My Best of MMS 2012 Series: Private Cloud 2012 Lessons Learned from Our Early Adopters

An open discussion on people who have built private clouds at customers.


In real live things don’t shrink that often.  Smile Free or real cheap back charge rates are not doing anything to help.

My take on this is that you should look at elasticity as a flexibility feature. Even if a cloud is no that elastic in both ways. You can shrink a cloud v1 to zero as you migrate the VMs to private cloud v2. Than dump the resources back into another pool, step by step or in one go. I’ll use whatever works to make my live easier.

Standardizing & Customer Centric Operations = Planning is Key!

These two can be hard to combine. It takes serious planning and as such an upfront investment.

Than you need to build it to optimize operations (cost & excellent service). This sounds nice but how good are we at this and what is the shelf life of a solution versus the investment?

There is a lot of preparation to do. There is a lot of things to consider. Databases, Storage, the network, security boundaries, disaster recovery planning. They advise not to do it cross domain. Hmmm … we need to address this. Seriously..

Testing => build decent scripts with variables & config files. This will help to deploy in test, acceptance & production without to many changes/work.

Make sure you define all service accounts, groups and permissions you need.

It’s all about planning and what’s being told are best practices that exist already, private cloud or not.

Self Service

  • Service Catalog is a prerequisite.
  • Self Service is key to the private cloud.
  • If people can they will do things differently. You’ll have to learn to deal with this.
  • Billing for services should be clear. Not to much detail. VMs & Storage are two good ones. Keep it simple and don’t go into memory & vCPU. Just set boundaries.


We dive into the System Center products and look at it from both the IT Pro and the consumer side of things.

Requests => Approval & Deployment

Approval process should be dynamic based on what is requested & who’s making is. You’ll also need SLAs & chargeback on these. Be careful not over complicating it or you encourage rogue IT.

RANT: IT should make things as easy as possible. And in this discussion I’m not won over for charge back. It often turns into an excel exercise. Internal IT becomes more and more like an external service provider or integrator in this model. The inherent strength of being part of the business and being in the best position to help that business move ahead is lost. Is this a complot of the integrators? It fits their model but basically a lot of that is broken very badly. The last thing internal IT should do is become like them. That will do nothing for “Business-IT alignment”. We need to leverage the possibilities of the private cloud for our business or we have no unique selling point. Not that the service providers do a better job, but at least they are not on the pay roll so the bean counters like that. And as long as they can use public cloud to get their needs served hey couldn’t care less about who does the private cloud thingy for them. So a functional IT is first and foremost what we need. That is customer centered. Alignment of business & IT is worthless without that. The latter happens ay to much.


Well yes this is important. We need reports, reviews, Service Improvement Plans, look for opportunities for automation.

Personal Best of MMS 2012 Series “Why We Fail–An Architect’s Journey to the Private Cloud”


The speaker (Alex Jauch) addresses cloud terminology confusion and points out that yet everyone wants it. So the pressure is on to deliver cloud.

But as an architect you can’t build with such vague notions of what it is. That just doesn’t work. 78% of enterprise IT Shops will deploy a private cloud by 2014 (Gartner) 62% of all IT Projects fail. For the record, building a private cloud is not an easy project.

For one, what are you building? What is it, way to may definitions. NIST seems to be one of the better definitions around. Specific, direct and actionable. We can work with that. I suggest you visit the NIST site for more information on:

  • Deployment Models:Private Cloud, Hybrid, Public.
  • Service models SAAS, PAAS, IAAS
  • The Essential Characteristics
    The Common Characteristics

Why We Fail

What happens:

  • Install Hyper-V
  • Deploy System Center
  • Build a solution

The essential element of cloud is that  “The cloud is a customer centric business model, not technology”.It’s approached to much as a technology problem and that’s why we fail.

The architect should not allow this to happen so he is to blame. The architectural practice is to marry business needs and wants to technology as a solution. This really hits home but there are more people involved and than there is the entire business / IT alignment fiasco as you can read in my blog The shortage of skilled employees, are we making it worse? , but the bucks ends with the architect..

How do we add value to the business? Commodities do not add value, they are necessities. So we need to decide what business we are in. Meeting standards is not a goal. Enabling business is the goal. So they think you’re doing a great job empowering them. After all they are paying for it.

The Take Away

Traditional IT needs to evolve (fast) to customer centric IT.  End user departments define the goals. Our operational proficiency used to be our pride but what does it mean to the customer? Problems that do not affect the business don’t matter. So talk to customers to find out about what they want and need. They don’t care about your skill set or certifications. You’ll need t extract the need from their wants.

The ability to take pain points away from customers. Small & medium sized projects do very well at this. But in a lot of companies they don’t promote you for those “smaller” projects. So the business also has to evolve.

I’d like to add that Old style IT is also promoted by  a lot of misguided security officers and business lawyers. Strict rules as a guidance and instrument are their instruments and no those are also not always in the business best interest.

This relates to IT Portfolio Management: Strategic, High Potential, Key Operational & Support. We need to realize that whatever we work on might be strategic or high potential will move to key operational and support. They all need different approaches and types of management. So choose your methodologies wisely. Don’t just pick one and force that square peg in the round hole. This is my advice to both business and IT. I’ve seen business decisions change support level products turned into high cost  high maintenance because due to bad decisions. So we might not have to be our brothers keeper towards the business but than again do we really need those bridging functions and those guys or gals need to be at the top of their game as I stated in The shortage of skilled employees, are we making it worse?

So keep things a simple and as effective as possible. Do it fast, ride and repeat. You’ll learn a lot and improve along the way. So here comes the build or buy decision and the link to the NetApp plug by the speaker. This is very dependent on the situation of the organization at hand. So the fast track has it’s place here. Is speed of delivery of key importance or absolute flexibility and adaptability? So it will depend. Yes the consultants answer. But being a real consultant is a very respectable job. I can’t hell it that the word has become meaningless due to missuses and inflationary titles for temps for hire.  The System Center stack and how NetApp improves and leverages all this is briefly discussed. He ties the fast track into the discussion of portfolio management and working in a customer centric way.


Why are we doing what we do? Think about it. There is a nice book on this subject  “Why we fail? by Alex Jauch.

Thinking About Windows 8 Server & Hyper-V 3.0 Network Performance


The main purpose of this post is, as mentioned in the title, to think. This is not a design or a technical reference. When it comes to designing a virtualization solution for your private cloud their are a lot of components to consider. Storage, networking, CPU, memory all come in to play and there is no one size fits all. It all depends on your needs, budget in combination with how good your  insight into your future plans & requirements are. This is not and easy task. Virtualizing 40 webservers is very different from virtualizing SQL Server, Exchange or SharePoint.  Server virtualization is different from VDI and VDI itself comes in many different flavors.

  • So what workloads are you hosting? Is it a homogeneous or a heterogeneous environment?
  • What kind of applications are you supporting? Client-Server, SOA/Web Services, Cloud apps?
  • What storage performance & features do you need to support all that?
  • What kind of network traffic does this require?
  • What does your business demand? Do you know, do they even know? Can you even know in a private cloud environment?
  • Do you have one customers or many (multi tenancy) and how are they alike or different in both IT needs and business requirements.

The needs of true public cloud builders are different from those running their own private clouds in their own data centers or in a mix of those with infrastructure at a hosting provider. On top of that an SMB environment is different from large enterprises and companies of the same size will differ in their requirements enormously due to the nature of their business.

I’ve written about virtualization and CPU considerations before (NUMA, Power Save settings for both OS & 10Gbps network performance) before. I’ve also discussed a number of posts about 10Gbps networking and different approaches on how to introduce it with out breaking the bank. In 2012 I intend to blog some more on networking and storage options with Windows 8 and Hyper-V 3.0. But I still need to get my hands on the betas and release candidates of Windows 8 to do so. You’ll notice I don’t talk about Infiniband. Well I just don’t circulate in the ecosystems where absolute top notch performance is so important that they can justify and get that kind of budget to throw at those needs.

To set the scene for these blog posts I’ll introduce some considerations around networking options with Hyper-V. There are many features and options both in hardware, technologies, protocols, file systems. Even when everything is intended to make live simpler people might get lost in all the options and choices available.

Windows Server 8 NIC features – The Alphabet Soup

  • Data Center Bridging (DCB)
  • Receive Segment Coalescing (RSC)
  • Receive Side Scaling (RSS)
  • Remote Direct Memory Access (RDMA)
  • Single Root I/O Virtualization (SR-IOV)
  • Virtual Machine Queue (VMQ)
  • IPsec offload (IPsecTO)

A lot of this stuff has to do with converged networks. These offer a lot of flexibility and the potential for cost savings along the way. But convergence & cost savings are not a goal. They are means to an end. Perhaps you can have better, cheaper and more effective solutions leveraging you existing network infrastructure by adding some 10Gbps switches & NICs where they provide the best bang for the buck. Chances are you don’t need to throw it all out and do a fork lift replacement. Use what you need from the options and features. Be smart about it. Remember my post on A Fool With A Tool Is Still A Fool, don’t be that guy!

Now let’s focus on couple of the features here that have to do with network I/O performance and not as much convergence or QOS. As an example of this I like to use Live migration of virtual machines with 10Gbps. Right now with one 10Gbps NIC I can use 75% of the bandwidth of a dedicated NIC for live migration. When running 20 or more virtual machines per host with 4Gbps to 8Gbps of memory and with Windows 8 giving me multiple concurrent Live Migrations I can really use that bandwidth. Why would I want to cut it up to 2 or 3Gbps in that case. Remember the goal. All the features and concepts are just tools, means to and end. Think about what you need.

But wait, in Windows 8 we have some new tricks up our sleeve. Let’s team two 10Gbps NICs put all traffic over that team and than divide the bandwidth up and use QOS to assure Live Migration gets 10Gbps when needed but without taking it away from others network I/O when it’s not needed. That’s nice! Sounds rather cool doesn’t it and I certainly see a use for it. It might not be right if you can’t afford to loose that bandwidth when Live Migration kicks in but if you can … more power & cost savings to you. But there are other reasons not to put everything on one NIC or team.


One thing all these have in common is that they are used to reduce the CPU load / bottleneck on the host and allow to optimize the network I/O and bandwidth usage of your expensive 10Gbps NICs. Both avoiding having a CPU bottleneck and optimizing the use of the available bandwidth mean you get more out of your servers. That translates in avoiding buying more of them to get the same workload done.

RSS is targeted at the host network traffic. VMQ and SR-IOV are targeted at the virtual machine network traffic but in the end the both result in the same benefits as stated above. RSS & VMQ integrate well with other advanced windows features. VMQ for examples can be used with the extensible Hyper-V switch while RSS can be combined with QOS & DCB in storage & cluster host networking scenarios. So these give you a lot of options and flexibility. SR-IOV or RDMA is more focused on raw performance and doesn’t integrate so well with the more advanced features for flexibility & scalability. I’ll talk some more on this in future blog posts.

Now with all these features that have there own requirements and compatibilities you might want to reconsider putting all traffic over one pair of teamed NICs. You can’t optimize them all in such a scenario and that might hurt you. Perhaps you’ll be fine, perhaps you won’t.


So what to use where and when depends on how many NICs you’ll use in your servers and for what purpose. For example even in a private cloud for lightweight virtual machines running web services you might want to separate the host management & cluster traffic from the virtual machine network traffic. You see RSS & VMQ are mutually exclusive. That way you can use RSS for the host/cluster traffic and DVMQ for the virtual machine network. Now if you need redundancy you might see that you’ll already use 2*2NIC with Windows 8 NLB in combination with two switches to avoid a single point of failure. Do your really need that bandwidth for the guest servers? Perhaps not but, you might find that it helps improve density because of better better host & NIC performance helping you avoiding the cost of buying extra servers. If you virtualize SQL servers you’d be even more interested in all this. The picture below is just an illustration, just to get you to think, it’s not a design.


I’m sure a lot of matrices will be produced showing what features are compatible under what conditions, perhaps even with some decision charts to help you decide what to use where and when.

The Private Cloud A Profitable Future Proofing Tactic?

The Current Situation

I’m reading up on the private cloud concept as Microsoft envisions we’ll be able to build ‘m with the suite of System Center 2012 products. The definition of private cloud is something that’s very flexible. But whether we’re talking about the private, hybrid or public cloud there is a point of disagreement between some on the fact that there are people that don’t see self-service (via a portal, with or without an approval process) as a required element to define a *cloud. I have to agree with Aidan Finn on this one. It’s a requirement. I know you could stretch the concept and that you could build  a private cloud to help IT serve it customers but the idea is that customers can and will do it themselves.

The more I look into system center 2012 and it’s advertised ability to provide private clouds the more I like it. Whilst the current generation has some really nice features I have found it lacking in many areas, especially when you start to cross security boundaries and still integrate the various members of the System Center suite. So the advancements there are very welcome. But there is a danger lurking in the shadows of it all. Complexity combined with the amount of products needed. In this business things need to go fast without sacrificing or compromising on any other thing. If you can’t do that, there is an issue. The answer to these issues is not always to go to the public cloud a hundred percent.

While the entire concept might seem very clear us techies (i.e. still lots of confusion to be found) and the entire business crowd is talking about cloud as if it’s a magic potion that will cure all IT related issues (i.e. they are beyond confused, they are lost) there are still a lot of questions. Even when you have the business grasping the concept (which is great) and have an IT team that’s all eager and wiling to implement it (which is fabulous) things are still not that clear on how to start building and/or using one.

In reality some businesses haven’t even stepped into the virtual era yet or only partially at best. Some people are a bit stuck in the past and still want to buy servers and applications with THEIR money that THEY own and is ONLY for them.  Don’t fight that to much The economics of virtualization are so good (not just financially but also in both flexibility & capabilities) that you can sell it to top management rather easily, no matter what. After that approval just sell the business units servers (that are virtual machines), deliver whatever SLA they want to pay for and be done with it. So that problem is easily solved.

But that’s not a cloud yet. Now that I’m thinking of it, perhaps getting businesses to adopt the concept will be the  hardest. You might not think so by reading about private clouds in the media but I have encountered a lot of skepticism and downright hostility towards the concept. No, it’s not just by some weary IT Pros who are afraid to lose their jobs. Sometimes the show stoppers are the business and users that won’t have any of it. They don’t want to order apps or servers on line, they want then delivered for them. I even see this with the younger workforce when the corporate culture is not very helpful. What ‘s up here? Responsibility. People are avoiding it and it shows in their behavior. As long as they want to take responsibility things go well. If not, technical fear masked as “complexity” or issues like “that’s not our job” suddenly appear.

There is more, a lot of people seems at their limit of what they can handle in information overload at every extra effort is too much.  Sometimes it’s because of laziness or perhaps even stupidity? Perhaps it’s a side effect of what Nicolas Carr writes about the: the internet is making us dumber and dumber as a species. But then again, we only have to look at history to learn that, perhaps, we’ve never been that smart. Sure we have achieved amazing things but that doesn’t mean we don’t act incredibly stupid as individuals or as a group. So perhaps things haven’t changed that much. It’s a bit like the “Meet the new boss, same as the old boss” sort of thing. But on the other hand things are often too complex. When things are easy and become an aid in their work people adopt technology fast and happily.

Sometimes the scale of the business is not of that nature that it’s worthwhile top deploy a cloud. The effort and cost versus the use and benefits are totally out of sync.

That’s all nice and well you tell me, but what’s are technologists to advice to their customers?

Fire & Maneuver

The answer is in the sub title. You can’t stand still and do nothing. It will get you killed (business is warfare with gloves on and some other niceties). Now that’s all good to know but how do we keep moving forward and scoring? There will always be obstacles, risks, fears etc. but we can’t get immobilized by them or we come to a standstill, which means falling behind. The answer is quite simple. Keep moving forward.  But how? Do what you need to do. Here’s my approach. Build a private cloud. Use it to optimize IT and to be ready to make use of * clouds at every opportunity. And to put your mind at ease you can do this without spending vast amounts of money that gets wasted. Just provide some scale up and scale out capacity & capability. The capability is free if you do it right. The capacity will cost you some money. But that’s your buffer to keep things moving smoothly. Done right your CAPEX will be less than not doing this. How can this be?

Private Clouds enable Hybrid Clouds

The thing that I like most about the private cloud is that it enables the use of hybrid cloud computing. On the whole and in the long run hybrid clouds might be a transition to public cloud but as I’ve written before, there are scenarios where the hybrid approach will remain. This might not be the case for the majority of businesses but still I foresee a more permanent role for hybrid clouds for a longer time that most trendy publications seem to indicate. I have no crystal ball but if hybrid cloud computing does remain a long term approach to server computing needs we night well see more and better tools to manage this in the years to come. Cloud vendors who enable and facilitate this approach will have a competitive advantage. The only thing you need to keep I mind that private or cloud computing should not bee seen as replacements or alternatives for the public cloud. They don’t have the elasticity, scale and economics of a public cloud. They are however complementary. As such they enable and facilitate the management and consumption of IT services that have to remain on premises for whatever reason.

Selling The Public Cloud

Where private cloud might help businesses who are cloud shy warm up to the concept, I think the hybrid cloud in combination with integrated and easy management will help them make the jump to using public cloud services faster. That’s the reason this concept will get the care and attention of cloud vendors. It’s a stepping stone for the consumption of their core business (cloud computing) that they are selling to businesses.

What’s in it for the business that builds one?

But why would a business I advise buy into this? Well a private cloud (even if used without the self-service component) is Dynamic Systems Initiative (SDI) / Dynamic Data Center on steroids. And as such it delivers efficiency gains and savings right now even if you never go hybrid or public. I’m an avid supported of this concept but it was not easy to achieve for several reasons, one of them being that the technologies used missed some capabilities we need. And guess what, the tools being delivered for the private could can/could fill those voids. By the way, I was in the room at IT Forum 2004 when Bill Gates came to explain the concept and launch that initiative. The demo back then was deploying hundreds of physical PCs. Times have changed indeed! But back to selling the private cloud. Building a private cloud means you’ll be running a topnotch infrastructure ready for anything. Future proofing your designs at no extra cost and with immediate benefits is to good to ignore for any manager/CTO/CIO. The economics are just too good. If you do it for the right reason that is, meaning you can’t serve all your needs in the public cloud as of yet. So go build that private cloud and don’t get discouraged by the fact that it won’t be a definition example of the concept, as long as it delivers real value to the business you’ll be just fine. It doesn’t guarantee your business survival but it will not be for your lack of trying. The inertia some businesses in a very competitive world are displaying makes them look like rabbits trapped in the beams of the car lights. Not to mention government administrations. We no longer seem to have the stability or rather slowness of change needed to function effectively. Perhaps this has always been the case. I don’t know. We’ve never before in history had such a long period of peace & prosperity for such a broad section of the population. So how to maintain this long term is new challenge by itself.

Danger Ahead!

As mentioned above, if there is one thing that can ruin this party it’s is complexity. I’m more convinced of this than ever before. I’ve been talking to some really smart people in the industry over the weekend and everyone seems to agree on that one. So if I can offer some advice to any provider of tools to build a private cloud.  Minimize complexity and the amount of tools needed to get it set up and working. Make sure that if you need multiple building blocks and tools the integration of them is top notch and second to none. Provide clear guidance and make sure it is really as easy to set up, maintain and adapt as it should be. If not businesses are going to get a bloody nose and IT Pros will choose other solutions to get things done.