data_ center_availability_report_2014
TRANSCRIPT
Veeam Data Center Availability Report 2014 The Challenge of The Always-On Business
Veeam Data Center Availability Report 2014 | 1 vee.am/availability14
ContentsExecutive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Survey Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Part 1: The Modern Data Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Part 2: The Always-On Business . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Part 3: The Availability Gap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Part 4: The Financial Cost of Downtime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Part 5: Availability Solutions and Capabilities – The Root of the Issue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Part 6: An Example of Lack of Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Part 7: Looking Ahead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Veeam Data Center Availability Report 2014 | 2 vee.am/availability14
The Veeam Data Center Availability Report 2014 investigates the increasing demands organizations face to provide an ‘Always-On Business’, what actions they are taking to meet those demands, and how successful their actions are. Following on from previous Veeam Data Protection Reports, this report investigates whether existing solutions can provide the always-on availability that businesses demand in the 21st Century.
In particular, the report shows that there are clear demands for 24/7 access to IT services and applications, with over 90 percent of enterprises increasing their requirements for minimizing downtime and guaranteeing access to data. It also demonstrates how organizations are modernizing their data center infrastructure, in order to meet these requirements.
Despite investments in the modern data center, there is an “availability gap” between the Always-On Business requirements and what legacy backup solutions can deliver in terms of Recovery Time Objective (RTO) and Recovery Point Objectives (RPO). Indeed, to meet the demands of the Always-On Business, organizations would need to recover mission-critical data in 60 percent of the time it takes them now and perform backup 1.5 times more often.
This report shows the scale of this gap and the financial impact of failing to meet business demands. Currently, organizations suffer application downtime 13 times a year, with the total cost of downtime and data loss reaching up to $10,163,114. In addition, organizations experience backup recovery failure twice a year. Because of this, organizations will lose at least $2 million a year through data loss and factors such as lost productivity and missed opportunities.
Figure 1: Respondent industries (%)
23
22
22
17
16
Retail, distribution and transportManufacturingFinancial servicesBusiness & professional servicesOther commercial sector
Retail, distribution and transport
Manufacturing
Financial services
Business & professional services
Other commercial sector
The concluding sections of the report break down the reasons why legacy backup solutions do not yet meet RTO and RPO requirements. It was highlighted that these solutions lack capabilities such as high-speed recovery (wanted by 60 percent of organizations), data loss avoidance (53 percent); verified protection (47 percent); using backup data as a production-like test environment for new patches or updates (38 percent); and complete visibility, including proactive monitoring and alerting of issues before any operational impact (36 percent).
It then outlines the actions enterprises are taking to deliver Availability for their Modern Data Centers. Considering the lack of capabilities above, it comes as little surprise that 78 percent of organizations plan to change their backup solution in the next 2 years.
This report is based on an online survey conducted in August and September 2014 by Vanson Bourne, an independent market research organization, of 760 senior IT decision makers from organizations across the United States, United Kingdom, Germany, France, Italy, the Netherlands, Switzerland, Brazil, Australia and Singapore that employ more than 1,000 people.
Survey BackgroundRespondents were selected from a cross-section of industries. These were: Retail, Distribution and Transport, comprising 23 percent of respondents; Manufacturing (22 percent); Financial Services (22 percent); Business & Professional Services (17 percent); and other commercial sectors (16 percent). As a result, the survey findings were evenly distributed across a variety of enterprises and sectors, with less risk of one sector skewing the statistics (Figure 1).
Executive Summary
Part 1: The Modern Data Center
Veeam Data Center Availability Report 2014 | 4 vee.am/availability14
Part 1: The Modern Data Center
Figure 2: Organizations that have modernized, are modernizing or are planning to modernize their data centers (%)
Data center is already modernized
Modernizing the data center now
Planning to modernize within 2 years
No plans to modernize data center
Business requirements have changed dramatically in a relatively short time. IT has become strategic for every organization, and business requirements for IT have dramatically changed. To stay competitive today businesses need to provision IT services faster, strengthen security and control, lower operational costs and increase business agility. To meet these requirements businesses
are building a modern data center by investing into modern technologies such as virtualization, modern storage and cloud. Currently, 81 percent of organizations are either modernizing or have already modernized their data centers, with a further 16 percent planning to do so within the next 2 years (Figure 2).
TOTAL25 56 16 3
26
5716
FR22
5820
BR
26
48
21
5 DE20
6416
IT
21
6514US7 83
10SG
1760
23
CH
43
3614
6 UK
26
6212AU
3030
2218
NL
AVERAGE
1
Veeam Data Center Availability Report 2014 | 5 vee.am/availability14
68 percent of those organizations modernizing their data centers are doing so in order to enable 24/7, always-on business operations (Figure 3). There appears to be a consensus on the technology needed to fully modernize a data center. 97 percent of organizations engaged
Figure 3: Business drivers for data center modernization (%)
Figure 4: Technologies organizations are investing or planning to invest in (%)
in data center modernization are either investing or planning to invest in server virtualization. Other technologies focused on are storage upgrades (95 percent), OS upgrades (94 percent), and data protection & disaster recovery (93 percent) (Figure 4).
68%
Enable always-on business
operations
70%
Lower operational costs for IT
67%
Strengthen security and
control
58%
Provision new applications and services
more quickly
49%
Give end-users more fast and
efficient IT services
97
95
94
93
88
87
83
82
79
79
78
70
20
Server virtualization
Storage upgrades
OS upgrades
Data protection and disaster recovery
Network virtualization
Virtual desktop infrastructure
Consolidating existing data centers
Private cloud
Public cloud: Infrastructure as a Service (IaaS)
Adding new data centers
Public cloud: Software as a Service (SaaS)
Public cloud: DR as a Service (DRaaS)
Other
Part 2: The Always-On Business
Veeam Data Center Availability Report 2014 | 7 vee.am/availability14
Part 2: The Always-On Business
Figure 5: Organizations increasing their availability requirements over the past 2 years (%)
Organizations that have increased requirements for minimizing application downtime
Organizations that have increased requirements for guaranteeing access to data
Over the last 10 years, businesses have seen end-users demand increased access to data and applications for many reasons, including:
• Workers departing from the traditional nine-to-five working day
• Globalization allowing businesses to branch out across multiple time zones
• Customers conducting business online at any time
• Supply chain and logistics integration and automation requiring constant access to operational systems and data
• The rise of the Internet of Things (IoT) meaning devices are permanently connected and monitored
Taken in concert, these demands mean that businesses must be “always-on.” More data and applications are considered to be mission-critical, and businesses have less patience for downtime, driving the need for increased Availability for their Modern Data Centers.
Over 90 percent of organizations surveyed were increasing their availability requirements to meet the always-on needs of their business. Specifically, 93 percent of respondents are increasing their requirements for minimizing downtime, while 92 percent are increasing their requirements for guaranteeing access to data (Figure 5).
9394
9897
9192
78
100 100
83
97
92
92
98
90 91
94
88
100
97
80
97
65
70
75
80
85
90
95
100
AVERAGE Australia Brazil France Germany Italy Netherlands Singapore Switzerland UK USA
Veeam Data Center Availability Report 2014 | 8 vee.am/availability14
Figure 6: Key drivers for minimizing downtime and guaranteeing access to data (%)
This is being driven by the demands of end-users, the most common of which is more frequent, real-time interactions between customers, partners, suppliers and employees (65 percent). The need to access applications across time zones (56 percent), increased adoption of mobile devices (56 percent), employees working outside regular hours (54 percent) and an increasing level of automation for decision making and transactions (53 percent) (Figure 6), were also flagged as key requirements.
Figure 6: Key drivers for minimizing downtime and guaranteeing access to data, %
65
56
56
54
53
46
45
1
More frequent, real time interactions betweencustomers, partners, suppliers and employees
24/7 access to applications because of globalisation across time zones
Increasing adoption of mobile devices
24/7 access to applications because of employeesworking outside regular working hours
Increasing level of automation for decision making and transactions
Zero-tolerance attitude from users for downtime and data loss
Rapidly growing volumes of data
No key drivers
Part 3: The Availability Gap
Veeam Data Center Availability Report 2014 | 10 vee.am/availability14
Figure 7: Proportion of mission-critical workloads (%)
Current mission-critical percentage of workload
Expected mission-critical percentage of workload in two years’ time
Part 3: The Availability GapData center modernization does not automatically result in increased availability of all data and applications. In fact, despite investing in virtualization, advanced storage and the cloud, many businesses are not able to meet their RPO and RTO service level agreements (SLAs). This opens an “availability gap” between the availability requirements of the always-on business and what the company’s backup solution can actually deliver.
To help measure availability businesses look at RPO and RTO for critical and non-critical applications. Business must be confident that its most mission-critical applications will be available 24/7. Currently, 46 percent of workloads are mission-critical, although this is expected to be 53 percent by 2016 (Figure 7).
64
46 48
60
39 39
47
41
52
45
42
51
5354
67
44 45
50
59
64
51
47
58
35
40
45
50
55
60
65
70
AVERAGE Australia Brazil France Germany Italy Netherlands Singapore Switzerland UK USA
Veeam Data Center Availability Report 2014 | 11 vee.am/availability14
Figure 8: Average recovery time against current RTO: mission-critical applications (hours)
Figure 9: Average recovery time against current RTO: non-mission-critical applications (hours)
The first SLA critical to the always-on business is the RTO; i.e. how quickly applications can be recovered. Faster recovery of applications means less downtime and less impact to the business in terms of lost sales and productivity. Currently, mission-critical applications take an average of 2.86 hours to
Average recovery time
Current RTO
Average recovery time
Current RTO
recover, against an RTO of 2.69 hours (Figure 8), while non-mission-critical applications take an average of 8.45 hours to recover against an RTO of 10.02 hours (Figure 9). As we can see, the average enterprise is either meeting or close to meeting its SLAs for recovering data.
Figure 8: Average recovery time against RTO : mission-critical applications (hours)
2,86
1,6 1,61,9
1,1
5,3
1,9
2,3
3,9
5,2
3,32,69
2,6
3,6
1,7
1,0
3,7
2,9
1,0
3,3
4,1
3,0
AVERAGE Australia Brazil France Germany Italy Netherlands Singapore Switzerland UK USA
Figure 9: Average recovery time against RTO: non-mission-critical applications (hours)
8,45
8,2
5,3
8,2
5,8
14,0 14,3
2,3
4,8
13,7
7,1
10,0212,3
10,4
7,76,4
13,6
10,8
3,24,5
22,0
8,0
AVERAGE Australia Brazil France Germany Italy Netherlands Singapore Switzerland UK USA
Veeam Data Center Availability Report 2014 | 12 vee.am/availability14
The second critical SLA is the RPO; i.e. in the event of an IT failure, what is the most recent data that can be recovered and so how much will be irrecoverably lost? The more often an organization backs up its data, the smaller its RPO, and therefore the lower its risk exposure to data loss. Currently, organizations have less success meeting their RPO SLAs, putting them at risk of excessive data loss.
Mission-critical applications are backed up every 4.81 hours against an RPO of 3.53 hours (Figure 10). Non-mission critical applications are backed up every 14.46 hours against an RPO of 11.53 hours (Figure11). Looking at the RPO and RTO figures together it is clear that organizations can recover their applications within the agreed time, but they are at risk of excessive data loss.
14,46
17,4
11,0
16,5
14,5
22,4
12,0
8,0
21,3
15,8
11,5
11,53
9,310,4
13,5
7,6
17,7
10,7
5,0
13,9
18,5
9,5
AVERAGE Australia Brazil France Germany Italy Netherlands Singapore Switzerland UK USA
Figure 10: Average backup frequency against RPO: mission-critical applications (hours)
6,1
4,81
4,8
3,9
5,76,1
5,7
2,9
4,4
5,3
4,14,6
3,53
1,1
5,4
3,9
2,2
4,4
3,7
1,8
5,2
3,2
4,0
AVERAGE Australia Brazil France Germany Italy Netherlands Singapore Switzerland UK USA
Figure 10: Average backup period against current RPO: mission-critical applications (hours)
Figure 11: Average backup period against current RPO: non-mission-critical applications (hours)
Average backup period
Current RPO
Average backup period
Current RPO
Veeam Data Center Availability Report 2014 | 13 vee.am/availability14
Yet even this performance is currently not enough to meet a company’s availability demands. 82 percent of respondents say that there is an “availability gap” between the level of availability they can provide now and what end-users demand in order to provide an always-on business (Figure 12).
In order to address this gap, respondents state that their data protection solutions would need to offer an RTO of 1.73 hours
and an RPO of 3.19 hours. As a result, organizations would need to recover mission-critical data in 60 percent of the time it takes them now (Figure 13 on page 14) and perform backup 1.5 times as often as they do now (Figure 14 on page 14). However, it is highly likely that RTO and RPO requirements will continue to shrink as businesses attempt to guarantee 24/7 IT services. This means that, despite their data center modernization efforts, organizations will fall further behind the SLAs that the always-on business demands.
UNITEDKINGDOM
FRANCE GERMANY ITALY
NETHERLANDS SWITZERLAND UNITEDSTATES
BRAZIL
SINGAPORE
82%
70% 78% 80%
74% 87% 88%
96%
AUSTRALIA
86%
100%
82%
AVERAGE
Figure 12: Percentage of organizations identifying an “availability gap” (%)
Veeam Data Center Availability Report 2014 | 14 vee.am/availability14
Figure 13: Difference between desired RTO and average recovery time: mission-critical applications (hours)
Average recovery time
Desired RTO2,86
1,6 1,61,9
1,1
5,3
1,92,3
3,9
5,2
3,3
1,73
1,01,3
1,7
0,7
2,2
1,51,3
1,9
2,8
2,0
AVERAGE Australia Brazil France Germany Italy Netherlands Singapore Switzerland UK USA
4,814,8
3,9
5,76,1
5,7
2,9
4,4
5,3
4,1
4,6
3,19 3,6
2,7
3,6
2,0
3,43,0
1,9 1,9
5,1
3,0
AVERAGE Australia Brazil France Germany Italy Netherlands Singapore Switzerland UK USA
Figure 14: Difference between desired RPO and average backup period: mission-critical applications (hours)
Average backup period
Desired RPO
Part 4: The Financial Cost of Downtime
Veeam Data Center Availability Report 2014 | 16 vee.am/availability14
Part 4: The Financial Cost of DowntimeOne significant effect of failing to meet data and application availability demands is that organizations are exposed to unnecessary costs. For instance, businesses may miss sales opportunities or have to restructure their operations at some expense when critical applications fail; not to mention the costs
of lost productivity during any downtime. Since there is no longer a “safe” period of downtime outside normal nine-to-five working hours, these costs are magnified. On average, organizations encounter unplanned downtime 13 times per year (Figure 15).
UNITEDKINGDOM
FRANCE GERMANY ITALY
NETHERLANDS SWITZERLAND UNITEDSTATES
BRAZIL
SINGAPORE
AUSTRALIA
AVERAGE
10 16 15 11 17
9 12 13 14 12
13
Figure 15: Number of incidents of unplanned application downtime per year
Veeam Data Center Availability Report 2014 | 17 vee.am/availability14
Figure 17: Cost per hour of application downtime (US$)
Mission-critical applications
Non-mission-critical applications
Figure 18: Cost per incident of application downtime (US$)
1,33
0,58
1,541,19
0,90
2,16
0,89
1,46
2,40
1,15
1,56
3,97
3,30
2,74
4,54
3,01
5,69 5,78
3,473,72
4,99
3,34
AVERAGE Australia Brazil France Germany Italy Netherlands Singapore Switzerland UK USA
Figure 16: Length of unplanned application downtime (hours)
Mission-critical applications
Non-mission-critical applications
This downtime lasts 1.33 hours for mission-critical applications and 3.97 hours for non-mission-critical applications (Figure 16). The average cost of one hour of downtime for a mission-critical application is $82,864, and for a non-mission-critical application
is $43,886 (Figure 17). This means that an incident of mission-critical application downtime costs, on average, $110,209; while non-mission-critical downtime costs, on average, $174,227 (Figure 18).
Figure 17: Cost per hour of application downtime (US$)
$78 873
$105 844
$82 338
$50 611
$71 255
$75 641
$129 216
$56 953
$138 519
$68 052
$20 518
$40 920
$52 253
$26 389
$62 580
$26 069
$62 544
$26 658
$49 225
$51 944
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
AVERAGE $82 864$43 886
Figure 18: Cost per incident of application downtime (US$)
TOTAL
UK
France
Germany
Italy
Netherlands
Switzerland
USA
Brazil
Australia
Singapore
$45 746
$163 000
$97 982
$45 550
$153 911
$67 320
$188 655
$136 687
$159 297
$106 161
$174 227
$67 709
$112 121
$237 229
$79 431
$356 080
$150 679
$217 028
$99 168
$245 633
$173 493
$110 209 AVERAGE
AU
Brazil
France
DE
Italy
Netherlands
Singapore
Switzerland
UK
USA
Veeam Data Center Availability Report 2014 | 18 vee.am/availability14
As well as the costs of downtime itself, there is also a cost associated with data loss – i.e. data that has not been backed up and so cannot be recovered in the event of downtime. Depending on the importance of the data itself, this can be a huge cost for an organization in terms of
Figure 19: Cost of data loss per hour (US$)
Figure 19: Cost of data loss per hour (US$)
TOTAL
UK
France
Germany
Italy
Netherlands
Switzerland
USA
Brazil
Australia
Singapore $70 913
$84 018
$70 844
$64 075
$44 386
$63 278
$107 694
$128 540
$56 928
$87 317
$61 948
$42 016
$26 226
$49 779
$37 843
$24 759
$43 550
$29 033
$61 157
$50 211
$43 360
$52 804
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
missed sales opportunities and lost productivity. Data loss for mission-critical applications costs on average $70,913 per hour of data lost. For non-mission critical applications, data loss costs an average of $42,016 per hour of lost data (Figure 19).
Mission-critical applications
Non-mission-critical applications
Veeam Data Center Availability Report 2014 | 19 vee.am/availability14
Comparing the frequency of back-up for applications, as shown in Figures 10 and 11, this shows that a single incident can cost organizations up to $341,091 in lost data for mission-critical applications, and up to $607,551 for non-mission critical (Figure 20).
Figure 20: Maximum cost of data loss per incident (US$)
Figure 20: Maximum cost of data loss per incident, US$$341 091
$406 647
$272 749
$362 665
$268 535
$362 583
$306 928
$565 576
$302 288
$361 492
$283 102
$607 551
$456 595
$548 565
$624 410
$359 253
$973 343
$349 557
$489 868
$1 067 988
$685 955
$606 718
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
Mission-critical applications
Non-mission-critical applications
Veeam Data Center Availability Report 2014 | 20 vee.am/availability14
In total, a single incident of downtime can cost organizations $451,300 for mission-critical applications and $781,778 for non-mission critical, adding the costs of downtime and data loss (Figure 21).
Figure 22: Maximum annual cost of downtime (US$)
Figure 21: Maximum cost per incident of downtime (US$)
$781 778
$524 304
$711 565
$861 639
$438 684
$1 329 423
$500 236
$782 604
$1 230 988
$931 588
$780 211
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
Figure 22: Maximum annual cost of downtime (US$)
$10 163 114
$5 243 040
$11 385 040
$12 924 585
$4 825 524
$22 600 191
$4 502 124
$9 391 248
$16 002 844
$13 042 232
$9 362 532
AVERAGE
AU
BR
FR
DE
IT
NL
SG
CH
UK
USA
With an average of 13 incidents per year, enterprises face an average annual cost of up to $10,163,114, depending on the nature of the application and how much data is lost in each case (Figure 22).
Figure 21: Maximum cost per incident of downtime (US$)
Veeam Data Center Availability Report 2014 | 21 vee.am/availability14
These costs are likely to be a significant reason why businesses are demanding constant availability of IT services. For instance, if businesses could meet the data loss SLAs currently demanded by the Always-On Business, as illustrated in Figure 14, for both mission-critical and non-mission-critical applications, then assuming the cost per hour of data loss, as shown in Figure 19, remains
constant, the maximum risk of data loss would be $226,212 for mission-critical applications and $134,031 for non-mission critical. This represents reduced risk of at least $100,000 per incident, or $1.3 million a year, from improved RPOs alone (Figure 23). Since improved RTOs are also likely to affect downtime, the cost savings can be significant.
Figure 23: Minimum reduced data loss risk per incident from improved RPOs (US$)
Per incident of mission-critical application downtime
Per incident of non-mission-critical downtime
Figure 23: Minimum reduced data loss risk per incident from improved RPOs (US$)$114 879
$105 863
$84 304
$129 432
$181 538
$149 336
$317 494
$192 986
$94 780
$473 520
$362 706
$416 153
$486 661
$310 725
$826 580
$262 748
$371 835
$971 583
$464 385
$446 194
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
Veeam Data Center Availability Report 2014 | 22 vee.am/availability14
Indeed, if businesses could meet a target RTO and RPO of 15 minutes or less, which modern data protection tools can provide, then the savings would be significant. Again assuming that the cost-per-hour of both data loss and application downtime, as in Figure 17, remains constant, then the maximum cost of an application failure would be $38,444 (Figure 24). If unexpected downtime happens 13 times a year, this represents a maximum annual cost of $499,772, meaning a bare minimum saving of $932,945.
These statistics show the true costs of the “availability gap”, and these costs will only increase as demand for the Always-On Business grows. Businesses need to act immediately in order to ensure that the costs of the availability gap don’t grow from tens to hundreds of thousands.
Figure 24: Maximum cost per downtime incident from meeting 15-minute RTPO US$
TOTAL
UK
France
Germany
Italy
Netherlands
Switzerland
USA
Brazil
Australia
Singapore $38 444
$40 723
$44 172
$36 603
$23 749
$33 633
$45 834
$64 439
$28 470
$56 459
$32 500
$21 476
$11 686
$22 675
$22 524
$12 787
$26 533
$13 776
$30 925
$19 217
$23 146
$26 187
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
Figure 24: Maximum cost per downtime incident from meeting 15-minute RTPO (US$)
Maximum mission-critical cost
Maximum non-mission-critical cost
Part 5: Availability Solutions and Capabilities – The Root of the Issue
Veeam Data Center Availability Report 2014 | 24 vee.am/availability14
Part 5: Availability Solutions and Capabilities – The Root of the Issue
Figure 25: Capabilities organizations would like in their data center, but cannot implement (%)
An inability to support the Always-On Business ultimately comes down to a business’s legacy backup solution: without sufficient capabilities, IT departments cannot guarantee the RTOs and RPOs that the business demands.
Organizations recognize this: 92 percent of respondents identified availability capabilities they would like to have in their data center, but were currently unable to implement. These capabilities included high-speed recovery, i.e. the ability to recover any
application or server in under 15 minutes, which 60 percent of organizations wanted. Other capabilities demanded were data loss avoidance, i.e. reducing data loss to 15 minutes or less (53 percent); verified protection, i.e. guaranteed recovery of every file and application every time (47 percent); using backup data as a production-like test environment for new patches or updates (38 percent); and complete visibility, including proactive monitoring and alerting of issues before any operational impact (36 percent) (Figure 25).
Figure 25: Capabilities organizations would like in their data center, but cannot implement, %
38%
Using backup data for a production-like
test environment
60%
High-speedrecovery
36%
Complete visibility
47%
Verified protection
53%
Data lossavoidance
Veeam Data Center Availability Report 2014 | 25 vee.am/availability14
Organizations also recognized the barriers preventing them from implementing these capabilities. The primary barrier, in every case, is the cost of new technology, followed by the complexity of development or a lack of expertise; the current product not providing the capabilities needed; and human resources constraints (Figure 26).
Figure 26: Factors preventing organizations from implementing capabilities (percentage of organizations reporting factor) (%)
Cost of the new technology
Complexity of deployment / lack of expertise
Current product does not provide these capabilities
Human resources constraints
Figure 26: Factors preventing organizations from implementing capabilities (percentage of organizations reporting factor), %
5355
53
46 47
27
3640 40 40
41
33
2427
32
1513
19
27 26
High-speed recovery Data loss avoidance Verified protection Using backup for aproduction-like test
environment
Complete visibility
Part 6: An Example of Lack of Capability
Veeam Data Center Availability Report 2014 | 27 vee.am/availability14
Part 6: An Example of Lack of Capability
Figure 27: Frequency of backup testing (days) Figure 28: Percentage of backups tested each quarter (%)
An example of how this lack of capability affects data and application availability in businesses can be found in testing and verification. When a backup is made, there is always a chance that it is damaged and will not recover when needed. By testing backups, organizations can verify that they will recover correctly and that nothing will be lost. However, without the correct capabilities, verification is a time-consuming task meaning that only a fraction of backups will be verified.
Organizations test their backups for recoverability on average every eight days (Figure 27). However, each quarter, organizations only test an average of 5.26 percent of their backups (Figure 28); meaning that the vast majority of backups are not verified and so could fail.
Figure 27: Frequency of backup testing, days
8
10
11
8
9
8
2
4
15
3
11
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
Figure 28: Percentage of backups tested each quarter, %
5,26
5,1
5,4
6,5
5,8
4,4
3,5
2
10,1
4,6
5,1
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
Veeam Data Center Availability Report 2014 | 28 vee.am/availability14
Figure 29: Percentage of backups that fail to recover (%)
Figure 30: Minimum annual cost of unavoidable data loss (thousands US$)
Figure 29: Percentage of backups that fail to recover, %
17
10
26
18
16
22
10
20
26
12
17
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
Figure 30: Minimum annual cost of unavoidable data loss, thousands US$
TOTAL
UK
France
Germany
Italy
Netherlands
Switzerland
USA
Brazil
Australia
Singapore$682
$407
$1 091
$1 089
$537
$1 450
$307
$980
$907
$723
$566
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
This is borne out by the fact that 16.74 percent of backups fail to recover (Figure 29). With unplanned downtime occurring 13 times a year, this means that every year organizations’ recovery will fail twice – greatly increasing the duration, data loss and cost of downtime.
Indeed, because of these failures, data loss will cost organizations a minimum of $682,182 a year; since the best case situation is that they will have to roll back to the last valid backup (Figure 30).
Veeam Data Center Availability Report 2014 | 29 vee.am/availability14
Figure 31: Minimum annual cost of application failure (thousands US$)
Figure 32: Percentage of organizations reporting more downtime than expected when performing patches or upgrades to applications (%)
Figure 32: Percentage of organizations reporting more downtime than expected when performing patches or upgrades to applications, %
TOTAL
UK
France
Germany
Italy
Netherlands
Switzerland
USA
Brazil
Australia
Singapore87
92
100
81
71
94
72
97
87
91
92
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
Figure 31: Minimum annual cost of application failure, thousands US$
TOTAL
UK
France
Germany
Italy
Netherlands
Switzerland
USA
Brazil
Australia
Singapore $2 115
$864
$2 885
$2 558
$1 038
$4 067
$913
$3 244
$2 196
$2 953
$1 840
AVERAGE
Australia
Brazil
France
Germany
Italy
Netherlands
Singapore
Switzerland
UK
USA
Adding the cost of this data loss to the minimum average cost of downtime, enterprises will lose at least $2 million from application failure every year (Figure 31).
Testing is not only necessary for validating the recoverability of backups. Testing patches or application updates in a production-like sandbox environment before rolling them out into production can also ensure that these application patches or upgrades will perform as expected, and that the business will not suffer more downtime than expected. However, this is currently not the case. 87 percent report more downtime than expected when they perform patches or upgrades to applications (Figure32).
Part 7: Looking Ahead
Veeam Data Center Availability Report 2014 | 31 vee.am/availability14
Part 7: Looking AheadAs we have seen, organizations are well aware of the need to deliver Availability for the Modern Data Center for the Always-On Business, and that they are not yet ready to do so. Indeed, in two years’ time the landscape should look very different. 78 percent of organizations plan to change their data protection product in the next two years, with the average timeframe being six months (Figure 33).
Figure 33: Percentage of organizations planning to change their data protection product in the next 2 years (%)
Change within 6 months
Change within 1 year
Change within 2 years
No plans to change
This will be critical if business demands are to be met. It is clear that the current trend of data center modernization is not giving organizations the capabilities they need to make the always-on business a reality. Instead, IT departments need to be certain that recovery time is as short as possible; that data loss is minimized; and that backups will recover as expected, when they are needed. Without these capabilities, businesses will have no option but to bear the increasing costs of the availability gap.
Figure 33: Percentage of organizations planning to change their data protection product in the next 2 years, %
41
30
13
15FR
54
36
64 BR
33
37
723
DE
40
24
828
IT
45
22
15
18
US
63
27
37
SG
50
237
20
CH18
25
16
40
UK
2634
1624
AU
3412
16
38
NL
40AVERAGE 27 11 22
© 2014 Veeam Software All rights reserved. All trademarks are the property of their respective owners.