Private cloud...what am I missing?

17 views
Skip to first unread message

Doug Tidwell

unread,
Jun 25, 2009, 1:28:01 PM6/25/09
to Cloud Computing Use Cases
Whenever I disagree with people whose opinions I respect, my
assumption is that I've missed something. (Learning Mode is always
more productive than Arguing Mode.) But as much as I've thought about
it, I can't understand why the private cloud isn't a valid use case.
The only difference between a private cloud and a public one is the
fact that none of my data or processing goes outside my firewall, but
that's a *huge* difference. (The DoD is particularly finicky on this
point.)

So with all due respect (which is a lot) to Sam Johnston and
http://twitter.com/cote and others, what am I missing? Does anybody
else think the private cloud is a useful distinction for this forum?
It's clearly an important offering for vendors, but that doesn't
necessarily mean it merits a lot of discussion here. I think it's
something we should cover, but I'm happy to be convinced otherwise.

Your thoughts, everyone?

Cheers,
-Doug

p.s. cote is a very entertaining and insightful analyst, but IBM
forbids me from engaging analysts directly. If somebody who's not
under that restriction would like to invite him, I think his thoughts
would be a great addition to the group.

Reuven Cohen

unread,
Jun 25, 2009, 1:41:57 PM6/25/09
to cloud-comput...@googlegroups.com
I think the problem with the private cloud as a use case for cloud computing , it's like saying the use case for the internet is the apache webserver. It's what you do with or on apache that matters.

I Bcc'd cote

reuven

Krishna Sankar (ksankar)

unread,
Jun 25, 2009, 1:49:18 PM6/25/09
to cloud-comput...@googlegroups.com
Doug,
You are right - the main difference between private and public
clouds are of scale and the spatial posture. Public clouds would have a
lot more customers but (relatively) less number of distinct application
archetypes while the private cloud will have less number of customers
but a lot more apps and many of them with very different multi-tiered
architectures with their own compute-storage-network semantics.

But there are two big differences that many miss including the
NIST definitions i.e. the need for Internet and the pay-as-you-go. So
long as we abstract these two (i.e. network and meter - with option to
pay-as-you-go or charge usage or other charge back mechanisms) to
include both private and public clouds, we could consider them in
aggregation.

Good point.
Cheers
<k/>

xetta

unread,
Jun 26, 2009, 12:26:14 PM6/26/09
to Cloud Computing Use Cases

Krishna, and all,

I agree completely with your points about scale and vareity of
applications; I had actually not thought about that as part of the
definition of a private cloud and you're right. The smaller scale and
lower diversity in applications may make the private cloud a different
use case, since you may need less complex tools to configure and
manage it. We're in a much smaller and homogeneous setting. On the
other hand, the use of internet and the pay-as-you-go work pretty much
the same in both public and private clouds, at least from a use case
point of view.

Private firms do need the Internet to make use of their clouds. I
wouldn't find much sense in building a whole private cloud
infrastructure and limiting it's use to the people in the same
facility. The fact that it is private, that is, that it is used only
by one entity, does not mean it is used only at one location. Private
clouds also have this centralization trend we see in the public clouds
and leverage on the Internet to achieve it.

Private clouds can and should work on a pay-as-you-go basis. What
changes within firms is the way you spread out the cost of the IT
infrastructure. Before the cloud, the depreciation of the IT assets
would be spread out according to metrics such as number of PCs or
number of employees in each business group. Now the costs of the IT
infrastructure can be charged to each business application group
according to how much computing they use. Going back to the NIST
definition, private firms can better manage their costs with the new
"measured service".

Best,

German

Lisa Noon, IBM

unread,
Jun 26, 2009, 3:01:15 PM6/26/09
to Cloud Computing Use Cases
Hello Doug,

As a matter of fact I was having this discussion the other day with an
architect at a major financial services company. He made the point
very clearly that for reasons of pure governance they will create
"bounded" clouds, meaning purposefully constructed and auditably (is
that a word :)) described boundaries. From the outside they conduct
penetration tests, and on the inside perform all manner of similar
tests, including virus scans, etc. So the point I put forth that I
think this gentleman would support is that private clouds have a
boundary attribute that is not characteristic of public (non-private)
clouds.

On Jun 25, 1:28 pm, Doug Tidwell <dtidw...@us.ibm.com> wrote:
> Whenever I disagree with people whose opinions I respect, my
> assumption is that I've missed something. (Learning Mode is always
> more productive than Arguing Mode.) But as much as I've thought about
> it, I can't understand why the private cloud isn't a valid use case.
> The only difference between a private cloud and a public one is the
> fact that none of my data or processing goes outside my firewall, but
> that's a *huge* difference. (The DoD is particularly finicky on this
> point.)
>
> So with all due respect (which is a lot) to Sam Johnston andhttp://twitter.com/coteand others, what am I missing? Does anybody

Rodrigo

unread,
Jul 1, 2009, 12:59:46 PM7/1/09
to Cloud Computing Use Cases
I tend to think of cloud computing as a business model underpinned by
virtualization. Just as the virtualization layer abstracts the
underlying hardware, there's a front-layer that abstracts the cost,
sla, workload definition from the underlying computing
infrastructure.

We,and many startups, are using EC2 because the low, standard,
predictable and variable cost with elastic provisioning of resources
in minutes. And this is all accessed on a self-service basis with full
control of the life cycle. This change is hugely revolutionary vs
what we faced before with dedicated hosting.

Now when I hear customers talking about building a private cloud,
generally they mean the ability to deliver all or a subset of the
"standard, predictable variable, elastic, self-service, etc" value.
So where's the difference between private cloud and public cloud? It
is wholly dependent on your role. If you are building the
technology,the technologies and issues are very similar. If you are a
consumer, you'd like to treat them both the same,

But if you are the owner or operator, the private cloud is very
challenging. You need to think of a couple of things, One, what are
the standard environments and workload variations I will allow? How do
I condense everything we do into a predictable unit cost? And, how do
I provide elasticity? There are many other challenges as well, but
those three are biggies for your typical IT geek.

-Rodrigo Flores
http://www.cloudfrontoffice.com (blog)
http://www.servicecatalogs.com (blog)




On Jun 26, 12:01 pm, "Lisa Noon, IBM" <ln...@us.ibm.com> wrote:
> Hello Doug,
>
> As a matter of fact I was having this discussion the other day with an
> architect at a major financial services company.  He made the point
> very clearly that for reasons of pure governance they will create
> "bounded" clouds, meaning purposefully constructed and auditably (is
> that a word :)) described boundaries.  From the outside they conduct
> penetration tests, and on the inside perform all manner of similar
> tests, including virus scans, etc.  So the point I put forth that I
> think this gentleman would support is that private clouds have a
> boundary attribute that is not characteristic of public (non-private)
> clouds.
>
> On Jun 25, 1:28 pm, Doug Tidwell <dtidw...@us.ibm.com> wrote:
>
> > Whenever I disagree with people whose opinions I respect, my
> > assumption is that I've missed something. (Learning Mode is always
> > more productive than Arguing Mode.) But as much as I've thought about
> > it, I can't understand why the private cloud isn't a valid use case.
> > The only difference between a private cloud and a public one is the
> > fact that none of my data or processing goes outside my firewall, but
> > that's a *huge* difference. (The DoD is particularly finicky on this
> > point.)
>
> > So with all due respect (which is a lot) to Sam Johnston andhttp://twitter.com/coteandothers, what am I missing? Does anybody

Fred Zappert

unread,
Jul 1, 2009, 2:24:39 PM7/1/09
to cloud-comput...@googlegroups.com
Hi,

One of my friends at a major bank is pushing hard to migrate their systems into an internal cloud.

Any enterprise with a large number of systems will achieve many of the benefits of improved resource utilization through virtualization  With that accomplished, and with applications segregated and sanitized to respect rules, regulations and legislation, it is easier to push those applications into a public cloud on an as needed basis.

Yes, a private cloud will be far more constrained on the upper limits of elasticity, but with average utilizations running at about 12%, the ability to use those cycles more effectively should be compelling. 

But often it isn't, because each silo is provisioned for some peak hour in peak day with a 2x factor of cross data-center redundancy thrown in.

So, the transition may only happen from the top-down if and when the capital budget gets tight.

Regards,

Fred.

Ravi

unread,
Jul 9, 2009, 9:04:13 AM7/9/09
to Cloud Computing Use Cases
In a private cloud, the savings in hardware costs are not going to be
massive. Over time we may see some decrease as we mature our processes
and technology, but today the bulk of the cost of IT operation comes
in the form of software licensing and manpower.

I don't see much discussion on pay-per-use for software such as Oracle
database. Amazon/Oracle images already offer a model and it looks
interesting. Is this model likely to expand to private clouds too and
if not why not? This is where private clouds are bound to reap
benefits. As the utilization number (12% bandied about) shows that
most of the computing resources are used only at certain times, the
software licensing costs should take that into account. In a large
enterprise setting, that could result in significant cost savings and
encourage more adoption of private clouds.

Regards,

Ravi Kulkarni
http://computingnebula.wordpress.com
http://kulkarnr.blogspot.com

wjhuie

unread,
Jul 10, 2009, 9:03:21 AM7/10/09
to Cloud Computing Use Cases
German,
Glad to see you making the statement that private clouds should work
on pay-as-you-go.

It sounds like you think private clouds have a more homogeneous
workload than public (which I think seems true based on scale).
Although as an enterprise I could conceive of a "wild wild west" on my
private cloud where I let people develop, experiment, etc. And then
I'd only move a really small subset of that work to a public cloud.

Sam Johnston

unread,
Jul 10, 2009, 9:34:25 AM7/10/09
to cloud-comput...@googlegroups.com, Cloud Computing Use Cases
I've had enterprise cusomers suggest the opposite too - doing
development outside for internal deployment.

That's not to say I think it's a good idea but worth bearing in mind.

Sent from my iPhone

wjhuie

unread,
Jul 10, 2009, 9:46:06 AM7/10/09
to Cloud Computing Use Cases
That's a good point Sam (and I'd love to know how you can follow these
threads on your iPhone!!!)

So it may be even more critical to avoid making assumptions about the
homo/heterogeneity of workloads / users on any cloud!

Security seems like an "easy" identifier but in the Gov sector it's
just as important to keep info away from your colleagues (if it's
private / need to know) as it is in a public setting, so I'm not sure
that's a good criteria.

Funding seems possible but getting the "perspective" correct might be
hard. I.e. a Gov private cloud, vs a DoD private cloud, vs. Private
clouds that various DoD departments may have.

So I'm not really sure how to best differentiate "public vs private".
The "risk" statement (i.e. who's responsible and impacted by failures)
seems tantalizing to me but perhaps just as tricky as the rest.

Anils

unread,
Jul 10, 2009, 4:06:19 PM7/10/09
to cloud-comput...@googlegroups.com
Yes Ravi...
 
The model that I had in mind which I wrote about earlier where there would be the broker for the User for an entity, be it Application, Platform or Infrastructure and then application vendors who would possibly be resellers (I am not sure, Pay per use could be implemented by them) and then the vendors having volume label subscription contracts with people like Oracle and so on.
 
Here again, the concept of peak & off peak sharing of resources could be thought about as well if we are really looking at having the spread across the 24hr period. Please do share your thoughts on this.
 
Cheers,
 
Anils

--
Anil Kunjunny

15a Ian Marwick Place, Birkenhead
t (09) 480-8045 | m 021 1623 120| e anil.k...@gmail.com

"You have to be the change that you want to see in the world"
                                                          - Mahatma Gandhi

Khürt Williams

unread,
Jul 11, 2009, 8:35:00 AM7/11/09
to Cloud Computing Use Cases
I am not sure I even under what private cloud means given that the
Cloud term was meant to refer to how the public Internet was
represented on network diagrams. If it is inside my firewall then how
is it "Cloud"?

On Jun 25, 1:28 pm, Doug Tidwell <dtidw...@us.ibm.com> wrote:
> Whenever I disagree with people whose opinions I respect, my
> assumption is that I've missed something. (Learning Mode is always
> more productive than Arguing Mode.) But as much as I've thought about
> it, I can't understand why the private cloud isn't a valid use case.
> The only difference between a private cloud and a public one is the
> fact that none of my data or processing goes outside my firewall, but
> that's a *huge* difference. (The DoD is particularly finicky on this
> point.)
>
> So with all due respect (which is a lot) to Sam Johnston andhttp://twitter.com/coteand others, what am I missing? Does anybody

Botchagalupe

unread,
Jul 11, 2009, 9:16:11 AM7/11/09
to Cloud Computing Use Cases
One of the most exciting things about public clouds is their ephemeral
nature. What makes something like AWS so exciting is that you can
build and application or a business with an "almost" instant
architecture. The infrastructure of a business or application can be
moved, modified, or destroyed automatically or if a pesky human is
needed - with a button. Most behind-the-firewall infrastructures have
been static in nature. Even a virtualized BTF environment is for the
most part a static representation of it's prior bare-metal brethren. I
think the key differentiator of a true private cloud is one that
provides the capability to build and destroy infrastructure the same
way as can be done with something like AWS. Today offerings like
Eucalyptus, 3Tera, and a few others have that "I know one when I see
one" DNA.

I have done some strategic planning with some enterprise customers on
the subject of private clouds and you can easily see the spark in
their eyes when they realize the kind of control and out of the box
thinking they can get with the ephemeral infrastructure possibilities.

John Willis (a.k.a botchagalupe)
johnmwillis.com

Joe Armstrong

unread,
Jul 11, 2009, 11:48:55 AM7/11/09
to cloud-comput...@googlegroups.com

I think the meaning of cloud has changed somewhat since it was first used to indicate "the internet" - which I agree, it was used in the past in that context.

Now the cloud means the promise of network based connectivity to infinite resources (VM's, disk).  If that set of infinite resources (or what would appear to be infinite) is contained within one enterprise's intranet then it may be called a "private cloud".

But whether the cloud is public or private or hybrid or brokered or federated or <insert your favorite single word description here> the consumers interface to that cloud should (must ?) be identical.

So just to finish the thought, since the consumer's interface to the various cloud topologies is identical the distinction of one topology from the other is just that - a topological distinction not a functional one.  Again, this is from the consumers standpoint.

From the administrators standpoint there is quite a big difference, mostly in the area of data locality... where is the data stored and where does the data go to be processed (on-premises or off-premises).

Cheers,

Joe

Botchagalupe

unread,
Jul 11, 2009, 12:24:31 PM7/11/09
to Cloud Computing Use Cases
Joe,

Hopefully the consumer's interface and the cloud topology will be like
the way we use a telephone today (we don't care what the phone
infrastructure looks like - we just want to make a phone call).
That's all on the app side. I think we first need to solve the "how
to do data centers" the correct way (i.e., the infrastructure). This
is where the private cloud can .. in the end enhance the user's
expectation. Today a good customer interface should not depend on
whether it is meatcloud/cloud/public/private/hybrid... as long as they
get the resources in line with their expectations. I know we all
believe that the cloud is the answer; however, they don't and
shouldn't need to know that.

John
johnmwillis.com

On Jul 11, 11:48 am, Joe Armstrong <josephwarmstr...@gmail.com> wrote:
> I think the meaning of cloud has changed somewhat since it was first used to
> indicate "the internet" - which I agree, it was used in the past in that
> context.
>
> Now the cloud means the promise of network based connectivity to infinite
> resources (VM's, disk).  If that set of infinite resources (or what would
> appear to be infinite) is contained within one enterprise's intranet then it
> may be called a "private cloud".
>
> But whether the cloud is public or private or hybrid or brokered or
> federated or <insert your favorite single word description here> the
> consumers interface to that cloud should (must ?) be identical.
>
> So just to finish the thought, since the consumer's interface to the various
> cloud topologies is identical the distinction of one topology from the other
> is just that - a topological distinction not a functional one.  Again, this
> is from the consumers standpoint.
>
> From the administrators standpoint there is quite a big difference, mostly
> in the area of data locality... where is the data stored and where does the
> data go to be processed (on-premises or off-premises).
>
> Cheers,
>
> Joe
>

Joe Armstrong

unread,
Jul 11, 2009, 1:10:52 PM7/11/09
to cloud-comput...@googlegroups.com

I couldn't agree more...

Joe

stingleypt

unread,
Jul 11, 2009, 10:28:04 PM7/11/09
to Cloud Computing Use Cases
Perhaps to your point and Doug's, there are a couple of use cases.
Unfortunately the terms public and private cloud are currently in use
in a variety of conflicting ways. One is if my organization builds an
internal cloud, some would refer to that as a private cloud because
it's in my organization and I can keep it private. Another use of the
term private cloud is when the gov sector procures cloud services from
the private sector. Analogously, is a public cloud one that is
available to the general public? The government's use and perspective
on the cloud kind of flip it inside out. The government is by
definition not the private sector, so if it establishes a cloud, it
can't be a private cloud, it have to be a public cloud (besides, the
government is in the business of public service, that would make its
clouds public clouds.)
At this point these terms may be so overloaded as to be essentially
meaningless.

What we need is a use case to express when my organization extends
itself to a cloud service provider to gain some service (such as
additional computing or storage). The cloud service provider,
operating as a shared service provider, encrypts my data and ensures
that it is not compromised (keepts it private). There is a second use
case where my organization sends data to a cloud provider, but this is
not an extension of my domain to another entity. Both entities
operate as island domains, but pass data (probably XML) in a
controlled manner between them. And there is a third use case where
my organization builds internal data centers using cloud architectures
and methodologies (employing on-demand virtualization).

I'm not sure how to apply the terms private or public to these three
use cases. Perhaps we should find different terms for the use cases
and put public and private in the same heap of overloaded and thus
meaningless terms as hybrid.

Pat Stingley
> > >> German- Hide quoted text -
>
> - Show quoted text -

Sam Johnston

unread,
Jul 12, 2009, 8:13:22 AM7/12/09
to cloud-comput...@googlegroups.com
On Sat, Jul 11, 2009 at 2:35 PM, Khürt Williams <khurtw...@gmail.com> wrote:

I am not sure I even under what private cloud means given that the
Cloud term was meant to refer to how the public Internet was
represented on network diagrams.  If it is inside my firewall then how
is it "Cloud"?

Amen. The evolution of virtualisation is NOT cloud computing.

A few decades ago network diagrams necessarily contained every node and link because that was the only way to form a connected graph. Then telcos took over the middle part of it and consumers used a cloud symbol to denote anything they didn't [need to] care about... they just stuff a packet in one part of the cloud and it would magically appear [most of the time] out of another. Another way of looking at it (in light of the considerable complexity and cost) is "Here be dragons" - same applies today as managing infrastructure is both complex and costly.

Cloud computing is just that same cloud getting bigger, ultimately swallowing the servers and leaving only [part of] the clients on the outside (although with VDI nothing is sacred). Consumers now have the ability to consume computing resources on a utility basis, as they do electricity (you just pay for the connection and then use what you want). Clearly this is going to happen, and probably quicker than you might expect - I admit to being surprised when one of my first cloud consulting clients, Valeo, chose Google Apps for 30,000 users over legacy solutions back in 2007. Early adopters, as usual, will need to manage risk but will be rewarded with significant cost and agility advantages, as well as immunised to an extent against "digital native" competitors.

You can be sure that when Thomas Edison rocked up 125 years or so ago with his electricity grid there were discussions very similar to those that are going on today. With generators ("Electricity as a Product") you have to buy them, install them, fuel them, maintain them and ultimately replace them, which sustained a booming industry at the time. We all know how those conversations ended... Eastman Kodak is the only company I know of today still running their own coal fired power station (though we still use generators for remote sites and backup - this will likely also be the case with cloud). Everyone else consumes "Electricity as a Service", paying a relatively insignificant connection fee and then consuming what they need by the kilowatt hour.

What we have today is effectively "Infrastructure as a Product" and what we'll have tomorrow is "Infrastructure as a Service" (though I prefer the term "Intrastructure Services" and expect it to be "infrastructure" again once we've been successful and there is no longer any point in differentiating).

Now if legacy vendors work out how to deliver products as services (for example, by using financing to translate capex into opex and providing a full maintenance and support service) then they may have some claim to the "cloud" moniker, but that's not what I'm seeing today. Most of the "private cloud" offerings are about hardware, software and services (as was the case in the mainframe era) rather than true utility (per hour) basis. Good luck competing with the likes of Google and Amazon while carrying the on-site handicap - I'm expecting the TCO of "private cloud" setups to average an order of magnitude or so more than their "public" counterparts (that is, $1/hr ala network.com rather than $0.10/hr ala Amazon EC2), irrespective of what McKinsey et al have to say on the subject.

In the context of the use cases, sure on-premises or "internal" cloud rates a mention but the "public/private" nomenclature is problematic for more reasons than I care to list. I personally call it "I can't believe it's not cloud", but that's not to say I leave it out of proof of concepts and pilots... I'm just careful about managing expectations. Ultimately the user and machine interfaces should be the demarcation point for such offerings and everything on the supplier side (including upfront expense) should be of no concern whatsoever to the user. I consider utility billing and the absence of capex to be absolute requirements for cloud computing and feel this ought to be addressed in any such document - suppliers might, for example, offer the complete solution at $1/hr with a minimum of 150 concurrent instance minimum (~= $100k/month).

Oh and if large enterprises want to try their hands at competing with the likes of Google and Amazon by building their own next generation datacenters then that's fine by me, though I equate it to wanting to build your own coal-fired power station when you should be focusing on making widgets (and it should in any case be done in an isolated company/business unit). I imagine it won't be long before shareholders will be able to string up directors for running their own infrastructure, as would be the case if they lost money over an extended outage at their own coal-fired power station when the grid was available.

Sam
Reply all
Reply to author
Forward
0 new messages