Professional Documents
Culture Documents
June 2012
This paper looks at the issues that mid-market IT and business managers
need to take into account when deciding how to deploy applications and
when they should consider turning to managed service providers (MSP) for
resources and advice. The document should be of interest to those who are
focused on delivering their organisation’s core value proposition, whilst
also considering how this is best underpinned by IT.
Business should focus Reliable software applications are essential to the smooth running of any business but, for
most organisations, IT is not a core skill. Selecting the right platform and location for running a
on core value given application and ensuring future flexibility involves a number of increasingly complex
choices. Many will be better off turning to partners with integration skills for advice and
resources, especially mid-market organisations that will not have the required expertise in-
house.
Physical, virtual or In the past it was common to run individual applications on dedicated physical servers; in some
cases this may still make sense. However, much of the time it makes sense to share physical
cloud? resources between applications by virtualising servers or creating private clouds. Public clouds
allow even greater economies to be achieved through sharing infrastructure with other
organisations.
Location, location, By definition, public clouds are hosted in third party data centres. However, dedicated
infrastructure can be too, either by co-locating existing equipment or making use of managed
location hosting services. Most mid-market organisations will find that the resilience and efficiency of
enterprise-class third party data centres far exceeds that of in-house facilities.
Flexible application Ensuring a given application workload can be moved from one type of platform to another is
essential to achieving many of the goals of reliable application deployment. This is the best way
deployment to ensure resilience, scalability and future proofing.
Application There is a long list of considerations for ensuring optimal deployment that will vary from one
application to another. These include how the application should be structured and its
deployment workloads broken down, what resources it needs, where it is in its life cycle and how
check list commercial components are licenced. Considerations will also vary for in-house developed and
commercially acquired applications.
Security and Two key considerations are security and compliance. Options for using certain cloud platforms
may be restricted for regulated data. That said, the perception that cloud platforms are
compliance inherently insecure is slowly being overcome. Businesses are coming to realise that the
reputation of cloud service providers is dependent on delivering higher levels of security than
many IT departments achieve in-house.
Advice and resources The choices are complex and most mid-market organisations will not have all the skills required
in-house, which is why many turn to MSPs for advice and resources. Integrator-MSPs are one-
for deployment stop-shops offering advice on end-to-end application deployment services whilst specialist
MSPs offer specific services, for example managed hosting. For an all-round service the first call
should be on the former who will engage with and integrate the latter as required.
Conclusions
All businesses are now reliant on IT to a greater or lesser extent and mid-market businesses, in particular, are unlikely to have the
full range of in-house skills needed to ensure the performance, scalability, security and compliance of their applications. Forward-
thinking mid-market organisations are turning to integrator-MSPs and benefiting from their business processes being supported by
reliable applications running on stable and flexible platforms.
For mid-market organisations this is a double-edged sword; they can more effectively compete with larger
organisations without having to build up internal IT expertise. However, entrust the task to the wrong partner and
the intended goal of delivering more reliable applications may not be achieved and the business could be derailed.
First, a casually run application with no redundancy may be a benefit for a while, but when some part of the
infrastructure that supported the application failed, the processes that now relied on it would collapse too. Second,
IT could also get expensive; running every application on its own server, sometimes with no centralised controls or
procurement, meant some applications were costing businesses more than they were saving through efficiency
gains.
The right platform, or mix of platforms, on which an application should be run is the one that best supports its
workloads at a given time. A platform should not be selected just because it is being touted as the next big thing,
but because it suits the application’s needs.
It may also make sense to leave an existing legacy application running on a dedicated physical server, at least for a
period of time, because it makes sense to wait for a longer-term release rather than upgrade to an inferior
technology on a new platform today. Furthermore, the existing hardware that it runs on may have an intrinsic value
that makes its use cost-effective for a further period of time rather than paying for something new.
Virtual platforms
Quocirca research has consistently shown that hardware servers are underutilised. This has been the driving force
behind the explosion in the last 10 years of virtualisation software for utility ‘x86-based’ servers. This has seen the
emergence of virtualisation platforms, including VMware vSphere, Microsoft Hyper-V, Citrix XenServer and Red Hat
Enterprise Virtualization (RHEV). These are all hypervisors that provide a virtual link between hardware resources
and infrastructure software.
There is nothing new about the idea of virtualisation - it was at the heart of mainframe operating systems - but what
was new was to apply it to all sizes of computers (even PCs). Virtualisation is not just about better utilisation, it is
about flexibility of deployment; for example, two applications, one running on Linux and one running on Microsoft
Windows, can share the same physical hardware whereas before they would have needed their own dedicated
servers.
Tools supplied with virtualisation software make it relative easy to take instances of applications running on
dedicated physical servers and move them, almost seamlessly, to virtual environments. Some still perceive virtual
platforms as only being suitable for non-critical workloads, but the truth is that a growing number of businesses are
now running all their applications in a virtualised environment as whole data centres are turned into virtual private
clouds.
‘Private’ cloud
Scaling up virtualisation itself, so that many physical servers appear as a single large virtual server, is, in effect, what
a computing cloud is. When this is run exclusively for the use of a single organisation then it can be termed a private
cloud. Applications are deployed to the private cloud with no need to assign resources, as the cloud will assign them
as needed, in so far as they are available. Of course, the availability of these resources is still limited by the
underlying physical platform and this will dictate how the cloud is constructed and run and the sort of applications
that can be deployed to it. Existing in-house data centres may be transformed into private clouds or cloud service
providers can provision them for their customers in industrial-scale data centres, often alongside their own public
cloud offerings.
Public cloud
Take the cloud concept and use it to build large-scale multi-tenancy platforms with hundreds, thousands or tens of
thousands of servers that can be shared by multiple organisations and you have a multi-tenant public cloud. There
are two flavours; infrastructure-as-a-service (IaaS), where the deployment is to a hypervisor (i.e. an application is
deployed with the operating system it requires to run) and platform-as-a-service (PaaS) where the public cloud is
like a huge shared Windows or Linux platform to which applications can be deployed directly.
PaaS is generally quite proprietary, for example Microsoft Azure (based on Windows), Force.com (from
salesforce.com) or Google App Engine. With IaaS, deployment is to the hypervisor and this is the way most cloud
service providers are building out public clouds. Examples include Amazon’s EC2, which is based on Xen (an open
source project that is also used by Citrix), and Rackspace Cloud Servers (based on OpenStack, an open source data
centre operating environment that supports various hypervisors).
Software-as-a-service (SaaS)
Whilst it is not strictly speaking a platform, SaaS is mentioned here as it is an option that should be considered when
thinking about deploying new applications. A SaaS application is one that is provided as ready to go over the
internet as a service. Just as with public cloud, the infrastructure is shared but there is no need to even install
application software although varying degrees of tailoring are possible. Perhaps the most obvious example is email;
why go to the bother and expense of installing a Microsoft Exchange or some other email server when you can just
Managed hosting
When a business is provisioning new applications or re-provisioning old ones, the direct purchase of the hardware is
no longer necessary. Managed hosting providers (MHP) offer pre-provisioned dedicated hardware: this is different
to public cloud because not all resources are shared, although some will be (most obviously the data centre). Some
MHPs may run their own data centres but others rent space from co-lo providers. MHP infrastructure can be used to
provide dedicated physical servers, virtual servers or private cloud.
The criticality of a particular IT service to a business is also a factor; a business cannot differentiate itself much by
running commodity IT functions such as data centres, hypervisors, databases or email servers better than its
competitors (unless it is a provider of hosted services), but it can be more competitive if it has unique IT services
that drive its core business processes (e.g. high speed share trading or a customer loyalty programme). Commodity
applications are good early candidates for outsourcing, leaving IT staff free to focus on core value. Even if a critical
application is kept in house it may make sense to outsource the management of utility parts of the stack that
supports it. Figure 4 summarises who owns the ‘stack’ for the different platform or location options.
Redundancy
As outlined above, the best way to provide a failover capability for many applications deployed on-premise will be
to have access to a cloud-based back up resource. For applications that are deployed in the cloud as a primary
platform, failover should be to a secondary cloud resource. For new applications, this means ensuring there is
sufficient flexibility to achieve this. This may be harder for legacy applications that may be tied to old hardware and
it may be a good reason for accelerating their replacement.
Peak planning
Some applications may happily run on a given platform with limited resources most of time, but require additional
resources for short periods. In the past this required over-provisioning of hardware. However, being able to make
use of additional on-demand resources only when they are needed, sometimes called “cloud-bursting”, avoids this
problem. New applications should be acquired with this flexibility in mind.
Future proofing
It is in the interests of cloud service providers to keep up with “Provisioning all
emerging technology. Even if one lags, another may be innovating.
Provisioning all infrastructure in-house limits an application to the infrastructure in-
technology stack that was available at the time it was initially house limits an
provisioned. Ensuring an application can be easily moved means that
if a given new technology can provide considerable benefits and
application to the
provide quick wins then an application can be moved to a platform technology stack that
provider that is quick on the uptake of such technology. was available at the
Understanding when this is the case will depend on the application-
specific considerations. time it was initially
provisioned”
Network
As discussed earlier, if access to a preferred
network service provider is required or there
is any special connectivity to other partners,
users or the public internet needed, this may
affect the choices made. There may also be
special considerations for access from mobile
devices as their use becomes more and more
common amongst businesses of all sizes
(Figure 5) that require particular load
balancing and data caching to be put in place.
Platform
Again, the platform options have been discussed above. As discussed, different workloads may best run on different
platforms, be they physical, virtual or cloud based.
Storage
Where data is stored is an important consideration for a number of reasons including latency, compliance and
capacity. Storage is provided as an integral part of cloud and other offerings from service providers but is also
offered as a discrete resource by many. When this is the case, storage can be provisioned separately; for example
for providing off-site backups, shared data stores etc. Similar considerations apply to those for deploying
applications when considering the suitability of dedicated versus shared storage infrastructure.
down by the Payment Card Industry Data Security Standard (PCI DSS), which limits what data can be retained and
also demands encryption. Some organisations may want to ensure their data does not become subject to the US
Patriot Act by keeping it well away from its jurisdiction.
Application structure
Much of the discussion so far has talked about applications as if they were a single entity that cannot be broken
down, which of course they can, into individual workloads and components. It may make sense to deploy different
parts of an application on different platforms. An ecommerce application may consist of a web server, an order
processing application and a transactions database. It may make sense to deploy the web server in the cloud, well
connected to customers, whilst the order processing application may be a legacy system best run on existing
hardware but co-located near the web server. For performance reasons, the database may be best deployed on a
dedicated physical server offered as a managed hosting service in a nearby facility. This will only work if the
application has been structured in way that allows such componentisation in the first place, which many will not
have been – especially older ones. Even when that is the case, it does not mean there is no choice to be made; a
more resilient application may be achieved by moving it, and the hardware it runs on, to a co-lo facility or by re-
provisioning it on a managed hosting platform.
© Quocirca 2012 - 10 -
Sourcing and integrating managed services
For current applications that are fully supported by the vendor, there will be advice available for how the application
should be run and the resources it needs. Some vendors will have adapted older applications to run in virtualised
environments or have plans to do so. Continuing with vendors that have a clear road map to support flexible choice
makes sense; if they do not, the long-term plan should be to ditch them and they should certainly not make it on to
the short list for new acquisitions.
Software licencing
Another consideration is software licencing. First, it may be the case that an organisation has perpetual licences for
certain infrastructure and applications software that need to be utilised in any long-term plan to avoid unnecessary
new investment. However, moving licences to a virtualised environment may not be straightforward and a
renegotiation with the supplier may be necessary. This is especially true as more and more software runs on multi-
core physical servers, as some vendors charge for a licence per core.
© Quocirca 2012 - 11 -
Sourcing and integrating managed services
of a dedicated hosted platform; an application vendor is hardly likely to point out that it is time to ditch your
investment in its software and switch to an alternative SaaS option instead.
Integrator-MSPs can also scale their investments across multiple customers and will have invested in the enterprise-
class management and monitoring tools that are needed to ensure applications have the resources they need at all
times to perform as required. Most mid-market organisations will not have funds to invest in such tools.
It is also likely that a given integrator-MSP will have deployed a given type of application before and many problems
that arise will have been encountered and handled previously. For most individual mid-market organisations a new
application roll out will be a first time experience, wherever it is deployed. A steep learning curve can be avoided by
turning to a partner with existing experience.
Even if a given organisation has a core application that it considers gives it truly unique business differentiation and
should be kept in-house, it could outsource the management of the infrastructure to an integrator-MSP that will put
in place an optimised bespoke service for the application and the infrastructure it relies on. As with any third party
engagement, the contract put in place should set out the desired outcomes and service levels. Of course, there
should also be a means to measure that these are being achieved.
© Quocirca 2012 - 12 -
Sourcing and integrating managed services
Conclusions
All businesses are now reliant on IT to a greater or lesser extent; mid-market businesses, in particular, are unlikely to
have the full range of in-house skills to ensure the performance, reliability, scalability, security and compliance of
the applications they rely on. To this end they should consider engaging with integrator-MSPs that have the skills
and resources they lack. This will leave them free to focus on their core business activity and avoid having to invest
in internal IT expertise.
© Quocirca 2012 - 13 -
Sourcing and integrating managed services
References
1. Quocirca “2012 – The year of Application Performance Management” – February 2012, interviews with 500 UK,
USA, German and French IT managers. To be published on www.quocirca.com later in 2012, freely available on
request from bob.tarzey@quocirca.com or downloadable at:
http://applicationperformance.dynatrace.com/2012_Application_Performance_Management_Outlook_Survey.
html
2. Goldman Sachs Global Investment Research; A paradigm shift for IT: The Cloud – November 2009
3. Quocirca “Outsourcing the problem of software security” – February 2012, interviews with 100 UK and US
enterprises. To be published on www.quocirca.com later in 2012, freely available on request from
bob.tarzey@quocirca.com or downloadable at:
http://info.veracode.com/Quocirca_Outsourcing_Software_security.html
4. Quocirca “The data sharing paradox”, Quocirca Sept 2011 – http://www.quocirca.com/reports/620/the-data-
sharing-paradox (SMB data from UK, France, Germany USA and Australia)
5. Quocirca; UK-only data taken from research published in “Next Generation Datacentre Cycle II – Cloud
findings”, April 2012 http://www.quocirca.com/reports/689/next-generation-datacentre-cycle-ii-cloud-findings
© Quocirca 2012 - 14 -
About niu Solutions
niu Solutions specialises in building bespoke enterprise IT and communications, as organisations need it, as-a-
service, and as they grow.
Harnessing the power of cloud technology and knowhow gained from over 20 years of delivering managed and
hosted services to the mid-market, niu provides tailored integration services that introduce resilience, security and
compliance as standard.
niu’s professional services team create high-performance IT environments integrated across cloud or private hosted
services, or in house services, on an application by application basis. Focussed on delivering applications and
infrastructure that enable more efficient business processes, niu tightly integrates platforms, processes and
technologies from best of breed providers and delivers them as a single, tailored solution, with a bespoke and
flexible service wrap.
Taking the time to understand an individual organisation’s challenges is key to niu’s service values, as this quickly
translates to a valuable competitive edge for the customer’s business. Organisations benefitting from niu’s bespoke
service delivery include MetroBank, Keystone and JD Sports.
Headquartered in the City of London, niu employs more than 100 people and has offices throughout the South East
of England.
Sourcing and integrating managed services
Quocirca has a pro-active primary research programme, regularly surveying users, purchasers and resellers of ITC
products and services on emerging, evolving and maturing technologies. Over time, Quocirca has built a picture of
long term investment trends, providing invaluable information for the whole of the ITC community.
Quocirca works with global and local providers of ITC products and services to help them deliver on the promise that
ITC holds for business. Quocirca’s clients include Oracle, Microsoft, IBM, O2, T-Mobile, HP, Xerox, EMC, Symantec
and Cisco, along with other large and medium-sized vendors, service providers and more specialist firms.
Details of Quocirca’s work and the services it offers can be found at http://www.quocirca.com
Disclaimer:
This report has been written independently by Quocirca Ltd. During the preparation of this report, Quocirca has
used a number of sources for the information and views provided. Although Quocirca has attempted wherever
possible to validate the information received from each vendor, Quocirca cannot be held responsible for any errors
in information received in this manner.
Although Quocirca has taken what steps it can to ensure that the information provided in this report is true and
reflects real market conditions, Quocirca cannot take any responsibility for the ultimate reliability of the details
presented. Therefore, Quocirca expressly disclaims all warranties and claims as to the validity of the data presented
here, including any and all consequential losses incurred by any organisation or individual taking any action based
on such data and advice.
All brand and product names are recognised and acknowledged as trademarks or service marks of their respective
holders.