The 'Cloud' is everywhere it seems, in the media, on the web and
bandied about by IT companies left right and centre. All the
cloud-vendors seem to agree that cloud is "the future of
computing". But what actually is cloud computing? Is cloud simply
another cunning industry buzzword for selling more servers, or is
cloud computing really a 'paradigm shift' in the way we access
computing resources?
Arguably it's a bit of both.
To understand if cloud computing does represent an evolution - if
not a revolution - in our thinking on the way we process digital
information, as well as ideas of data ownership versus access, we
need to first understand the fundamentals of what cloud computing
is and then examine how this might benefit a business.
In June 2009, IT market analysts Gartner identified the five key
characteristics of cloud computing:
Service-based - Consumers subscribe to services designed to
directly fulfill specific needs.
Scalable and elastic - Computing resources and services can scale
up or down depending on the requirements of the consumer.
Shared - Services leverage a common platform or infrastructure to
maximise the efficiency of resources.
Metered-by-use - Consumers are presented with various payment
models based on the usage of services required rather than the cost
of equipment.
Uses internet technologies - Services are delivered using the
internet, typically internet protocol (IP) as the service delivery
mechanism.
In essence, a range of IT services and resources such as software,
desktop environments, networks and communications systems are
delivered directly to the consumer over the internet, from a
central data centre. While this approach is sometimes called
'infrastructure-as-a-service', it does not mean an end to keyboards
or monitors. Rather the software, servers, networking equipment,
storage and backup space can all be 'virtualised' - you simply
create your desired computing environment or infrastructure 'in the
cloud'.
By separating the physical hardware from the services you can
accurately measure the consumption of those services and therefore
be charged accordingly. As a consumer you only need to pay for the
resources you need at any given time - you don't need to buy all
the physical computing hardware to support your computing
needs.
Consumers of electricity don't rush out and buy a power station.
They purchase and consume the amount of power they need to go about
their daily lives; they rarely worry about how that electricity
gets to them. Why should computing be any different? Why should you
need to invest in computing infrastructure when you can simply
purchase what you need to get the job done and stick it on a
utility bill. Does all this sound familiar? The idea of computing
as a utility is not new. John McCarthy, then a researcher in
artificial intelligence at MIT, stated in 1961 to the MIT
centennial conference that: "If computers of the kind I have
advocated become the computers of the future, then computing may
someday be organised as a public utility just as the telephone
system is a public utility. … The computer utility could become the
basis of a new and important industry."
Nearly 50 years on, we have seen the growth of mainframes,
networked terminals, networked computers, the internet, ASPs, MSPs,
SaaS and so on, each a stepping stone on the path to achieving true
utility computing. One could argue that the emergence of cloud
computing is the first large-scale implementation of public utility
computing - it neatly fits together the many different complex
parts of the puzzle and presents them to the consumer in
easy-to-swallow metered-by-use chunks.
Cloud computing itself is often confused with SaaS
(software-as-a-service), which is hardly surprising as SaaS
solutions do display the characteristics outlined above with one
exception. SaaS delivers specific applications to the end-user's
desktop via a web browser, the application functionality is
rendered as a web page and then controlled through that webpage.
Google Docs and Salesforce.com are examples of this approach. Cloud
is somewhat different, while still using the internet as the
service delivery layer, it can deliver the entire environment to
the end-user, that is operating system, applications,
communications (such as VoIP) and networking functionality. This
means that existing and well-established software applications such
as Microsoft Office can be delivered as a service, even the
operating system itself can become a service. To really understand
the difference simply fire up Google Docs and attempt to install
your own application on there - you can't. With a cloud hosted
desktop, you can.
What about MSPs (managed service providers) and ASPs (application
service providers)? Aren't they the same thing as cloud? They are
very similar in essence but again not the same. Typically, an MSP
will lease dedicated server systems to a company which will either
be on-premise or in a server facility somewhere. They then provide
a management layer and support services. This means that the
consumer is effectively locked to that supplier and their hardware
and scalability means pulling the box out of the rack and inserting
more RAM. An ASP is slightly different again. Many ASPs have
evolved into SaaS and MSPs, but pure-play ASPs deliver specific
applications to the end user usually over a terminal session - once
again application-based rather than service-based.
Cloud computing, in principle, is not dissimilar to the mainframes
of the past - huge computers accessed by dedicated networked
terminals. Nowadays the huge computer lives in a data centre, 'the
cloud' - the terminal is your PC, thin client or mobile device and
you access your services over the internet. A simple way of
understanding cloud is this: All those servers and cables and boxes
and IT junk cluttering up your office space, and sucking up your
management time making it all work, are gradually replaced by a big
computer somewhere else that you never need to see or worry about.
And in the same way that electricity is agnostic - you can use
electricity to power a refrigerator, television, toaster or a lava
lamp - raw computing power in the form of processor time, RAM and
storage is agnostic too. It can be your operating system, telephone
exchange, backup system or anything else you might need. Remember,
the services consumed are distinctly separate from the device
powering those services.
Xhead: Benefits of cloud computing
An example of how cloud can directly benefit a small business is
where that business has a several locations and between five and 15
computer users at each location. Chances are the business is too
small and the locations too dispersed to employ a dedicated IT
manager, which means that the business is supporting two or three
independent networks, backup file servers and the like using IT
support contractors. The challenge is to maintain consistency and
security across each individual business unit while still
guaranteeing access to the services needed to run the business. In
this scenario cloud looks attractive - cloud services are not
location dependent, this small business could fulfil its entire
computing requirements directly from the cloud and the business
could function as a single unit from the IT perspective, rather
than the existing dispersed information silos. At the grass roots
level, cloud only needs terminals - existing PCs are fine for this
- and internet connectivity.
The main business benefit, in addition to cost savings, is that
cloud-based IT services move the IT budget from a capital expense
to an operational expense.
This means:
The cost of ownership can be managed as part of monthly
accounts
There is no upfront capex
Much lower management and maintenance costs
Service continuity
The move to cloud is non-disruptive. From a consumer's perspective,
nothing actually changes: they sit down at their desk, fire up
their computer and there are all the applications and services
exactly as before, the only difference is that their desktop and
applications are now sitting on the big computer somewhere else,
not under their desk.
As hardware becomes redundant in the business, computers can be
replaced by thin clients. These are small units that you plug a
keyboard and mouse into, then connect to the internet. No moving
parts or in-built functionality means that the shelf life of these
boxes is considerably longer than a regular PC and they consume a
fraction of the power.
When the next whizz-bang operating system or fancy
'solve-all-your-problems-at-the-push-of-a-button' application comes
along, businesses no longer need to worry about hardware or
software upgrades - they simply update their cloud services.
Xhead: Drawbacks of cloud computing
Naturally, managed services providers, SaaS players and IT support
companies are not entirely happy with cloud computing. Nor probably
are IT departments, whose main activity - much to their chagrin -
is the maintenance and upgrading/updating of their in-house
computer resources. (Some authorities put maintenance up to 80 per
cent of IT staff's activities.) Cloud might free up most of their
time for more 'productive' pursuits, but it also might involve many
of them suddenly found without a job.
SaaS players are nervous because shifting end-user operating
systems from local to remote hardware mitigates many of the
benefits promised by these web-based applications. Managed service
providers probably have the most to fear as their entire business
model is built around leasing hardware and providing services that
they can then manage. However, it is likely that MSPs will embrace
cloud when the consumers demand elasticity and scalable computing
and no longer want to be tied to lengthy leasing agreements - they
are also in a good position to provide these services as they have
expertise and existing infrastructure.
At the consumer level, the main concerns are service availability,
bandwidth and security.
Service availability - what happens when the internet goes down?
Are the users cut off from their 'virtual desktops'? Does all work
stop?
The same principles that work at the data centre end can be applied
at the consumer end - failover. At a moderately low cost, a second
broadband connection can be installed at the consumer's premises.
With the right router, load balancing can be achieved over two
connections; if one drops out, the other takes up the slack.
Admittedly a consumer may see a slight drop in the quality of
service until the primary line is restored, but not the complete
failure of services envisaged by sceptics. Additionally, because
the services are on a remote computer if there is a power cut, an
internet drop-out or a wombat chews through the cable, when you do
regain access your services come back exactly as you left them with
no loss of data.
Bandwidth - as the internet connection is used extensively, there
are also some concerns that bandwidth and data transfer may become
an issue. In fact, this is quite the opposite. In many cases the
data transfer rates actually drop. Because you are accessing data
at the data centre, the only information that is passing through
the net are highly compressed screenshots of your remote computer -
the data itself never actually moves from the cloud.
But does sharing computing hardware with other businesses pose a
threat to data?
Security is the perennial IT challenge and a multi-billion dollar
industry. With hardware shared between businesses there is a
perception that you may be able to gain access to other companies'
data or that a virus could spread far more rapidly affecting
thousands of virtual machines in the blink of an eye. In fact, the
nature of virtualisation is such that each machine exists in its
own 'virtual container'; if one crashes it simply crashes the
container, not the whole system. Furthermore, applying group
application policies and security updates over a range of devices
becomes far simpler when they are centrally managed in a cloud.
Instead of relying on an individual user or support technician to
update security policies, this can all be automated from a
centralised control panel. Push a button and all the desktops are
magically updated. While enterprise systems have had this ability
for some time, it is only recently that these benefits have become
available to smaller businesses without the capital or expertise
needed to implement autonomous IT infrastructures.
Xhead: Cloud computing today
The cornerstone of cloud computing was laid as early as 2001 when
Microsoft launched its service provider licensing agreement (SPLA).
The agreement allowed services providers to commercially host
Microsoft products as well as re-sell them on a pay-as-you-go
basis. This meant that vendors could effectively 'rent' software
without having to cover the licensing costs upfront. The hosting
industry saw immediate benefits, leading to a drop in the price of
hosting and broadening of services on offer.
Meanwhile, a number of software development companies were
developing and refining their virtualisation platforms - software
that would allow creating virtual containers to host specific
applications and operating systems independently on the same
computer. Unfortunately, many of the cost-benefits of approaching
computing in this manner were eroded by the high cost of the
virtualisation software itself. Enter Microsoft again with Windows
Server 2008. One of the new enterprise features was extremely
low-cost virtualisation, effectively levelling the playing field.
The Server release, allied with improved broadband connectivity,
cost-effective virtualisation and the ability to license products
on a monthly basis, meant that cost rather than management becomes
a strong motivating factor. This means that small and medium
businesses could now afford to get their hands on a piece of the
cloud computing pie.
Even with over 90 per cent of Australian businesses having
broadband access, few of these businesses use the capacity for
running line-of-business services. Typically, they use it for basic
operations like email and web-browsing. Line-of-business
applications are either running locally on the user's machines or
on an application server within the corporate network.
Cloud offers them the direct benefits of optimising their use of
the net and simultaneously minimising in-house IT resources.
However, current players in Australia only have different pieces of
the puzzle - until recently nobody has supplied the full range of
services. The massive costs involved in setting up Australian-based
data centre infrastructure capable of delivering cloud-based
services effectively locks smaller service providers out of the
market, so it is hardly surprising that early cloud provisioning
has been limited to the carrier end of the industry where cloud
services are bolted on to existing service offerings. While
Australian businesses can access IP-based services from abroad, the
latency of such distant services (or the amount of time it takes
for a piece of data to get from A to B and back again) becomes an
issue, so adoption en-masse is yet to occur. As cloud services do
become available locally, we expect the business uptake to be a
rapid one.
Xhead: The future of cloud computing
Given that the big players are very active in the delivery and
creation of a new cloud-based service delivery landscape would
suggest that the future is very bright for cloud. The increasing
capacity of the local broadband network is also a strong indicator
that the carriers see a big future in IP and a big future in IP
means a big future for cloud computing. Gartner itself predicts
that by 2020 over 40 per cent of the world's IT services will be
delivered as cloud-based services - that's a whopping big chunk of
the IT market.
Is there life after the cloud? Will there come a day where
everybody buys their computing resources like they currently buy
electricity? While the benefits to business consumers are obvious,
there are less immediate or compelling benefits to the regular home
user - this may explain why the obvious candidate for piping cloud
services into the home, cable, is still largely used for providing
the internet connectivity itself rather than directly delivering
computing resources.
While the business community will get the early adopter advantage
of embracing cloud, over time we will see greater competition which
will lead to vendors seeking new ways to package and sell cloud. At
time of writing cloud services are available, however many vendors
still tie the consumer in to contracts. Sometimes, if you look
under the hood, some systems currently sold as cloud are really
just dedicated servers, which rather negates the benefit of having
a 'metered-by-use' service. However, it is understandable as the
definition of cloud computing is still somewhat fluffy. The next
generation of true cloud services will be far more granular in
their measuring and billing, much as mobile phone providers are
today - pretty soon we will be talking about 'processing minutes'
and 'dedicated RAM time', one line on your statement might read
"User 3 - Microsoft Word Access - 46minutes - $4.60".
Maybe we are finally catching a glimpse of that shifting paradigm
we've been promised for so long.
Simon Probert is the managing director of Skybox, one of
Australia's first local 'no-contracts' cloud computing services
providers.
The 'Cloud' is everywhere it seems, in the media, on the web and
bandied about by IT companies left right and centre. All the
cloud-vendors seem to agree that cloud is "the future of
computing". But what actually is cloud computing? Is cloud simply
another cunning industry buzzword for selling more servers, or is
cloud computing really a 'paradigm shift' in the way we access
computing resources?
Arguably it's a bit of both.
To understand if cloud computing does represent an evolution -
if not a revolution - in our thinking on the way we process digital
information, as well as ideas of data ownership versus access, we
need to first understand the fundamentals of what cloud computing
is and then examine how this might benefit a business.
In June 2009, IT market analysts Gartner identified the five key
characteristics of cloud computing:
- Service-based - Consumers subscribe to
services designed to directly fulfill specific needs.
- Scalable and elastic - Computing resources and
services can scale up or down depending on the requirements of the
consumer.
- Shared - Services leverage a common platform
or infrastructure to maximise the efficiency of resources.
- Metered-by-use - Consumers are presented with
various payment models based on the usage of services required
rather than the cost of equipment.
- Uses internet technologies - Services are
delivered using the internet, typically internet protocol (IP) as
the service delivery mechanism.
In essence, a range of IT services and resources such as
software, desktop environments, networks and communications systems
are delivered directly to the consumer over the internet, from a
central data centre. While this approach is sometimes called
'infrastructure-as-a-service', it does not mean an end to keyboards
or monitors. Rather the software, servers, networking equipment,
storage and backup space can all be 'virtualised' - you simply
create your desired computing environment or infrastructure 'in the
cloud'.
By separating the physical hardware from the services you can
accurately measure the consumption of those services and therefore
be charged accordingly. As a consumer you only need to pay for the
resources you need at any given time - you don't need to buy all
the physical computing hardware to support your computing
needs.
Consumers of electricity don't rush out and buy a power station.
They purchase and consume the amount of power they need to go about
their daily lives; they rarely worry about how that electricity
gets to them. Why should computing be any different? Why should you
need to invest in computing infrastructure when you can simply
purchase what you need to get the job done and stick it on a
utility bill. Does all this sound familiar? The idea of computing
as a utility is not new. John McCarthy, then a researcher in
artificial intelligence at MIT, stated in 1961 to the MIT
centennial conference that: "If computers of the kind I have
advocated become the computers of the future, then computing may
someday be organised as a public utility just as the telephone
system is a public utility. … The computer utility could become the
basis of a new and important industry."
Nearly 50 years on, we have seen the growth of mainframes,
networked terminals, networked computers, the internet, ASPs, MSPs,
SaaS and so on, each a stepping stone on the path to achieving true
utility computing. One could argue that the emergence of cloud
computing is the first large-scale implementation of public utility
computing - it neatly fits together the many different complex
parts of the puzzle and presents them to the consumer in
easy-to-swallow metered-by-use chunks.
Cloud computing itself is often confused with SaaS
(software-as-a-service), which is hardly surprising as SaaS
solutions do display the characteristics outlined above with one
exception. SaaS delivers specific applications to the end-user's
desktop via a web browser, the application functionality is
rendered as a web page and then controlled through that webpage.
Google Docs and Salesforce.com are examples of this approach. Cloud
is somewhat different, while still using the internet as the
service delivery layer, it can deliver the entire environment to
the end-user, that is operating system, applications,
communications (such as VoIP) and networking functionality. This
means that existing and well-established software applications such
as Microsoft Office can be delivered as a service, even the
operating system itself can become a service. To really understand
the difference simply fire up Google Docs and attempt to install
your own application on there - you can't. With a cloud hosted
desktop, you can.
What about MSPs (managed service providers) and ASPs
(application service providers)? Aren't they the same thing as
cloud? They are very similar in essence but again not the same.
Typically, an MSP will lease dedicated server systems to a company
which will either be on-premise or in a server facility somewhere.
They then provide a management layer and support services. This
means that the consumer is effectively locked to that supplier and
their hardware and scalability means pulling the box out of the
rack and inserting more RAM. An ASP is slightly different again.
Many ASPs have evolved into SaaS and MSPs, but pure-play ASPs
deliver specific applications to the end user usually over a
terminal session - once again application-based rather than
service-based.
Cloud computing, in principle, is not dissimilar to the
mainframes of the past - huge computers accessed by dedicated
networked terminals. Nowadays the huge computer lives in a data
centre, 'the cloud' - the terminal is your PC, thin client or
mobile device and you access your services over the internet. A
simple way of understanding cloud is this: All those servers and
cables and boxes and IT junk cluttering up your office space, and
sucking up your management time making it all work, are gradually
replaced by a big computer somewhere else that you never need to
see or worry about. And in the same way that electricity is
agnostic - you can use electricity to power a refrigerator,
television, toaster or a lava lamp - raw computing power in the
form of processor time, RAM and storage is agnostic too. It can be
your operating system, telephone exchange, backup system or
anything else you might need. Remember, the services consumed are
distinctly separate from the device powering those services.
Benefits of cloud computing
An example of how cloud can directly benefit a small business is
where that business has a several locations and between five and 15
computer users at each location. Chances are the business is too
small and the locations too dispersed to employ a dedicated IT
manager, which means that the business is supporting two or three
independent networks, backup file servers and the like using IT
support contractors. The challenge is to maintain consistency and
security across each individual business unit while still
guaranteeing access to the services needed to run the business. In
this scenario cloud looks attractive - cloud services are not
location dependent, this small business could fulfil its entire
computing requirements directly from the cloud and the business
could function as a single unit from the IT perspective, rather
than the existing dispersed information silos. At the grass roots
level, cloud only needs terminals - existing PCs are fine for this
- and internet connectivity.
The main business benefit, in addition to cost savings, is that
cloud-based IT services move the IT budget from a capital expense
to an operational expense.
This means:
- The cost of ownership can be managed as part of monthly
accounts
- There is no upfront capex
- Much lower management and maintenance costs
- Service continuity
The move to cloud is non-disruptive. From a consumer's
perspective, nothing actually changes: they sit down at their desk,
fire up their computer and there are all the applications and
services exactly as before, the only difference is that their
desktop and applications are now sitting on the big computer
somewhere else, not under their desk.
As hardware becomes redundant in the business, computers can be
replaced by thin clients. These are small units that you plug a
keyboard and mouse into, then connect to the internet. No moving
parts or in-built functionality means that the shelf life of these
boxes is considerably longer than a regular PC and they consume a
fraction of the power.
When the next whizz-bang operating system or fancy
'solve-all-your-problems-at-the-push-of-a-button' application comes
along, businesses no longer need to worry about hardware or
software upgrades - they simply update their cloud services.
Drawbacks of cloud computing
Naturally, managed services providers, SaaS players and IT
support companies are not entirely happy with cloud computing. Nor
probably are IT departments, whose main activity - much to their
chagrin - is the maintenance and upgrading/updating of their
in-house computer resources. (Some authorities put maintenance up
to 80 per cent of IT staff's activities.) Cloud might free up most
of their time for more 'productive' pursuits, but it also might
involve many of them suddenly found without a job.
SaaS players are nervous because shifting end-user operating
systems from local to remote hardware mitigates many of the
benefits promised by these web-based applications. Managed service
providers probably have the most to fear as their entire business
model is built around leasing hardware and providing services that
they can then manage. However, it is likely that MSPs will embrace
cloud when the consumers demand elasticity and scalable computing
and no longer want to be tied to lengthy leasing agreements - they
are also in a good position to provide these services as they have
expertise and existing infrastructure.
At the consumer level, the main concerns are service
availability, bandwidth and security.
Service availability - what happens when the internet goes down?
Are the users cut off from their 'virtual desktops'? Does all work
stop?
The same principles that work at the data centre end can be
applied at the consumer end - failover. At a moderately low cost, a
second broadband connection can be installed at the consumer's
premises. With the right router, load balancing can be achieved
over two connections; if one drops out, the other takes up the
slack. Admittedly a consumer may see a slight drop in the quality
of service until the primary line is restored, but not the complete
failure of services envisaged by sceptics. Additionally, because
the services are on a remote computer if there is a power cut, an
internet drop-out or a wombat chews through the cable, when you do
regain access your services come back exactly as you left them with
no loss of data.
Bandwidth - as the internet connection is used extensively,
there are also some concerns that bandwidth and data transfer may
become an issue. In fact, this is quite the opposite. In many cases
the data transfer rates actually drop. Because you are accessing
data at the data centre, the only information that is passing
through the net are highly compressed screenshots of your remote
computer - the data itself never actually moves from the cloud.
But does sharing computing hardware with other businesses pose a
threat to data?
Security is the perennial IT challenge and a multi-billion
dollar industry. With hardware shared between businesses there is a
perception that you may be able to gain access to other companies'
data or that a virus could spread far more rapidly affecting
thousands of virtual machines in the blink of an eye. In fact, the
nature of virtualisation is such that each machine exists in its
own 'virtual container'; if one crashes it simply crashes the
container, not the whole system. Furthermore, applying group
application policies and security updates over a range of devices
becomes far simpler when they are centrally managed in a cloud.
Instead of relying on an individual user or support technician to
update security policies, this can all be automated from a
centralised control panel. Push a button and all the desktops are
magically updated. While enterprise systems have had this ability
for some time, it is only recently that these benefits have become
available to smaller businesses without the capital or expertise
needed to implement autonomous IT infrastructures.
Xhead: Cloud computing today
The cornerstone of cloud computing was laid as early as 2001
when Microsoft launched its service provider licensing agreement
(SPLA). The agreement allowed services providers to commercially
host Microsoft products as well as re-sell them on a pay-as-you-go
basis. This meant that vendors could effectively 'rent' software
without having to cover the licensing costs upfront. The hosting
industry saw immediate benefits, leading to a drop in the price of
hosting and broadening of services on offer.
Meanwhile, a number of software development companies were
developing and refining their virtualisation platforms - software
that would allow creating virtual containers to host specific
applications and operating systems independently on the same
computer. Unfortunately, many of the cost-benefits of approaching
computing in this manner were eroded by the high cost of the
virtualisation software itself. Enter Microsoft again with Windows
Server 2008. One of the new enterprise features was extremely
low-cost virtualisation, effectively levelling the playing field.
The Server release, allied with improved broadband connectivity,
cost-effective virtualisation and the ability to license products
on a monthly basis, meant that cost rather than management becomes
a strong motivating factor. This means that small and medium
businesses could now afford to get their hands on a piece of the
cloud computing pie.
Even with over 90 per cent of Australian businesses having
broadband access, few of these businesses use the capacity for
running line-of-business services. Typically, they use it for basic
operations like email and web-browsing. Line-of-business
applications are either running locally on the user's machines or
on an application server within the corporate network.
Cloud offers them the direct benefits of optimising their use of
the net and simultaneously minimising in-house IT resources.
However, current players in Australia only have different pieces
of the puzzle - until recently nobody has supplied the full range
of services. The massive costs involved in setting up
Australian-based data centre infrastructure capable of delivering
cloud-based services effectively locks smaller service providers
out of the market, so it is hardly surprising that early cloud
provisioning has been limited to the carrier end of the industry
where cloud services are bolted on to existing service offerings.
While Australian businesses can access IP-based services from
abroad, the latency of such distant services (or the amount of time
it takes for a piece of data to get from A to B and back again)
becomes an issue, so adoption en-masse is yet to occur. As cloud
services do become available locally, we expect the business uptake
to be a rapid one.
Xhead: The future of cloud computing
Given that the big players are very active in the delivery and
creation of a new cloud-based service delivery landscape would
suggest that the future is very bright for cloud. The increasing
capacity of the local broadband network is also a strong indicator
that the carriers see a big future in IP and a big future in IP
means a big future for cloud computing. Gartner itself predicts
that by 2020 over 40 per cent of the world's IT services will be
delivered as cloud-based services - that's a whopping big chunk of
the IT market.
Is there life after the cloud? Will there come a day where
everybody buys their computing resources like they currently buy
electricity? While the benefits to business consumers are obvious,
there are less immediate or compelling benefits to the regular home
user - this may explain why the obvious candidate for piping cloud
services into the home, cable, is still largely used for providing
the internet connectivity itself rather than directly delivering
computing resources.
While the business community will get the early adopter
advantage of embracing cloud, over time we will see greater
competition which will lead to vendors seeking new ways to package
and sell cloud. At time of writing cloud services are available,
however many vendors still tie the consumer in to contracts.
Sometimes, if you look under the hood, some systems currently sold
as cloud are really just dedicated servers, which rather negates
the benefit of having a 'metered-by-use' service. However, it is
understandable as the definition of cloud computing is still
somewhat fluffy. The next generation of true cloud services will be
far more granular in their measuring and billing, much as mobile
phone providers are today - pretty soon we will be talking about
'processing minutes' and 'dedicated RAM time', one line on your
statement might read "User 3 - Microsoft Word Access - 46minutes -
$4.60".
Maybe we are finally catching a glimpse of that shifting
paradigm we've been promised for so long.
Simon Probert is the managing director of Skybox, one of
Australia's first local 'no-contracts' cloud computing services
providers.