The Key to Private Cloud Is Removing IT Stratification
One of the leading problems plaguing IT organizations is the high costs of operations and maintenance. The industry average is roughly 70% with some organizations going as high as 90%. Picking apart these costs one often finds a stratified organization focused on narrow bands of computing with little crossover between the bands. Moreover, the weighting of political density between layers often makes it too risky for basic collaboration between the stratified layers. Hence, when problems arise, each layer attempts to solve the problems only with the tools at their disposal. The result is the Operation Petticoat wired together with chewing gum and bras that we call IT.
JP’s IT Axiom #124: Design flaws at the top of the stack will highlight limitations at the bottom of the stack. Likewise, the design at the bottom of the stack impacts performance at the top of the stack.
There’s no escaping the fact that a poorly-designed application will put undue burden on the operating infrastructure. A “chatty” application impacts bandwidth. Improperly designed database queries will consume memory and disk capacity. Poorly-designed storage architecture will limit the amount of I/O per second (IOPS) and, thusly, limit the speed of retrieval of data to the application. IT transformation is about moving from a stratified organization to an agile organization through the use of DevOps culture and other collaborative techniques.
Short of correcting this organizational challenge, the stratified layers will attempt to correct issues using the tools at their disposal. Hence, infrastructure & operations (I & O) will scale linearly with memory, servers and storage to correct design flaws in the application. Software engineering will add specialized code to work around limitations in the infrastructure, such as timeouts and latency. Removal of the stratification in favor of collaborative teams means that issues can be rooted out and solved appropriately.
Moreover, this stratification has greater implications for delivery of private cloud services to the organization. Indeed, while many organizations focus on delivering Infrastructure-as-a-Service (IaaS) from their private cloud, it begs the question, “What is cloud strategy for the organization?” IaaS implies that the consumer will manage their own applications in the cloud and that IT is simply the supplier of infrastructure services. I posit that this is merely an extension of the stratification of IT with the I & O layer delivering within their swimlane. However, it misses the greater opportunity for the business a whole, which is to deliver reliability, quality, trust and scalability for data and applications in a consistent manner.
Hence, IT organizations should be focused on delivering Platform-as-a-Service (PaaS) to the business as this will provide a consistent way to design, build, deploy and manage applications resulting in lowering operational overhead while delivering greater overall agility. By delivering IaaS, the business loses the opportunity for this consistency as engineering teams are now responsible for building and deploying their own application runtime platforms. Even if a single vendor’s application platform is used, the various configurations will make it more difficult to support, lead to longer repair cycles and add undue complexity to operational concerns.
Private cloud computing represents a unique opportunity for the business to reduce operating overhead significantly through the three C’s: consolidation, consistency and congruence. To achieve this goal, IT needs to break down the stratified layers and formulate workload teams comprised of members from various parts of the IT organization and together become responsible for the workload’s availability, performance and consumer experience.
IT-as-a-Service and the Facebook Effect
I’ve recently been theorizing around a new model for IT transformation. There’s anecdotal evidence that, in general, business problems tend to change slower than the rate of technology innovation. Thus, we can discern that IT has focused on the application of technical innovation to solve existing business problems in more effective ways versus using technology innovation to solve only new problems or continually having to evolve the solution to an existing business problem.
The figure below illustrates a generally-observed pattern in IT. Existing problems move to the new platform under the guise of “technical refresh” while technology innovation introduces new business problems to be resolved. Ultimately, what we learn is that there’s very little stickiness with legacy platform selection and that users will eventually attempt to migrate their solution domain forward onto the latest solution set in an attempt to derive lower costs, easier support, better performance and overall improved customer experience.
However, something interesting is occurring in this most recent migration that is best described as service-orientation. This change has profound impact for the IT industry that we will call the “Facebook effect” for lack of a more widely-understood term. The Facebook effect is best explained as follows:
In general, small populations of Facebook users will leave the service completely and , for many, overall time on the site may diminish, however, Facebook, as a service, is seemingly entrenched to a point where it cannot be unseated.
That is, another service will most likely not emerge offering the same function and capabilities driving a migration of users away from Facebook to this new service. This is because the user has invested in customizing the service to a point where it is extremely painful to recreate in a completely separate service.
Over time, Facebook adds more features and the underlying performance is improved through technology innovation and continuous platform improvement, but the consumer of the Facebook service is relatively unaware and unaffected by these changes. In contrast, today, enterprise IT users are very aware of underlying platform changes. For example, a migration away from a client/server version of the application to the Web-based version of the application will represents a significant shift in user experience.
As IT organizations start to adopt, and more succinctly, perfect, a more service-oriented approach toward IT delivery, I believe, they will start to experience the Facebook effect within their own domain. There will be less significant change at the service level with more material changes continuing to occur in the platform. Moreover, IT will start users will start to invest their time in building connections with and automating their processes around the service. This will greatly limit enterprise IT’s ability to arbitrarily change the service in a way that impacts the user. The net impact of this on the business is yet to be seen, but in general, it is clearly the change that we have heard forecasted for years about business taking the lead in information technology direction.
The Path to the Intelligent Cloud
Cloud Needs Application Architects To Understand IaaS
Application development has been moving in the direction of platform abstraction. That is, the need for developers to have detailed knowledge of the infrastructure that the application was being deployed on was becoming less important with increasing sophistication of the application platform for which they were developing. Cloud computing is now reversing this course of action, at least in the short term.
Actually, the platform abstraction is a bit of a misnomer since the implementation resulted in operations struggling to tweak the infrastructure to meet performance requirements. Additionally, most applications typically had their own dedicated hardware allowing for specialization to meet the needs of the applications deployed on that hardware.
So, more accurately, cloud computing illustrates the flaws in the approach of pure platform abstraction and a ‘Chinese Wall’ between application development and operations as operations now has fewer tweaks at their disposal to make an application perform in a multi-tenancy environment. Hence, it is imperative that application architects begin to incorporate into their design the impacts of operating in the cloud into their architectures. Application architects must be able to understand how the application will perform given the environment that the application will be operating under.
Impacts that application architects will need to think about in this cloud world include:
- Databases – running a highly-available database in the cloud is a daunting task; especially without direct control over the storage. Environments like Amazon offer database services that deliver greater performance than can be achieved if you put up your own database in their IaaS, but there are also pitfalls.
- Software failover – applications can now implement failover far less expensively using commodity hardware. Hence, failover should now be developed into the application instead of relying on the application platform or hardware infrastructure. Given that application architects have not focused on this use case in many cases, it will require some education and experience before this can become common.
- Virtual networking – virtual networks enable the application development team to take control over their own application’s networking infrastructure. Once again, the lack of experience here means that there are likely to be many misconfigurations that impact the performance and availability of the application in addition to enabling security flaws.
- Instrumentation, logging and monitoring – these are areas that the application development teams have been pushing responsibility off onto the application platforms. However, without visibility beyond the hypervisor, it’s imperative that they incorporate this back into the applications or they may have significant issues troubleshooting or auditing their applications.
As my famous Uncle Winthrop liked to say, “Now that I've given you a band saw, I need to teach you how to use it or you will just be wasting a lot of wood and in the worst case might lose a few fingers.”
Why Poor Data Classification in Government Will Impact BYOD
In recent discussions with IT leaders from both federal and Department of Defense sides of US government, representatives stated that they are having a heck of a time accommodating expansive growth in mobile computing. This is critical given that today, in most cases, agencies and departments still have control over which mobile devices can be used. In the future, these executives realize that the changing demographics of contractors and employees means they will not only need to support continually growing traffic, multiple presentations and increased asset management, but will also have to deal with a wide spectrum of mobile devices due to Bring Your Own Device (BYOD).
This idea that these executives will one day soon have to loosen their grip over endpoints is a major concern. Contrary to belief it is not about power and supremacy over their domain. Most users have no concept of the level of complexity for managing access and availability of data and applications when there is no control over the endpoint; nor should they. While network security solutions have improved dramatically over the past decade, improper use of the tools and ever increasing abilities of hackers means that “locking the front door” isn’t good enough to solve this problem by itself.
One of the keys to solving this issue in a way that doesn’t alienate users, but also ensures confidentiality and security of government data is going to be segmentation. Segmentation is the act of distributing the data across multiple tiers from unclassified to compartmentalized and providing greater levels of restrictions on access at each layer. For example, unclassified information should be hosted in the public cloud with no permanent connections back to any data center housing higher classified documents. For Official Use Only (FOUO), Confidential, and Secret data should require minimum Virtual Private Network (VPN) connections for access. This means that the mobile devices must support the VPN protocols in use in order to establish a connection. Finally, Top Secret and above information should require on-premise wireless or wired support only combined with two-factor authentication, VPN and the ability to remove any data downloaded to the mobile device when the VPN connection is broken.
However, the biggest issue for government is going to be segmenting data that is either improperly classified or comprised of various levels of classifications. I was once part of a project where the requirements documents were improperly labeled FOUO, which raised many problems sharing them with foreign counterparts even though the project was developing a collaboration portal to work with foreign government officials through what was going to be a publicly hosted application. This is just one small instance of tens of millions within the government. Moreover, it seems more recent projects have seen serious disagreement among government IT employees and contractors as to what are the appropriate classification levels for certain pieces of data. In one of these cases a very junior security professional within the government was demanding aggregated publicly available information could not exist in a publicly hosted cloud.
I don’t relish these government IT executives position with regard to the growing mobile demand. BYOD is going to amplify this problem exponentially.
Adventures in Cloudwashing: Are You the Cloud Or in the Cloud?
Anyone who is working intimately with cloud computing and having critical conversations regarding this medium will eventually be party to the “cloudwashing” conversation.
Cloudwashing: the activity of associating all your products with cloud computing even though it doesn’t meet core attributes that define cloud computing as created either by de jure or de facto processes.
Depending upon your role in information technology cloudwashing is either a major concern or a significant benefit. If you are a product vendor looking to gain attention for your product, then cloudwashing is very useful. It opens up interest in your product to an entirely new audience that may see benefit in your offering even if it really doesn’t align fully with cloud computing goals and values. If it attracts eyeballs without significant additional costs, then it’s positive.
If you are an IT manager, director, vice president, CIO or other, and your goal is to say that you’ve successfully delivered a cloud computing initiative, then cloudwashing is good. After all even if it’s traditional managed services, if you can label it cloud, the pure confusion and lack of agreement on what is cloud will work in your favor. That is, who can argue with you that your initiative isn’t cloud computing if they cannot define cloud computing in a concrete and definitive manner.
If you are a security officer concerned with privacy and security of your businesses’ data and access, then cloudwashing may work in your favor as well. Most cloudwashing is typically managed services repackaged, which means it doesn’t really incorporate key attributes of cloud computing that concern security professionals, such as multitenancy and shared hardware platforms. So, the security professional who cares less what the business wants to say they’ve accomplished as long as they’ve done their job securing information and access may be better served by a cloudwashed product or service than a real cloud computing product or service.
So, why should we care if a business wants to cloudwash their offering? For one, its disingenuous, and anyone that’s willing to sell you something in a disingenuous manner should trigger you to question that company’s values and ethics. Ultimately, it could mean the difference of them being their when you need them and you out on the street with your briefcase in hand looking for your next job.
Secondly, it damages the entire industry. While you can technically argue that there’s no one single globally-accepted definition of cloud computing, NIST and Cloud Security Alliance are two very credible organizations that have settled on a common definition. There may be others also promoting this definition as well. So, to imply that there’s no standard undermines the work of these organizations in attempting to help businesses and individuals understand cloud computing. Moreover, without a common agreement, there is no means to develop agreed upon metrics for comparing offerings from different vendors.
Finally, there should be delineation for vendors that are in the cloud and those that are the cloud. This differentiation is unclear when vendors cloudwash their offerings. Being ‘in the cloud’ means that you are selling a service that relies upon yet another provider’s cloud computing service. This is important for businesses to know that are signing up for a service. In the same way that you would want to know that FedEx or UPS uses a third party delivery agent in a foreign country and that they are not handling your package from start to finish, you should want to know who ultimately is touching your data and applications. Cloudwashing makes it difficult to distinguish those who are “in the cloud” from those that “are the cloud”.
Of note, I know there are those that will argue that you’re buying a service and you shouldn’t care, but that’s a naïve opinion that demonstrates a lack of understanding for the business requirements for audit and compliance. Maybe some startup in Bumpuck, Nowheresville that’s hosting a free service can get away with this approach, but real businesses that hire professional accounting and auditing services are responsible for understanding the implications for using that service; this often means knowing what software and services is being used to provide a service.
Start Building Your Next Generation IT Department Now
David Johnson’s Blog piece really got my goat. In this piece, “Meet Jamie - A HERO With The Power To Force Change,” Johnson paints a sales representative that has rejected his IT department’s choices for device support in favor of an unsanctioned Bring Your Own Device (BYOD) strategy as a hero. Frankly, I believe Johnson does a major disservice to the IT industry with this piece, once again painting them as inept and unable to keep pace with the speed of business.
As someone who has consistently been advocating for pushing the envelope within IT, my goal has been to help business establish a framework for operating in a rapidly changing industry and meeting the expectations of its users. And, like a good mentor, I can push my students to reach their potential, but don’t you come bash them for not being able to rise to the challenge.
How many of you out there believe your IT department is a hindrance or a hurdle to you getting your work done? Okay, let’s put that feeling to the side for one minute. It’s a valid concern, but let’s look across the table at the challenge from ITs perspective.
- Operational costs are continuing to rise and continually consuming a greater portion of the annual IT budget
- Technological shifts are coming faster and are more disruptive with each shift
- In many businesses, the same IT group is responsible for desktop support, mobile device support, application management, operations management and telecommunications
- The IT department is supporting many different businesses, not one. From the IT perspective, marketing, accounting, executive, sales, distribution, warehousing, supply-chain & logistics are all different businesses
- Enterprise class software and hardware generally sucks
Now, let look at prioritization of efforts in the environment I just described. Who should get the most attention? In the realm of this, does it seem reasonable that they may not be organized in a way to support your desire to bring in the latest mobile device or tablet?
All this aside, let’s turn back to the answer we put aside earlier regarding your concern that IT is a hurdle or hindrance to getting your job done. The reality is that you should expect that the business provide the appropriate levels of support to ensure you operate with maximum productivity. It’s in their best interest and yours. This is not a question of “should we”, but “how can we?”
Frankly, I believe IT departments have no choice but to innovate with regard to how they staff, organize and deliver service. The old world nature of silos of expertise are toast. Lack of cross-domain understanding of IT-related issues accounts for 90% of slowdowns, lack of response and service disruptions. This type of change cannot occur overnight. It must be an evolution unless the business is also willing to accept service disruption as a possibility in order to speed the change within their own organization. I am not stating this facetiously either; it is a risk management decision that may be viable and worthwhile if it will enable the organization to triple their efforts and operate more effectively.
It’s time to put away the pitchforks and torches with regard to the IT department. The best thing you can do to help yourself is not become a rogue BYOD agent, but convince your management to contribute to innovation investment within IT to support your needs for greater productivity.
Become the Platform
Steve Yegge, a Google engineer, recently posted a long rant on Google+ about how Amazon does everything wrong and Google does everything right. Probably the most traffic generated for Google+ since they launched, which is why he most likely still has a job. While it was painfully excruciating to get through, I wanted to make sure I read the entire entry because the focus wasn’t really on Google at all, but on a transformative idea of Jeff Bezos, founder and CEO of Amazon.
As Steve points out at some point Bezos got “it” and he realized the power and the value in his company wasn’t just to be an online retailer, but to the be THE platform for online retailing. Now, to fully understand this, Bezos didn’t say to his team build me a platform that handles every aspect of online retailing including managing the supply-chain and facilitating third-party suppliers to sell directly through Amazon. What he said was simpler than that, but significantly more powerful; he said, every team in the business will expose their data via a service interface.
Now, I’ve written about Enteprise SOA till I’m blue in the face and still had to deal with arrogant dweebs argue about REST vs. SOAP or top-down vs. bottom-up just more people who didn’t get that SOA is an architecture that should be applied to the business, not to just the software. Now, I have the greatest proof point imaginable for my argument, Amazon is the embodiment of Enterprise SOA; no two teams can communicate data without going through their own public interfaces or face termination. I guarantee that format, protocol, size, shape, smell, whatever attribute you want to convey about SOA all became irrelevant after a few months of your fellow workers kicking the crap out of you for having a broken or dysfunctional interface.
This isn’t even the best part of the story; it’s just the beginning. What Bezos effectively created by this one mandate was to turn Amazon into a platform. Amazon today is probably the most powerful retail platform in the world. The underlying software helps that platform to run smoothly, but the platform is more than the software. It’s the people, the processes and the technologies working together in harmony to move products from buyer to seller in both physical and digital forms.
Additionally, what Amazon realized in due course of this act is that the computing platform developed to drive their retail platform can also play a significant role in helping other businesses become a platform as well. Hence, Amazon Web Services is the embodiment of that effort providing the same methodology as Amazon the retail platform uses to the world-at-large.
On the other side of the jungle sits a million pound behemoth attempting to stay valid in this fast moving cloud market, where lots of small and mid-sized competitors, as well as some large competitors, are all already vying for leadership positions. Companies, such as Dell, Google, HP, Oracle, Cisco, Microsoft, Unisys, and Harris are established firms with solid client bases that are all looking to deliver cloud services to the enterprise. So, what can IBM do as a Johnny-come-lately to the cloud game to compete in this arena?
Obviously, IBM believes it’s been playing in this cloud game for some time, but perception is reality and when people talk cloud, IBM is typically not part of the conversation. And, that’s when it hits me, IBM needs a BHAG (Big Hairy Audacious Goal) to turn this perception around. They’re not going to win this game by throwing money and people at the problem, it’s a different world led by a different mindset and Grandpa’s enterprise computing approaches aren’t going to cut it. IBM needs to become the platform! They need to embody everything they know and have been delivering for the past 100 years, package it and deliver it through service interfaces. They need to make every team work with every other team only through service interfaces. And, most importantly, they need to change the conversation from “what is cloud computing” to “what is cloud computing about”.
Now, I suppose this same approach could work for HP and Microsoft as well since they too both struggle to stand out against in the field of cloud computing. However, at least HP and Microsoft are part of the conversation. Maybe I’m just rooting for the underdog like I always do, which is why I still haven’t given up on the Washington Redskins … yet! It would be fun to watch a behemoth like IBM come stomping from the backfield, crushing those with existing market penetration and moving to the front of the pack to compete against the leaders in cloud computing.
Cloud Conundrum: Private Cloud Computing With A Pay-for-Use Model
One of the touted benefits of cloud computing is supposed to be that it is a metered service. Much like your water and electric, the cloud is supposed to allow users to access compute resources as needed and be charged only for what is used. This is a great model for users of cloud computing since the risk is nominal. However, it’s a very costly investment on the part of cloud service providers since they need to create, market, manage and support the service without any guarantees that the service will be used. Moreover, pricing this usage so as to be profitable can be a complex initiative since users expect to pay less than it would cost for them to acquire and manage the resources themselves.
Now, some pundits and vendors can talk all they want about what they believe cloud is and isn’t, but the truth of the matter is that customers are speaking up, especially in the federal government, and they want usage-based charge models. Here’s the rub; many of these users also only want to be the only tenant. Hence, we have customers that want a low-risk, private cloud solutions on one hand and cloud service providers that offer either public usage-based charge models or private temporal-based charge models on the other hand.
The answer to this problem is not a simple one. The reason why these two options exist is to balance the risk. If you want to be the only tenant, then many cloud service providers are willing to share the risk, but want commitments for usage based on time. These contracts are analogous to your mobile phone contracts, where there are fees for early termination. Meanwhile, the cloud service provider can significantly reduce the costs and offer a usage-based charge model if they can share the resources among a wide enough audience such that they are likely to have a very high utilization rate.
However, recently I have learned that there are some companies, such as ESCgov, that are offering up alternative means of acquiring private cloud computing services based on usage. Due to the complexity these alternatives are not one-size-fits-all, but instead, are highly-dependent upon the individual business opportunity. Each opportunity needs to undergo an underwriting process that examines multiple variables, such as expected usage, other existing options for acquisition, and the private cloud architecture. Given that these businesses can qualify that the consumption patterns exist, they will develop custom private cloud services for customers that charge based on usage-based models.
I recently watched a video from Stanford University’s Entrepreneurship program in which the speaker made a very interesting statement, “big problems = big opportunities.” While challenging, answering the call from customers for private cloud pay-for-usage models could lead to the creation of the next Amazon Web Services.
Cloud and Tablets Favor the Content Publisher, LANs and PCs Favor the Content Creator
I recently added a Vizio tablet to my list of technological acquisitions. It’s a relatively good Android-based tablet that is very reasonably priced compared to equivalent functional models. However, I realized today a pattern emerging regarding my usage of the device--I’m more willing to pay for content when using my tablet than I was on when using my PC.
The primary reason for the tablet investment was to gain a hands-on experience with what the future is shaping up to look like. I would not be the first to state that tablets are consumption devices, but what I haven’t seen clearly stated is that the future is rosy for content publishers, such as the music industry, book publishers, magazine publishers and websites. While these devices are nothing more than scaled down operating systems, they are clearly not designed to offer a multi-windowed multi-tasking experience. This means that it’s highly unlikely you will find a burgeoning market for the creation of content, such as movies, music, art, books, and even blogs that run on these devices.
I am not saying some application developers won’t develop tools to support creation in this manner, but I am saying I don’t believe there will be a large audience of users who will be creating content on tablet and mobile platforms. Additionally, as cloud-based services continue to emerge for the storage of tablet and mobile content, such as Amazon’s Cloud Player and Google Music, the barrier for acquiring and loading the content on alternative devices through alternative means increases when compared to the ability to click a button and have it show up on your cloud storage system.
The thought of acquiring a CD, ripping it and then copying the songs onto my phone or ipod is already too burdensome a task compared to clicking one button, running the Amazon MP3 application on my tablet or phone and voila -- my music. Plus, these devices don’t come with typical peripherals needed to perform these operations. The tablets and phones that do have USB ports are for tethering to PCs for maintenance and loading, not extension.
That said, I will never consider writing even a blog entry on that device because I would most likely want to slit my wrists before finishing the first paragraph. Heck, I can’t even stand IM’ing on that device because of the pain of typing on it. Which means that while the future is less rosy for the PC market, there is still going to be a need for power machines to help develop the content that is being consumed by our phones and tablets. It also means that I’m going to want responsive access to the files that I’m working on and most likely will not desire to wait for this data to be transferred over the Internet. So, this data will continue to be stored locally either on the PC or on a local area network, which may or may not be backed up to the cloud.
All-in-all, the tablet and mobile computing market has dramatically shifted the balance of power in favor of the cloud. Indeed, it might be fair to say that it’s these devices that have really breathed life into the need for cloud computing since most users are consumers, not producers, of content. The one gray area of course is social media, which turns all users into content producers, but typically at a level or volume that is achievable on the mobile device specifications and uses alternative input effectively, such as camera and audio. This begs the question, if the PC/LAN era is shrinking due to this fact, can it still be considered commodity or will it become specialized and will equipment for publishers cost more in the future due to less volume?
Disclaimer
My Most Recent Books
:Other Links
Web Profiles
Recent Posts
- The Key to Private Cloud Is Removing IT Stratification
- IT-as-a-Service and the Facebook Effect
- The Path to the Intelligent Cloud
- Cloud Needs Application Architects To Understand IaaS
- Why Poor Data Classification in Government Will Impact BYOD
Tags
Recent Comments
- JC Oberholzer on SOA is Dead, Long Live Distributed Computing
- Windows Azure and Cloud Computing Posts for 5/3/2012+ - Windows Azure Blog on The Key to Private Cloud Is Removing IT Stratification
- Gustavo Tomás on SOA is Dead, Long Live Distributed Computing
- Beware the 'Facebook Effect' when service-orienting information technology | ZDNet on IT-as-a-Service and the Facebook Effect
- JP Morgenthal on Cloud Needs Application Architects To Understand IaaS
Blogroll
Other Links
Archives
- April 2012
- March 2012
- February 2012
- January 2012
- December 2011
- November 2011
- September 2011
- August 2011
- July 2011
- June 2011
- May 2011
- April 2011
- March 2011
- February 2011
- January 2011
- November 2010
- October 2010
- September 2010
- July 2010
- February 2010
- January 2010
- December 2009
- November 2009
- October 2009
- September 2009
- August 2009
- July 2009
- June 2009
- May 2009
- March 2009
- February 2009
- January 2009
- September 2008
- August 2008
- May 2008
- December 2007
- November 2006
- April 2005
