extreme cases,
loss of life.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 2

KEYNOTE REPORT: Annual outage analysis 2024

UII KEYNOTE REPORT 131 • MARCH 2024

RESILIENCY, RISK AND SECURITY, FINANCE AND STRATEGY,
IT ARCHITECTURES, SYSTEMS, NETWORKS

Annual outage
analysis 2024
The causes and impacts of IT and data center outages
Contents
Introduction ———————————————————————————————————————————— 5
Avoiding digital infrastructure failures remains paramount for data center owners
and operators. This report analyzes recent Uptime Institute data on IT and data center
outage trends: their causes, costs and consequences.

Authors
Douglas Donnellan, Research Analyst
Andy Lawrence, Executive Director of Research

Outage frequency and severity———————————————————————————————— 7

Outage causes—————————————————————————————————————————— 11
Cloud and third-party provider reliability——————————————————————————— 15
© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 1

Power outages————————————————————————————————————————— 17
Networking outages——————————————————————————————————————— 18
System and software outages————————————————————————————————— 20
The human factor———————————————————————————————————————— 21
Cost of outages————————————————————————————————————————— 25
Summary————————————————————————————————————————————— 26
Appendix: Sources and methodology——————————————————————————— 27

Uptime Institute Intelligence is an independent unit of Uptime Institute dedicated to identifying,

analyzing and explaining the trends, technologies, operational practices and changing business models
of the mission-critical infrastructure industry. For more about Uptime Institute Intelligence, visit
uptimeinstitute.com/ui-intelligence or contact [email protected].

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 3

KEYNOTE REPORT: Annual outage analysis 2024

Contents (continued)
Figures and tables
Table 1 ——————————————————— 5 Figure 9 ———————————————————16
How Uptime Institute tracks outages Most common causes of major
third-party outages
Table 2 ——————————————————— 6
Outage severity rating Figure 10 —————————————————— 17
Most common causes of major
Figure 1 ——————————————————— 7 power-related outages
While majority of operators experienced
an outage, most had negligible impact Figure 11 ——————————————————19
Most common causes of major
Figure 2 ——————————————————— 9 network-related outages
Physical site redundancy still climbing
Figure 12 ——————————————————20
Table 3 ———————————————————10 Most common causes of major IT
Publicly reported outages tracked by system- / software-related outages
Uptime, 2016 to 2023
Figure 13 ——————————————————21
Figure 3 ———————————————————10 Most common causes of major
Proportion of publicly reported outages human error-related outages
that were serious or severe, 2016 to 2023
Figure 14 ——————————————————22
Figure 4 ——————————————————— 11 Most say outages preventable
Power remains the number one root with better management
cause of outages
Figure 15 ——————————————————23
Figure 5 ——————————————————— 12 Durations of publicly reported
Most common causes of IT outages, 2017 to 2023
service-related outages
Table 4 ———————————————————24
Figure 6 ———————————————————13 Ten major outages in 2023 and 2024
Causes of publicly reported outages 2023
Figure 16 ——————————————————25
Figure 7 ———————————————————14 Half of impactful outages cost
Publicly reported outages by sector, more than $100,000
2016 to 2023
Figure 8 ———————————————————15
Most say cloud only resilient enough
for some workloads

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 4

KEYNOTE REPORT: Annual outage analysis 2024

Introduction
This report on digital infrastructure outage trends is the sixth edition in an ongoing
series from Uptime Institute Intelligence that analyzes IT service resiliency. The analysis
is based on data from a variety of sources, including publicly available reports (e.g.,
information reported in news and social media), Uptime surveys (e.g., the Uptime Institute
Global Survey of IT and Data Center Managers 2023 and the Uptime Institute Data Center
Resiliency Survey 2024) and other data aggregated and anonymized from Uptime
members and partners. Each of these data sources has limitations (see Table 1).

Table 1 How Uptime Institute tracks outages

Tracking outages is neither simple nor consistent. Some outages are visible and well
publicized, others remain confidential. Some managers, staff and customers may be aware
of outages, while others in different roles may not. In addition, some major slowdowns or
disruptions may not be classified as outages. Uptime Institute uses multiple means to track
the overall trends and incidents, but none provide a clear picture on their own.
This table shows the methods used by Uptime (see the Appendix for further information).

Accuracy Methodology Limitations

Mainly big outages and interruptions to
News / social media
consumer-facing services
Public reports Poor Outage trackers May lack details

Company statements Sources may be untrusted or poorly informed

Answers may vary according to role and sample

Uptime Institute Fair / good
Industry surveys by
surveys Uptime Institute
All responses anonymous

Uptime Abnormal Good / very Detailed, accurate site / Information primarily facility / site-based
Incident Report good facility-level data
(AIRs) database shared under a non-
disclosure agreement All data anonymous

UPTIME INSTITUTE 2024

In this report we use three primary sources:

• The Uptime Institute Annual Global Data Center Survey 2023. This was conducted
in March and April of 2023, with more than 850 operator respondents. This may be
referred to as Uptime’s annual survey. In most cases, the outages discussed from this
survey refer to data center-specific issues and are termed “data center outages” unless
otherwise stated.
• The Uptime Institute Data Center Resiliency Survey 2024, conducted in January
and February 2024, with responses from more than 400 operators and 450 vendor
respondents. This survey is conducted annually. The outages discussed from this survey
in most cases refer to all IT service outages, including those caused by data center
issues, and are termed “IT services outages” unless otherwise stated.
• Publicly reported / tracked outages monitored by Uptime in 2023. Because this data
is less reliable and based on third-party sources, it is separated and discussed in the
blue-tinted boxes throughout this report.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 5

KEYNOTE REPORT: Annual outage analysis 2024

Throughout this report we discuss and categorize outages according to their obvious
or perceived severity, using terms such as “serious” or “severe”. The way we categorize
outages according to severity is shown in Table 2.

Table 2 Outage severity rating

Category Service outage Impact of outage
1 Negligible Recordable outage but little or no obvious impact on services.

2 Minimal Services disrupted. Minimal effect on users / customers / reputation.

3 Significant Customer / user service disruptions, mostly of limited scope, duration or effect. Minimal
or no financial effect. Some reputational or compliance impact(s).

4 Serious Disruption of service and/or operation. Ramifications include some financial losses,
compliance breaches, reputational damage and possibly safety concerns. Customer
losses possible.

5 Severe Major and damaging disruption of services and / or operations with ramifications
including large financial losses and possibly safety issues, compliance breaches,
customer losses and reputational damage.

UPTIME INSTITUTE 2024

Although the headlines may sometimes suggest otherwise, IT service reliability is

improving. Decades of innovation, investment and improved management have significantly
increased the reliability of critical IT systems, networks and data centers. However,
operators are also facing new challenges from increased demand, the adoption of software-
based optimization techniques and a growing number of cyberthreats. Adverse weather
events, which are increasing in both intensity and frequency due to climate change, and the
use of more renewable energy in the power grid have added further risks.
There are several reasons why major failures appear to be more common. These include the
high levels of critical IT in use, society’s high dependency on IT, and increased visibility through
news and social media. However, the majority of IT service outages are brief and have a
limited impact. This report focuses on the significant, serious and severe outages (categories
3, 4 and 5) that can have serious consequences for businesses and governments.
In 2024, more major failures are likely to be caused by software, systems or configuration
errors than in previous years — a reflection of the growing complexity of the IT, the use
of hybrid IT and associated networking. Meeting the growing demand for digital services
while navigating this complexity requires careful planning and ongoing investment in
resiliency strategies.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 6

KEYNOTE REPORT: Annual outage analysis 2024

Outage frequency and severity

As the global data center footprint expands to meet new demand, the overall number
of data center-related outages is expected to increase. However, Uptime data reveals a
consistent, downward trend in the frequency and severity of outages relative to the overall
growth in IT. This trend has been observed for several years.
More than half (55%) of operator respondents to the 2023 Uptime Institute data center
survey report having an outage in the past three years — down from 60% in 2022 and
69% in 2021. At the same time, only one in 10 outages in 2023 was categorized as either
serious or severe (see Figure 1). This is an improvement of four percentage points from
the 2022 response and an improvement of 10 percentage points compared with 2021.
When analyzing this data, Uptime focuses on data center outages and not all IT service
outages. To collect more precise survey data and to improve accuracy, respondents to
our annual survey are now asked about outages at the specific facility that they are most
familiar with, rather than the largest site within the data center operator’s organization.
This may have led to some shifts in the data — however, our control questions suggest the
effect on longitudinal comparisons has been minor.

Figure 1 While majority of operators experienced an outage, most had negligible impact
On a scale of 1 (negligible) to 5 (severe) how would you classify your data center’s most
impactful outage in the past three years, either in your own facility or because of a third-party
service provider? (n=781)

Severe 4%

Serious 6%
Did you
experience an
45% outage in the 55% Significant 17%
NO past three YES
years?
Minimal 32%

Negligible 41%

(Responses for "Don't know" are not included.)

UPTIME INSTITUTE GLOBAL SURVEY OF IT AND DATA CENTER MANAGERS 2023

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 7

KEYNOTE REPORT: Annual outage analysis 2024

What do these outage rates tell us? The answer is complicated because the technology is
never static and external factors (such as weather or utility reliability) can play a significant
role. Uptime Intelligence can, however, make the following observations:
• Overall decrease. The data collected over several years suggests that outages are
decreasing relative to the overall rise in IT. This is likely due to many different reasons
(see below).
• No complacency. Although the outage frequency has decreased, there is no room
for (or any sign of) complacency. Instead, there is strong consensus across the
sector, including among regulators, that outage rates are still concerning. The high
financial and/or reputational costs, which can result from a data center outage, mean
that resiliency consistently registers as one of the top concerns among industry
stakeholders and is a strong driver of investment.
• Public cloud. The move to the public cloud does not necessarily mean that there will
be fewer outages. However, it may mean that, for example, “third-party supplier” is
registered as the cause of more IT service disruptions while fewer on-premises data
center outages are recorded.
• Long COVID. The COVID-19 pandemic had a significant impact on the data center
industry, particularly in terms of decreasing and then increasing demand, straining
supply chains and distorting outage rates. These aftershocks are still being felt in
2024, even if indirectly, while their longer-term impact remains unclear. For example,
supply chain disruptions continue to stall capital projects, which has led many
organizations to delay maintenance and infrastructure upgrades. It is possible that
these factors have temporarily reduced the rate of incidents, which can sometimes
cause an outage, and that a rebound effect in the near- to medium-term will be seen.
• Grid instability. There is evidence that the global shift toward more transactive, dynamic
and renewable power grids is reducing, or will reduce, grid reliability. If this is the case,
data centers may experience an increase in outages. Many outages occur when an
uninterruptible power supply (UPS) or generator fails to respond to a grid disruption.
• Climate change. Extreme weather events — such as high and low temperatures,
high winds and floods, and forest fires — exacerbated by climate change have been
associated with data center outages over the past few years. This trend is likely to
intensify and will increase the outage risks until pre-emptive action is taken.
• Adoption of new technology. To comply with anticipated and recently passed
regulations around the reporting and improvement of resiliency and energy
performance, operators may adopt technologies and practices that require careful
management. These may even add new risks, for example:
• The use of distributed, software-based resiliency (i.e., moving traffic and workloads
dynamically) can reduce outage risks and their associated impact over time, but
during an introductory period, these may increase.
• The use of liquid cooling may reduce some thermal risks, but the impact of
component failure may reduce thermal ride-through times, which, in some cases,
increases risk.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 8

KEYNOTE REPORT: Annual outage analysis 2024

Despite the increase in risk factors, Uptime’s annual survey data up to 2023 suggests that
the rate of outages per facility is falling. What could be driving this trend? One factor stands
out: Uptime research finds that, year-on-year, most organizations are investing more in
physical infrastructure redundancy (see Figure 2).
This trend contradicts expectations that multisite approaches will undermine expensive,
physical site redundancy strategies. While the industry may indeed move further toward
distributed and software-based resiliency models, maintaining and increasing site-level
redundancy remains a high priority for most operators.

Figure 2 Physical site redundancy still climbing

How have redundancy levels changed in the past three to five years in your data center?
Enterprise (n=346)
POWER 39% 58% 3%
COOLING 37% 61% 1%
Colocation / data center provider (n=186)
POWER 35% 60% 4%
COOLING 33% 64% 3%
Cloud / hosting / SaaS provider (n=46)
POWER 37% 61% 2%
COOLING 33% 63% 4%

Increased redundancy No change Decreased redundancy

(All figures rounded.)

UPTIME INSTITUTE GLOBAL SURVEY OF IT AND DATA CENTER MANAGERS 2023

PUBLICLY REPORTED OUTAGES

Major public outages — the headline makers

In addition to regular resiliency and other surveys, Uptime Intelligence also tracks
major (sometimes high-profile) outages that are reported by the media or other
publicly available sources, such as social media and government. These collected and
aggregated reports provide further insights into the nature and impact of outages.
Table 3 shows the number of outage reports Uptime collected from publicly available
sources over the past eight years. The numbers do not necessarily represent the
underlying number of outages. In recent years, media and social media have reported
an increasing number of outages, even though many are trivial and the reports are
sometimes inaccurate. Therefore, we no longer collect reports of small, category 1
outages where it is difficult to verify details and where no great financial loss, disruption
or reputational damage occurred. In 2023, Uptime collected details of 110 outages.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 9

KEYNOTE REPORT: Annual outage analysis 2024

Table 3 Publicly reported outages tracked by Uptime, 2016 to 2023

2016 2017 2018 2019 2020 2021 2022 2023 Total
27 57 71 165 118 109 111 110 768

UPTIME INSTITUTE 2024

The proportion of publicly reported outages that were either serious or severe continued
to increase in 2023. Figure 3 shows that category 4 and 5 outages combined increased
by 12 percentage points compared with 2022. This may reflect the growing dependence
of other industries on IT — when outages happen, the consequences are often
widespread and affect millions of users; however, it may also mean that the media have
become less interested in reporting minor outages. Nevertheless, it is likely that major
incidents do cause greater financial and reputational damage than in previous years and,
as such, are more likely to be reported on by public media outlets.

Figure 3 Proportion of publicly reported outages that were serious or severe, 2016 to 2023

11% Severe 5%
(category 5)
29%

22% Serious
(category 4)

67%
Minimal / significant 65%
(categories 2 and 3)

2016 2017 2018 2019 2020 2021 2022 2023

(n=27) (n=57) (n=71) (n=165) (n=118) (n=109) (n=111) (n=110)
(Category 1, negligible, outages omitted for all years.)

UPTIME INSTITUTE 2024

This data suggests that each year there will probably be 10 to 20 high-profile IT outages
or data center events across the world that cause serious or severe financial loss,
business and customer disruption, reputational loss and, in extreme cases, loss of life.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 10

KEYNOTE REPORT: Annual outage analysis 2024

Outage causes
Establishing the root cause of a data center outage is imperative for preventing repeat
instances of disruption and for identifying areas that require greater investment to mitigate
the risks. However, assessing outage data poses challenges due to the multifaceted nature
of most incidents, which often stem from a combination of factors.
Uptime’s annual surveys consistently show that disruptions to on-site power distribution
are the most common cause behind impactful outages (see Figure 4). This is unsurprising
given the intolerance of IT hardware to any significant power disturbances, such as voltage
fluctuations or complete loss of power, that last more than fractions of a second.
Conversely, failures or underperformance of cooling equipment are generally tolerated for
longer durations, often measured in minutes, due to thermal ride-through mechanisms
or network traffic redirection capabilities. While IT-originating failures may occur more
frequently, they often have isolated, minor effects that go unrecorded and primarily impact
specific applications or datasets.
Third-party provider issues have seen a marginal but consistent uptick since 2020, rising
by five percentage points to account for nearly one in 10 outages in 2023. This steady
increase reflects the growing reliance on cloud / hosting, software as a service (SaaS)
and colocation providers.

Figure 4 Power remains the number one root cause of outages

What was the primary cause of your data center’s most recent impactful incident or outage?
(n=108)

Power 52%

Cooling

Third-party provider

IT systems (hardware / software)

1%
3% 1%
Network

Fire / fire suppression

19% 7%
Information security-related
8%
Not known 9%

UPTIME INSTITUTE GLOBAL SURVEY OF IT AND DATA CENTER MANAGERS 2023

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 11

KEYNOTE REPORT: Annual outage analysis 2024

For a more general view of outage causes that extend beyond (but include) the data center,
Uptime’s annual resiliency survey also asks about the most common causes of any end-
to-end IT service outages, regardless of whether they were the most recent or the most
impactful. The responses have consistently shown that network-related outages are more
common, and are considerably ahead of power-related outages (see Figure 5).

Figure 5 Most common causes of IT service-related outages

Which of the following issues has been the most common cause of any IT service outages that
may have affected your organization over the past three years?
(n=406) 2023 2024 (n=442)
31% Networking / 31%
connectivity

IT system /
18% software 22%

23% Power 18%

7% Cooling 7%

Third-party
10% IT service 8%

No IT service
9% outages 10%

2% Other 3%

UPTIME INSTITUTE DATA CENTER RESILIENCY SURVEY 2024

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 12

KEYNOTE REPORT: Annual outage analysis 2024

PUBLICLY REPORTED OUTAGES

Software or configuration outages rise with network complexity

Software tools are increasingly used to manage capacity planning, energy efficiency, network
optimization and monitor hardware. While these tools can help data center operators keep
up with the growing demand for high-bandwidth and low-latency application performance,
which should ultimately improve resiliency, they also add complexity and, consequently, new
risks. This partly explains a shift in the causes of incidents recorded in public reports over
the past few years. Compared with data from 2022, outages that stemmed from software
or configuration issues in IT systems and network infrastructure increased by five and 10
percentage points respectively (see Figure 6).
More complex and interconnected networks can also increase cyber risks. Cybersecurity
incidents — which have been on the rise in recent years — are now a clear cause of
outages. These attacks, often severe and resulting in the loss of private data, may also
be underreported, indicating a potentially higher impact.

Figure 6 Causes of publicly reported outages 2023

IT (software / configuration)

Network (software / configuration)

1%
1%
23% 2%
Power
4%
Cyberattack / ransomware

Fiber 6%
Fire

Cooling 9%
22%
Network (cabling)

Provider / partner issue

10%
Capacity / demand
11%
Other 11%
(n=90)
(Category 1, negligible, outages omitted.)

UPTIME INSTITUTE 2024

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 13

KEYNOTE REPORT: Annual outage analysis 2024

PUBLICLY REPORTED OUTAGES

Commercial operators in the limelight

Over the past eight years, third-party commercial operators of IT and/or data centers
combined (cloud / internet giants, digital services, telecommunications, etc.) have
accounted for more than two-thirds (67%) of publicly reported outages recorded by Uptime
(see Figure 7). This reflects the growth of professional, outsourced IT services, from
colocation to cloud and hosting. Moving to third-party venues and services may reduce
risks but failures still occur, and they can be very serious.
Compared with the average over this period, financial sector outages fell in both 2022 and
2023. This may be influenced by stricter regulations and oversight following a series of
large, high-impact outages before 2021.
The telecommunications sector, on the other hand, has experienced an uptick in outages.
This has been driven by various factors: rising demand for connectivity and capacity
across all sectors has put a strain on networks and legacy infrastructure, while the
criticality of mobile networks, in particular, means outages can have an outsized impact.
The increasing use of standardized and less expensive data centers — compared with
earlier, resilient but limited designs — may have increased some risks while lessening
others. The adoption of technologies, such as software-defined networking, 5G and
network function virtualization, adds complexity to these systems.

Figure 7 Publicly reported outages by sector, 2016 to 2023

Cloud / internet giant
22%
21%
Digital services
25%
24%
Financial services
19%
9% 2016-2023
(n=n=696)
Government
7% 2023 (n=99)

7%
Telecoms
21%
30%
Transportation
7% (Only top 6 outage categories are shown.)
(Category 1, negligible, outages and outages in which the root
8% cause was not determined or disclosed are omitted for all years.)

UPTIME INSTITUTE 2024

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 14

KEYNOTE REPORT: Annual outage analysis 2024

Cloud and third-party provider reliability

Cloud services are engineered with the aim of achieving minimal failure rates. Major cloud
and IT service providers deploy layers of software and middleware; strategically distribute
capacity across systems, networks and data centers; and are able to redirect workloads and
traffic away from areas experiencing issues. As a result, these architectures offer high levels
of service availability, particularly at scale.
However, no architecture is immune to failure. The inherent challenges associated with
managing intricate, large-scale software, data and networks mean that outages are
inevitable. Given the rise in public cloud adoption across multiple industries, these failures
will affect millions of users and receive significant media attention.
Despite the expectation that cloud, hosted and other internet-based services will play
an increasing role in corporate IT, operators remain wary when it comes to trusting their
organization’s most sensitive applications and data to third parties. Only one in seven (16%)
operator respondents to 2024 Uptime resiliency survey say the public cloud is resilient
enough for all their organization’s mission-critical workloads, while a similar proportion
(15%) do not trust the public cloud with any mission-critical workloads (see Figure 8).

Figure 8 Most say cloud only resilient enough for some workloads
Regarding public cloud services, do you think public cloud is resilient enough to run all of your
organization’s mission-critical IT workloads, run most of them, run only some of them, or is public
cloud not resilient enough to run any of your organization’s mission-critical IT workloads? (n=441)

Public cloud is NOT resilient Public cloud is resilient

enough to run any of our 15% 16% enough to run all of our
mission-critical workloads mission-critical workloads

Public cloud is resilient

29% enough to run most of our
mission-critical workloads
Public cloud is resilient enough
to run only some of our 40%
mission-critical workloads

(Responses for "Don't know" are not included.)

UPTIME INSTITUTE DATA CENTER RESILIENCY SURVEY 2024

In the past three years, many organizations have backed away from a “cloud-first”
strategy and are taking a more cautious, selective approach. This has shown up in Uptime
survey data, which, to some extent, has predicted the slowdown of large-scale corporate
migrations to the cloud.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 15

KEYNOTE REPORT: Annual outage analysis 2024

The main reasons preventing organizations from further adopting cloud services
for critical applications have shifted over the years. Rather than a lack of clarity and
transparency regarding providers’ operational resiliency (as it has been in previous
years), most operators (64%, n=240) in 2024 cite data security concerns as the main
barrier to increasing the adoption of public cloud services. Notably, only one in five (20%)
respondents cite resiliency concerns as their organization’s main deterrent.
It is likely that these data security concerns follow widely publicized cyberattacks on some
cloud providers, in which services were taken offline and confidential information was
compromised. Indeed, one in five (20%) operators that experienced an outage due to a
third-party IT provider cite a malicious cyberattack as the root cause — an uptick of seven
percentage points from 2022. However, software or configuration errors from third-party
IT service providers are more than three times (62%) as likely to cause an outage as
cybersecurity issues (see Figure 9).

Figure 9 Most common causes of major third-party outages

Has your organization experienced a significant, serious, or severe IT service outage(s) that
was caused by a problem with a third-party IT service provider over the past three years? If
so, what are their most common causes? Choose no more than three. (n=202)

Software or
configuration error 62%

Networking /
connectivity issues 35%

53% 39% Mechanical or

electrical failure 22%
NO YES

Malicious cyberattack 20%

8%
Don’t
know
Overload / congestion
/ capacity issues 19%

(Only top 5 outage categories are shown.)

(The sum percentages for most common causes exceed 100% due to respondents being asked to choose up to three options.)

UPTIME INSTITUTE DATA CENTER RESILIENCY SURVEY 2024

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 16

KEYNOTE REPORT: Annual outage analysis 2024

Power outages
Power-related disruptions often lead to the most severe downtime incidents. While diagnosis
and restoration of power are often quick, restarting the IT equipment and synchronizing
the databases can take several hours — assuming these systems were not damaged during
the disruption.
Uptime’s annual survey data consistently shows that, while power-related issues are the
most common cause of impactful outages for data centers (see Outage causes), they also
represent a growing share of overall outages year on year.
Challenges with electrical grids may exacerbate this trend in the years ahead. Grid reliability is
under threat due to a combination of factors, including aging infrastructure and transmission
systems, escalating demand, the decommissioning of older power generation plants, severe
weather events and an increasing reliance on intermittent renewable energy sources.
In terms of power-related outages at their site, 30% of operators who responded to the
2024 Uptime resiliency survey suffered an impactful outage caused by a problem with
power systems in the past three years (see Figure 10). UPS issues are cited as the most
common underlying cause of these outages — as they have been in every year since we
have conducted this survey.

Figure 10 Most common causes of major power-related outages

Has your organization experienced a significant, serious, or severe IT service outage(s) that
was caused by a problem with a power system over the past three years? If so, what are their
most common causes? Choose no more than three. (n=431)

UPS failure 42%

Transfer switch
(e.g. utility) failure 30%
Generator failure 28%
Transfer switch between
66% 30% paths (A/B) failure 18%
NO YES
Power distribution
unit failure
16%
Single-corded IT
device(s) failure 14%
3% Controls failure 8%
Don’t know

(All figures rounded. Only top 7 categories are shown.)

(The sum percentages for most common causes exceed 100% due to respondents being asked to choose up to three options.)

UPTIME INSTITUTE DATA CENTER RESILIENCY SURVEY 2024

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 17

KEYNOTE REPORT: Annual outage analysis 2024

There are several reasons for this. Uptime engineers working across many data centers
report the following as the most common problems with static UPS systems:
• Fans fail frequently because they are usually inexpensive and constantly in operation.
A single fan failure does not take down a unit, but the failure of multiple fans may.
• Snubber capacitors can fail from wear and tear. Regular preventative maintenance
will reduce the number of failures.
• Batteries fail because of age. They require good management, close monitoring
and adherence to replacement schedules. Many batteries fail because they are not
monitored closely enough by experienced technicians.
• Inverter stack failures are least common. These are more likely to occur when the unit
is overloaded, although wear and tear can also cause failures.
The frequency of UPS problems is more likely to increase with age, so supply chain /
replacement problems may lead to more failures. Operators of data centers with no
trusted concurrent maintainability designs (the ability to bypass any item of equipment
for maintenance without interrupting overall service) can be more likely to postpone
maintenance or replacement.
Generators are reliable, but require regular scheduled maintenance, fuel checks and testing.
Automatic transfer switch (ATS) units are generally robust, but failures may occur with active
controls or with a loss of direct current (DC) power to those controls. Other less common
failures are caused by mechanical issues, such as worn-out bearings or a jammed switch.

Networking outages
Networking issues have caused an increasing portion of IT service outages in recent years.
The 2024 Uptime resiliency survey finds that the two most common causes of networking-
or connectivity-related outages are configuration / change management failure and third-
party network provider failures — similar numbers to previous years (see Figure 11).
As demand patterns for applications evolve, data center networks also undergo
corresponding changes. The growing use of virtualization to accommodate this demand
heightens the reliance on software components, such as management, monitoring and
automation systems.
These tools can help prevent network-related incidents that are caused by human error.
However, they require script modifications when network changes occur, which can lead to
errors during reconfiguration. When organizations utilize multiple hardware vendors, this
becomes more challenging because it requires further maintenance and the adaptation of
multiple scripts with each introduced change.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 18

KEYNOTE REPORT: Annual outage analysis 2024

Figure 11 Most common causes of major network-related outages

Has your organization experienced a significant, serious, or severe IT service outage(s) that was
caused by a problem with networking / connectivity over the past three years? If so, what are
their most common causes? Choose no more than three. (n=444)

Configuration / change
management failure 46%
Third-party network
provider failure 42%
Hardware failure 33%
Firmware / software error 20%
46% 46%
NO YES Line breakages 19%
Network / congestion failure 15%
8% Corrupted firewall /
13%
Don’t routing tables issues
know
Cyberattack 12%
Weather-related incident 7%

(Responses for "Other" and "Don't know" are not included.)

(The sum percentages for most common causes exceed 100% due to respondents being asked to choose up to three options.)

UPTIME INSTITUTE DATA CENTER RESILIENCY SURVEY 2024

Configuration errors, firmware errors and corrupted routing tables all play a significant role in
network-related failures, while the more traditional worries of weather and cable breaks take
on a much smaller role overall. Congestion and capacity issues can also cause failures, but
these are often the result of programming or configuration issues.
In a complex and high-throughput environment, small errors can propagate across networks,
resulting in cascading failures that can be difficult to stop, diagnose and fix. The high number
of network or software failures has clearly contributed to the rise in telecommunications
failures in publicly reported outages.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 19

KEYNOTE REPORT: Annual outage analysis 2024

System and software outages

IT system and software-related failures and bugs are a persistent, endemic problem for IT
operations, but over time enterprises have developed processes, tools and architectures to
minimize their impact, with considerable success. Even so, more serious outages persist
as a challenge for operators, with more than one-third (38%) experiencing significant
disruptions over the past three years (see Figure 12).
These outages, similar to those caused by network-related issues, stem from the complexity
and scale of modern systems, coupled with the increasing reliance on software for
ensuring availability across distributed data center sites (some operated by third parties).
Issues with database synchronization, load balancing and traffic management can lead to
partial or complete service downtime across multiple data centers or availability zones.

Figure 12 Most common causes of major IT system- / software-related outages

Has your organization experienced a significant, serious, or severe IT service outage(s) that
was caused by a problem with IT systems / software over the past three years? If so, what are
their most common causes? Choose no more than three. (n=426)

Configuration / change
management issue 60%

Firmware / software fault 38%

Hardware failure 30%

50% 38%
NO YES Data synchronization /
corruption 21%

Capacity /
12% congestion issue 21%
Don’t know
Cyberattack /
security issue 16%

(Responses for "Other" and "Don't know" are not included.)

(The sum percentages for most common causes exceed 100% due to respondents being asked to choose up to three options.)

UPTIME INSTITUTE DATA CENTER RESILIENCY SURVEY 2024

Software problems primarily arise from configuration and change management issues,
patches, upgrades and other changes, which can lead to instability and unanticipated
errors. These errors become more difficult to contain once they have propagated across
networks. Hardware and software faults are less likely than configuration and change
management issues to cause an outage — but taken together, they still contribute to a
significant number of outages.
Compared with 2022, cyberattacks, including ransomware and distributed denial of
service (DDoS) attacks, increased by six percentage points. When such incidents occur, the
consequences can be severe, leading to data loss, financial loss and reputational damage.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 20

KEYNOTE REPORT: Annual outage analysis 2024

The human factor

Data center operators face considerable challenges when it comes to preventing and
mitigating downtime caused by human error. This is because failures can stem from
various factors, such as the adequacy of training, the effectiveness of procedures in place,
staff fatigue, resource availability and the complexity of equipment operation. There is also
uncertainty around how such incidents should be defined; for instance, whether a machine
failure caused by a software error at the factory constitutes human error.
As a result, Uptime tends to analyze human error as a contributing factor rather than the
sole or primary cause of outages. Drawing on 25 years of data, Uptime estimates that
human error, whether directly or indirectly, contributes to a significant majority — ranging
from two-thirds to four-fifths — of all downtime incidents.
In recent surveys on resiliency, Uptime has tried to understand how the makeup of some of
these failures relates to human error. Figure 13 shows that human error-related outages
are mostly caused either by staff failing to follow procedures (even if they have been
agreed upon and codified) or by the procedures themselves being inadequate.

Figure 13 Most common causes of major human error-related outages

Has your organization experienced a significant, serious, or severe IT service outage(s) that was
caused by human error over the past three years? If so, what are their most common causes?
Choose no more than three. (n=418)

Data center staff failing

to follow procedure 48%
Incorrect staff processes
/ procedures 45%

Installation issues 23%

51% 39% In-service issues 20%
NO YES

Insufficient staff 15%

10%
Preventative maintenance
Don’t
frequency issues
14%
know
Data center design
or omissions
10%

(Responses for "Other" and "Don't know" are not included.)

(The sum percentages for most common causes exceed 100% due to respondents being asked to choose up to three options.)

UPTIME INSTITUTE DATA CENTER RESILIENCY SURVEY 2024

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 21

KEYNOTE REPORT: Annual outage analysis 2024

While even the most robust methods of training and effective processes for staff cannot
prevent all possible failures, nearly four in five (78%) operators report that better
management and processes would have prevented their organization’s most recent
downtime incident (see Figure 14). This proportion has been highly consistent. Every year
since 2020, at least 75% of operators have noted the preventability of their most recent
outage; this suggests that there is a major opportunity to make significant reductions
in downtime.

Figure 14 Most say outages preventable with better management

Would your data center’s most recent impactful downtime incident have been preventable
with better management / processes or configuration? (n=108)

78% 22%
YES NO

UPTIME INSTITUTE GLOBAL DATA CENTER SURVEY 2023

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 22

KEYNOTE REPORT: Annual outage analysis 2024

PUBLICLY REPORTED OUTAGES

Are outages getting longer?

The longer an outage, the more likely it is to be expensive and disruptive, resulting in
damaged reputations. It is also more likely to receive media attention. Data from reported
publicly reported outages shows that, even from this group of high-profile outages, the
majority (64%) are resolved within 12 hours. However, there is one concerning trend: the
number of outages that have not been fully recovered after 24 hours is rising (see Figure 15).
There may be several reasons for this rise; for example, fires are on the increase, as are
the complications of synchronizing distributed data and management systems. Major
ransomware attacks, which usually require the shutdown of all potentially affected
systems, are also becoming more common. Nearly half of the outages that lasted longer
than 48 hours were caused either by fire or cyberattacks (45%, n=20).

Figure 15 Durations of publicly reported outages, 2017 to 2023

4%
4% 18%
7%
9%
23% 8%
27%

63%
37%

2017 2018 2019 2020 2021 2022 2023

(n=57) (n=71) (n=142) (n=117) (n=109) (n=108) (n=110)

0-4 hours 4-12 hours 12-24 hours 24-48 hours >48 hours

(Category 1, negligible, outages are not included.)

UPTIME INSTITUTE 2024

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 23

KEYNOTE REPORT: Annual outage analysis 2024

PUBLICLY REPORTED OUTAGES

Recent serious / severe outages

Most years, there are 15 to 20 outages publicly reported that Uptime classes as “severe”
or “serious” (category 4 or 5). This means that, depending on the services involved, the
outage(s) may result in high costs, reputational damage, threats to life or safety, and
serious breaches of compliance rules.
Some examples of major publicly reported outages in 2023 and early 2024 are shown in
Table 4. The list is largely made up of telecommunications, cloud and / or digital services
companies, where outages can affect a wide variety of customer.

Table 4 Ten major outages in 2023 and 2024

Company Severity Year
name Industry rating (quarter) Cause Impact

Federal Stored files in a pilot-alert system were

Configuration
Aviation mistakenly deleted. More than 30,000
Government 5 2023 (Q1) error / IT
Administration domestic flights were grounded, delayed or
(US)
software
canceled. Stocks of major US airlines affected.

Cybercriminals encrypted company data,

Dish Network Ransomware disrupting services for 300,000 users. Some
Telecommunications 3 2023 (Q1)
(US) / cyberattack services were still affected weeks later.
Stock value fell by more than 6%.

A cyberattack took cloud services offline for

Western Ransomware a week, locking out users from their stored
Digital services 5 2023 (Q2)
Digital (global) / cyberattack files and potentially compromising sensitive
company and personal data.

Some banking services were offline for

Zenith Bank millions of users after a backup power
Financial services 5 2023 (Q2) Fire
(Nigeria) system caught fire. Issues with the disaster
recovery infrastructure prolonged the outage.

Cybercriminals “shut down all systems,”

CloudNordic,
Ransomware including its website and email, and
Azerocloud Cloud / internet giant 5 2023 (Q3)
/ cyberattack encrypted customer systems. Most
(Denmark)
customers “lost all data.”

A cyberattack forced the company to shut

Kyivstar Ransomware down network connections, disrupting
Telecommunications 5 2023 (Q4)
(Ukraine) / cyberattack a regional air-raid warning service, IT
infrastructure and some banking services.

Configuration A networking issue caused transport delays,

Optus error / banking issues and cut hospital phone lines
Telecommunications 5 2023 (Q4)
(Australia) network for 12 hours, affecting more than 10 million
software users and 400,000 businesses.

A faulty signal closed chilled water tank

Citibank and Cooling
valves, which raised temperatures and took
DBS Bank Financial services 4 2023 (Q4) equipment
services offline. The central banking authority
(Singapore) failure
issued significant penalties for the downtime.

A water leak caused a transformer to

explode, cutting power to the facility’s two
Windstream
Telecommunications 5 2023 (Q4) Fire main switches. Backup generators then
(Nebraska, US)
failed, taking down emergency services
communications for nearly 12 hours.

Multiple government services, including

Fulton County, Ransomware court filings, e-records, phone lines and
Government 4 2024 (Q1)
Georgia (US) / cyberattack email, were down following a cyberattack by
the LockBit group.

UPTIME INSTITUTE 2024

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 24

KEYNOTE REPORT: Annual outage analysis 2024

Cost of outages
Even though operators appear to have reduced the likelihood of the most serious
and severe outages, those that do happen tend to be expensive. More than half of the
respondents (54%) to Uptime’s 2023 annual survey say their most recent significant,
serious or severe outage cost more than $100,000, with 16% saying that their most recent
material outage cost more than $1 million (see Figure 16).
These cost figures are slightly lower than in previous years. However, as noted in
the Outages frequency and severity section of this report, changes to the survey’s
methodology may have affected year-on-year comparisons.
The high cost of outages stems from multiple factors, including inflation, penalties for
breaching service level agreements (SLAs), labor costs, callouts and the expense of
replacing parts. However, the primary driver is the increasing reliance of corporate
economic operations on digital services and data centers. The failure of critical IT services
frequently results in immediate business disruption and revenue loss.
Uptime does not calculate an average cost of outages because the insights gained are
rarely useful — businesses and outage impacts vary widely. Each year, a few large outlier
outages are so costly that they can distort the overall picture. Some result in compensation,
fines and lost business, with costs adding up to millions or even tens of millions of dollars.
The high costs resulting from outages will likely increase over time as the dependency on
digital services also rises. Stronger SLAs, expected by some businesses because of this
growing reliance, could make outages even more costly, as will more and higher regulatory
fines and compensation for customers who experience a service disruption. This, in turn,
justifies better analysis of the causes and costs of outages and the continued or increased
investment in resiliency.

Figure 16 Half of impactful outages cost more than $100,000

Please estimate the total cost of this downtime incident (from outage to full recovery) for your data
center, including direct, opportunity and reputation costs, using the following options. (n=94)

46% $ $ $ $ $ $ $ $ $ $
Under $100,000
$ $ $ $ $ $ $ $ $ $
$ $ $ $ $ $ $ $ $ $ 38%
$ $ $ $ $ $ $ $ $ $ $100,000 to
$1 million
$ $ $ $ $ $ $ $ $ $
$ $ $ $ $ $ $ $ $ $
$ $ $ $ $ $ $ $ $ $
$ $ $ $ $ $ $ $ $ $
16%
$ $ $ $ $ $ $ $ $ $ Over $1 million

$ $ $ $ $ $ $ $ $ $
UPTIME INSTITUTE GLOBAL DATA CENTER SURVEY 2023

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 25

KEYNOTE REPORT: Annual outage analysis 2024

Summary
High availability and resiliency (which means outage prevention and effective recovery) are
priorities for all involved in the digital infrastructure supply chain. It is sometimes assumed
that progress in this area is as reliable as Moore’s law has been in the past three decades.
This is not the case: Uptime’s data shows that progress is gradual, hard-won and — when
failures occur — expensive.
This year’s data shows that progress is being made in improving the outage rate relative to
overall IT capacity. This can be attributed to a range of measures: greater investment, the
combined effects of software-based resiliency and on-site physical redundancy, improved
training, the outsourcing and greater professionalism of some third-party operators, and
overall continuing vigilance.
But Uptime’s research also identifies several trends in 2024 that could undermine
progress in outage rates and overall resiliency.
First, the widespread adoption of distributed architectures — where IT workloads are
spread across multiple sites — aims to mitigate localized failures. However, Uptime data
suggests that this shift may play a role in the increase in network, software or system-
related incidents. This will likely be a transitional effect.
Second is the increase in system and network complexity, which includes the use of more
automated and remote management at a facility level, supported by software-based
optimization strategies. The greater use of both on-site and distributed systems enables
data center operators to improve maintenance, energy management, capacity planning
and incident monitoring — but it also requires greater integration with IT and operational
technology (OT) systems, a heightened awareness of security risks and rigorous testing.
The greater use of these systems can create additional access pathways and widen the
attack surface for malicious actors, which may be a factor in the notable rise in outages
caused by cybersecurity incidents.
Third is the ongoing challenge of recruiting and training staff as well as the establishment
of proven management processes. To make significant progress in reducing downtime,
owners and operators will have to allocate increased efforts and resources to these areas.
Finally, there are two areas of increasing risk that are external to the data center itself:
grid stability, which may deteriorate due to high demand and the transition to intermittent,
renewable energy; and the effects of climate change (notably, extreme weather). While
operators can do little about these directly, there are many steps that they can take to
reduce their exposure.
In summary, outage prevention requires ongoing vigilance and investment — and currently,
the digital infrastructure industry is on an improving trajectory. Robust data center design,
detailed attention to IT architectures and topology, physical infrastructure redundancy,
testing, improved training and continuous review will continue to be necessary if this is
to be maintained.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 26

KEYNOTE REPORT: Annual outage analysis 2024

Appendix

Sources and methodology

Uptime Institute currently has four main data sources for monitoring data center and IT
outages or incidents that could lead to outages:

Uptime Institute Global Survey of IT and Data Center Managers

This long-running series of global annual surveys asks detailed questions about outages. Some of
the findings are discussed here. This represents the most statistically significant dataset relating
to outages in the critical infrastructure industry.

Uptime Institute Data Center Resiliency Survey

This global survey specifically focuses on outages and resiliency-related issues. The results
are compared and contrasted with those from the Uptime Institute Global Survey of IT and Data
Center Managers, which is conducted midyear.

Uptime Institute Intelligence’s public outages database

Since the beginning of 2016, Uptime Institute has collected data about major IT outages from
media reports and other public sources (social media, outage detection sites, etc.). This has
enabled us to collect information on major outages that become visible to the public and the
media, and to identify patterns over time.

Uptime Institute’s Abnormal Incident Report (AIRs) database

This is a long-standing confidential system for global Uptime Institute members to share details
of incidents under a nondisclosure agreement. Most incidents recorded do not actually lead to
outages — many are “near misses”. We do not include such incidents in the analyses described in
this report.

Uptime Institute Professional Services

This is a less formalized source of information. Uptime Institute conducts Digital Resiliency
Assessments and root-cause analyses of failures on behalf of clients globally. Although these
assignments are confidential, the experience garnered from these incidents informs our analyses.

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 27

KEYNOTE REPORT: Annual outage analysis 2024

Uptime compiles a database of public outages, which is used for some of the findings in
this report. However, it is not a comprehensive list of all the outages that have occurred.
There are some limitations that should be considered:
• If a failure is not reported or picked up by the media or Uptime, it will not be recorded.
This immediately means there is a bias toward coverage of large, public-facing IT
services, particularly in geographies with well-developed and open media.
• Uptime limits failures to those that had a noticeable impact on end users — a major
fire during data center commissioning, for example, may never be registered. All
category 1 outages have been eliminated, i.e., small, short failures where the business
or reputational impact is negligible.
• The amount of information available varies widely from outage to outage. In some
of the analyses, it has been necessary to include outages for which the cause is “not
known” — which means information was never disclosed.
• Although IT system failures are included, cybersecurity breaches are generally not
included, apart from those that lead to complete service interruptions.

Other related content published by Uptime Institute includes:

The Uptime Institute Global Data Center Survey 2023

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 28

KEYNOTE REPORT: Annual outage analysis 2024

About the authors

Andy Lawrence
Andy Lawrence is Uptime Institute’s Executive Director of Research. He has spent
three decades analyzing developments in IT, emerging technologies, data centers and
infrastructure, and advising companies on their technical and business strategy.
[email protected]

Douglas Donnellan
Douglas Donnellan is a Research Analyst at Uptime Intelligence covering sustainability in
data centers. His background includes environmental research and communications, with
a strong focus on education.
[email protected]

All general queries About Uptime Institute

Uptime Institute Uptime Institute is the Global Digital Infrastructure Authority. Its Tier Standard is the IT industry’s most
405 Lexington Avenue trusted and adopted global standard for the proper design, construction, and operation of data centers
9th Floor — the backbone of the digital economy. For over 25 years, the company has served as the standard for
New York, NY 10174, USA data center reliability, sustainability, and efficiency, providing customers assurance that their digital
infrastructure can perform at a level that is consistent with their business needs across a wide array
+1 212 505 3030
of operating conditions. With its data center Tier Standard & Certifications, Management & Operations
[email protected] reviews, broad range of related risk and performance assessments, and accredited educational curriculum
completed by over 10,000 data center professionals, Uptime Institute has helped thousands of companies,
in over 100 countries to optimize critical IT assets while managing costs, resources, and efficiency.
Uptime Institute is headquartered in New York, NY, with offices in Seattle, London, Sao Paulo, Dubai,
Singapore, and Taipei.
For more information, please visit www.uptimeinstitute.com

© COPYRIGHT 2024 UPTIME INSTITUTE. ALL RIGHTS RESERVED 29