IPv6 Industry Survey Results

painlosososΛογισμικό & κατασκευή λογ/κού

30 Ιουν 2012 (πριν από 5 χρόνια και 4 μήνες)

304 εμφανίσεις

 
 
BT Connect

IPv6 Industry Survey Results
June 6, 2012

 
 
IPv6 deployment has begun in earnest!
Many organizations have taken action to
deploy IPv6 in some form since last
year’s survey. Thirteen per cent of survey
respondents have taken such action,
while only five per cent had done so in
last year’s survey.

The number of respondents also grew by nearly 50 per cent
this year to 876, so the 13 per cent represents a significant
increase in deployed IPv6 networks. Another 20 per cent
are engaged in the process of IPv6 implementation as of
now, and 24 per cent plan to begin deploying IPv6 within
two years. So if all goes according to today’s plans, within
two years time, 57 per cent of respondents will have
deployed or will be in the process of deploying IPv6. An
additional 16 per cent of respondents are currently
assessing the costs and benefits of deploying IPv6. The
remaining 27 per cent have either decided not to begin
deployment within the next two years, feel IPv6 is
unnecessary or will follow a bottom-up end-user demand-
driven (“BYOD”) approach to IPv6 deployment.


Despite the progress in IPv6 deployments since last year,
many organizations still face obstacles. Leading among
these was the inability to demonstrate a strong business
case with 22 per cent of responses. Other leading obstacles
included the complexity of infrastructure upgrades,
perception that the only benefit of IPv6 is larger address
space and the cost of equipment upgrades.

While many enterprises plan to deploy IPv6 only on Internet-
facing servers initially, most plan full deployment throughout
their networks as a follow-on goal. The most common or
planned approach to IPv6 deployment was dual-stack for
both enterprises and service providers.














Key Findings


Fifty-seven per cent of respondents have deployed,
are deploying, or plan to deploy IPv6 within two years.
Thirteen per cent of respondents have already
deployed IPv6 on all or a portion of their networks.
This is up sharply from five per cent who had deployed
in 2011. Another 44 per cent are in the process of
deploying or plan to begin deployment within two
years.


While respondents believe value and benefits may be
realized from IPv6 deployment, stronger business
case justification is still needed to demonstrate
sufficient ROI.


Along these lines, leading among the hurdles to IPv6
deployment, the business case trumps complexity.
The biggest hurdle identified was the inability to
demonstrate a clear business case with a 22 per cent
response, followed by the complexity of the
infrastructure upgrade with 17 per cent, which is down
from prior survey results in the 22-28 per cent range.


The initial shock and ensuing reaction to IANA’s IPv4
address depletion announcement has apparently
waned with the proportion of respondents expressing
a “huge concern” about IPv4 depletion dropped back
to 25 per cent from 35 per cent last year. The 2011
result of 35 per cent was more than double the result
from 2008 at 16 per cent and was likely driven by the
proximity of the announcement only two months prior
to the survey window last year.


Most organizations will not be satisfied simply
deploying IPv6 on a portion of their networks but will
look to deploy more broadly as 55 per cent of
respondents agreed with the statement:
IPv6 is
required for deployment across my entire network.



The dual-stack deployment approach led other IPv4-
IPv6 co-existence strategies that enterprises and
service providers are utilizing or planning to utilize.
The transition to IPv6 is
one of the most important
steps we will take
together to protect the
internet as we know it.”
Vint Cerf
Google’s Chief Internet Evangelist

 
 



IPv6 deployment has begun in earnest with 57
per cent of our IPv6 survey respondents
indicating deployment within two years. Internet
ubiquity is at stake and organisations need to
plan for IPv6.”

Tim Rooney, BT Diamond IP product management

 
 

















Introduction

In early spring, 2011, BT Diamond IP conducted the third
of a then-triennial survey regarding opinions about IPv6
deployment and relative merits. This survey nearly
coincided with IANA’s announcement of the allocation of
its final /8 blocks of IPv4 space to Regional Internet
Registries (RIRs). During the survey interval, the APNIC
RIR serving the Asia/Pacific region, which received the
largest of these final allocations, had already exhausted its
IPv4 address space for all intents and purposes. Results
of this survey understandably indicated a large jump in
concern about IPv6.

To help determine the after-effects of these major events a
year later, BT Diamond IP moved to an annual format and
conducted a web-based survey on its website from April
23, 2012 through May 15, 2012. The survey was
completed by 876 IT or Operations professionals from
around the globe and spanning multiple industries. The
survey was posted on www.btdiamondip.com and
invitations to participate were sent to individuals identified
as IT and Operations professionals throughout the world.

All survey responses were automatically tabulated into a
survey tool. Any individual skipped questions were not
included in tabulations. Each chart highlighting unique
responses in this report includes the number of valid
responses for that particular question (e.g. n=500
indicates 500 responses). Percentages shown in charts
may not equal 100 per cent due to rounding.





IPv6 Level of Concern


Last year’s survey was conducted about two months
after IANA’s announcement of the allocation of its final
/8 blocks of IPv4 space to Regional Internet Registries
(RIRs). As such, concern was extremely high given
this “sudden” announcement of IPv4’s demise. This
year’s survey indicates a leveling off of this “reaction”
but also points to real movement towards IPv6
deployment.





In 2011, 35 per cent of respondents expressed a huge
concern and another 46 per cent indicated moderate
concern. This year’s results yielded 25 per cent and 50
per cent response rates respectively. A view of this
across industry verticals as identified by respondents is
shown in Figure 2, ranked in increasing order of low
concern. As one might expect, nearly all service
providers showed at least moderate concern with only
11 per cent expressing low concern; but perhaps these
have already deployed IPv6! The other verticals
besides education/non-profit expressed low concern in
the upper 20’s percentage-wise.

From a regional perspective, respondents from Europe
and Asia answered nearly identically, with about 30 per
cent expressing huge concern and about 50 per cent
moderate concern, which topped averages in other
regions of 18 per cent and 48 per cent respectively.
Organization size also influenced responses, with
those from larger organizations expressing larger
concern.




Figure 1: Concern about IPv4 address exhaustion (n=876) 
 
 
IPv6 Deployments


Evidently concern expressed in last year’s survey manifested itself with a jump in IPv6 deployments reported this year. Thirteen
per cent of survey respondents have deployed IPv6 on all or part of their networks, while only five per cent had done so in last
year’s survey. The number of respondents also grew by nearly 50 per cent this year to 876, so that 13 per cent represents a
significant increase in deployed IPv6 networks. Another 20 per cent are engaged in the process of IPv6 implementation as of
now and 24 per cent plan to begin deploying IPv6 within two years. So within two years time, 57 per cent of respondents
reportedly will have deployed or will be in the process of deploying IPv6. An additional 16 per cent of respondents are currently
assessing the costs and benefits of deploying IPv6. The remaining 27 per cent have either decided not to begin deployment
within the next two years, feel IPv6 is unnecessary or will follow a bottom-up BYOD approach to IPv6 deployment.

Figure 3: IPv6 deployment status (n=876)
Figure 2: IPv4 address exhaustion across verticals
 
 

Figure 4: IPv6 deployment status by year
While Figure 3 illustrates the percentage breakdown of responses from this year’s survey, Figure 4 provides a historical
perspective. Interestingly, not only have more deployments begun, but a decreasing percentage of respondents has not
considered or assessed IPv6 over the years as well.

It’s instructive to illustrate deployment status by vertical, region and organization size as well. Figure 5 depicts results by vertical,
ordered by the sum of the top three criteria = {already deployed + in the process of deploying + will begin within two years}.
Evidently, government organizations lead the way, followed by service providers, education/non-profits and finally enterprises.
This is not surprising given the spate of government mandates over the last decade, and the requisite relationship between IP
addresses and subscriber revenue for service providers. Multinational enterprises lead their generally smaller, more localized
cousins, national/regional enterprises.

Figure 5: IPv6 Deployment Status by Vertical
 
 

 
Figure 6: IPv6 Deployment Status by Region (n=738)
Figure 6 shows results for IPv6 deployment status by region, ordered alphabetically. Respondents from Europe seem to be a bit
ahead of those in other regions, while those from Asia, North America and Middle East/Africa trail slightly. Unfortunately, the
data set for Central/South America is statistically insignificant but is shown for completeness.
Figure 7 shows that the larger the organization in terms of IP addresses, employees or subscribers, the further along in IPv6
deployment they are in terms of actual deployments completed, in-progress, or planned. This of course makes sense in that
these large consumers of IP addresses would be among the earlier adopters to support their larger capacity requirements.

 
Figure 7: IPv6 Deployment Status by Organization Size (n=738)

 
 








IPv6 Non-Deployments

We asked those respondents who had no plans to deploy IPv6
what steps they were taking, if any, to support IPv6
communications. Responses were rather evenly split among
taking no steps, relying on Internet Service Provider (ISP)
translation services, implementing in-house translation services,
tunneling and to a lesser degree, explicitly disallowing IPv6
communications.

Figure 8: Mitigation steps for non‐deployments 

Perceptions about the need for IPv6

We asked about respondents’ opinions
about the value of IPv6 within their
organizations and for the Internet at
large. Unfortunately, some of these
questions were posed in the negative
sense in keeping consistent with the
first 2005 survey. So some of the
commentary in this section cancels out
the double negative by inferring “most
agreed with the affirmative” instead of
the technically correct “most disagreed
with the negative” format.

For example more respondents
disagreed than agreed with the
statements that IPv6 has value but does
not link to business drivers and that
IPv6 does not provide any benefits to
my infrastructure or organization. We
infer from this that more respondents
agreed than disagreed that IPv6
provides benefits and offers business
value. On the other hand, more
respondents agreed with the statement
that IPv6 deployment does not offer a
strong enough ROI. We conclude that
while respondents recognize benefits
and value, these are not yet sufficient to
produce a strong return on the
investment required for deployment.

Figure 9 illustrates the overall results
regarding the need and value of IPv6.
Over 60 per cent of respondents agreed
or strongly agreed that IPv6 deployment
is required to communicate with IPv4
and IPv6 Internet users. Only about 18
per cent disagree or strongly disagree
with this statement. As a consistency
validation, the converse statement that
it is not necessary to implement IPv6
yielded nearly proportional contrary
results.
 
 

Figure 9: IPv6 perceived value (n= 876) 
Historical perspectives on IPv6’s value from past surveys are summarized in Figure 10. The rating scale in this figure was
devised by assigning values of one to five for strongly disagree, disagree, neutral, agree and strongly agree respectively. Hence
a value of “three” indicates a neutral average response, while above three indicates agreement and below 3 disagreement.
Responses from 2011 and this year are nearly identical, but fewer respondents agreed that the ROI for IPv6 is not strong
enough and that IPv6 does not link to business drivers than in 2008 or 2005.
 
 
Figure 10: Historical IPv6 perceived value
 
 

In terms of IPv6 features, Figure 11 summarizes respondents’ value ratings over this and past surveys, with a rating of five
being most valuable. As you can see, responses have consistently rated these attributes above neutral. In reality, security,
quality-of-service and flow labels offer equivalent value in both IPv4 and IPv6, while the other attributes do offer contrasts in
protocol operation. Expanded address space is certainly a unique advantage of IPv6, while improved mobility with more
efficient routing, address autoconfiguration, efficient packet routing with fragmentation performed on the perimeter and the
simplified header structure likewise offering feature improvements.


Figure 11: Historical IPv6 features ratings 
 
 
IPv6 Deployment Approaches
We asked survey participants what techniques they have used or plan to implement in support of IPv6 deployment, for service
providers vs. enterprise respondents. Figure 12 illustrates responses for this survey contrasted with those from last year for
service providers. Most answers appear roughly the same as last year, with the exception that segmented dual stack
approaches on the backbone only or customer-facing only, have seemingly been joined to a larger proportion of service
providers supporting dual stack throughout their networks. Note that multiple responses were permitted to these particular
questions about deployment techniques, which tends to level out results to some degree.
 
Figure 12: Service provider deployment approaches 
Enterprise respondents likewise favored dual stack as the deployment technique of choice as illustrated in Figure13. A larger
proportion of respondents indicated constrained deployment to Internet facing servers this year than last, while fewer have no
plans to deploy or to fully deploy IPv6.
 
Figure 13: Enterprise deployment strategies
 
 
IPv6 Deployment Obstacles
We asked survey participants two questions regarding obstacles to deployment. The first asked about the biggest hurdle to
overcome when proposing IPv6 deployment within the organization. As Figure 14 indicates, the inability to demonstrate a
strong business case was the top response with 22 per cent of respondents.

This top answer was actually a new choice in this year’s survey, as was the “we have overcome all hurdles” i.e., none of the
above. Figure 15 shows how responses to this question have varied over past surveys. Interestingly, there has been a steady
decrease in the obstacles of equipment upgrade cost, application conversion, network services support, training, and security
product availability. Evidently IPv6 awareness is growing and vendors are rolling in IPv6 support over time.


Figure 14: Single biggest obstacle to IPv6 deployment (n=761) 
 
 

Figure 15: Historical IPv6 deployment obstacles 
The second question relating to IPv6 deployment obstacles asked what would most help in convincing the organization to
deploy IPv6. Results are summarized in Figure 16. The top answer was a guidebook covering deployment approaches and
processes with coverage of address planning, IPv4-IPv6 co-existence, security and network management
1
. Case studies
showing a positive ROI was next most popular, followed by a government or industry mandate, IPv4 address exhaustion itself,
and case studies illustrating how IPv6 helped an organization achieve a competitive advantage.

Figure 16: Most helpful in overcoming IPv6 deployment obstacles (n=761)
                                                           
 
1

Members of the BT Diamond IP team are currently working on such a book, and availability will be announced closer to publication.

 
 
IPv6 Deployment Threshold
This year we asked whether organizations considered a
threshold in terms of the proportion of the Internet that was
“IPv6-enabled” that would impel them to deploy IPv6 with
some urgency. Figure 17 reflects results, which vary widely,
though most organizations either have already deployed,
are deploying, or have no set threshold but will await
industry and business conditions to warrant deployment.

Figure 17: IPv6 density threshold for triggering deployment 
(n=741) 
Conclusions
IPv6 has generated substantial momentum within the
industry with a majority of survey respondents indicating full
or partial deployment or plans for deployment within two
years. There is no deadline for IPv6 deployment but the time
for planning your IPv6 is now.
For more information about how BT can help please visit
ipv6.bt.com. To learn about BT Diamond IP and IPv4-IPv6
address management solutions please visit
www.btdiamondip.com.
Survey Demographics


Figure 18 summarizes survey respondent demographics.
Geographically, 54 per cent of respondents indicated they
were from North America, 28 per cent from Europe, 14 per
cent from Asia, 3 per cent from Middle East/Africa and 1 per
cent from Central/South America. This geographic
distribution represents a shift with more respondents from
Europe and Asia and fewer from other regions compared to
prior years’ IPv6 surveys.
.

Figure 18: Survey respondent locations (n=738) 
.
From a network sizing perspective, Figure 19 illustrates that
39 per cent of respondents each managed networks of less
than 10,000 IP addresses and 30 per cent between 10,000
and 100,000 addresses. Twenty-one per cent of
respondents managed networks of 100,000 to one million
addresses and 10 per cent managed networks larger than
1million addresses. This represents nearly proportional
respondents from the mid-ranges with more smaller
organizations and fewer larger “one million plus” IP address
organizations as compared to last year’s results.


Figure 19: Survey respondents’ organisation sizes (n=738)
 
 
Selected Verbatim Comments

To add color regarding respondents’ perspectives, this
section lists selected anonymous comments entered in
various free form text areas of the survey:
“IPv6 will be required in the future, because the use of NAT
with IPv4 will not scale to enterprise levels, it will cause
performance bottlenecks. Also NAT is NOT a substitute for
security.”
“IPv6 is not the future, anymore. It must be NOW.”
“Address management is going to be the most difficult part
of deployment. Unless a responsive DNS structure is in
place (one that has mechanisms to gather information about
hosts on a LAN and apply some kind of naming intelligence
to it), it will be a disaster.”
“I believe there are unforeseen consequences with
implementing IPv6 across the board.”
“For years, regulators and Internet experts have warned
about the exhaustion of IPv4’s limited pool of addresses.
Basically more new devices, platforms and services
incorporate support for IPv6, but so far mass migration has
been delayed. The lack of public understanding and general
fears of the difficulty and complexity of the migration will add
to the slow pace of IPv6’s adoption.”
“IPv6 is a necessity on Internet facing services in the short
term. In the long term I see network equipment vendors
reducing feature support for IPv4 as they focus on IPv6 for
new products. This will make a full transition to IPv6 pretty
much mandatory to stay current with vendor support.”
“IPv6 has been "two years away" for over 10 years. The
infrastructure challenges IPv6 would have solved elegantly
for my company have been solved through necessity on
IPv4. We will deploy IPv6 for our public facing services
when there is customer demand.”
“As a protocol IPv6 is fine. The problem is that all the other
tools and infrastructure requirements aren't quite there yet.”
“Critical business need is to support access to our public-
facing websites from external, IPv6-only clients, prospects,
etc.”
“IPv6 is inevitable - the question is when will it obtain the
critical mass necessary to overcome inertia and resistance
to change.”
“No business drivers is what keeps getting thrown back at
us. We have done nothing other than some rudimentary
testing and this is by our mainframe networking guys!”
“Seems to me that a lot of people do not realize the need
and impact when ipv4 runs out. For many organizations ipv6
will never have a ROI.”
“IPv6 is like taxes and death. You cannot avoid it.”
“IPv6 requires specific aspects not only to IPv6 addressing
and IPv6 address allocation, as your survey might suggest.
It also requires IPv6 DNS, Domain registration, routing
processes, coexistence IPv4<->IPV6 on all layers, etc. To
implement this kind of things is one side of the story; to
operate such an environment is a very big challenge at the
end of the day.”
“Think business RISK”
“Internal to our network since we use private address space,
IPV6 is not mandatory.”
“Too many companies are not taking IPV6 seriously
enough.”
“My life (and our network) would be easier if we were
completely "IPV6 ready". I am tired of shuffling RFC1918
blocks between amalgamated business units and
configuring NAT/PAT on our Firewalls. Most of our
equipment is ready. Most of our staff isn't.”
“Because of our footprint in Asia, we are looking
aggressively toward plans for IPv6.”
“Do it, now!”
“Don't do it, we don't need it yet.”
“There is still a lot of training that is required and legacy
applications that need to be re-written. Also due to address
structure device table sizes will be impacted which could be
costly.”
“One of the biggest overall concerns with rolling out IPV6 is
having to support parallel networks (IPv4 & v6) and the
added load & security risks/threats that potentially doubles
the effort/workload on infrastructure & human resources.”
“My organization has a fully functional IPv6 network. We
wish some of the IPv6 tools were mature...or even available.
But we have been running for nearly two years with no real
trouble.”
“IPv6 is highly dependent on who a company's clients and
partners are.”
“We have a need to implement IPV6 for Internet facing
servers primarily. We must be able to support access for
students and faculty from International and Mobile sources
which have a business requirement to access us.”
“Inside the Enterprise, nothing is mature enough yet. It is
Y2K level of work with no justification.”
“Researching IPv6 for internal data center use initially to get
experience and then consider expanding.”
“Deployment will be a multiyear project in large
organizations. We will focus on everything publicly facing,
with plans for deployment this year. Deployment in the
organisation will take years, due to dependencies on IPv4
for a lot of peripherals, and the desire to keep the period to
maintain a dual stack infrastructure as short as possible.”
“Migration-Plan: IPv4 -> establishing an internal translation
service -> IPv4/IPv6 dual-stack beginning at internet edge -
> establishing an internet-service translation service ->
unconfiguring IPv4 internal”
“Windows7 and Server 2008 natively speak IPv6 and when
you run it over an IPv4 everything gets tunneled and you
loose visibility and control.”
“Management supports external-facing Internet
infrastructure. Expending funds for internal-only
implementations would require additional business
justification.”
“I do not see IPv6 replacing IPv4 with NAT in small to
medium sized businesses in the foreseeable future.”
“The faster it happens, the better for the world.”
 
Offices worldwide
The services described in this publication are subject to availability
and may be modified from time to time. Services and equipment
are provided subject to British Telecommunications plc’s
respective standard conditions of contract. Nothing in this
publication forms any part of any contract.
British Telecommunications plc 2012.
Registered office: 81 Newgate Street, London EC1A 7AJ
Registered in England No: 1800000