, , , , , , , , , , , , , ,

Leading Edge or Bleeding Edge – Cloud Strategy at the Enterprise Level

This paper is primarily aimed at the “Cloud” and the promises of its capabilities now and in the future. But this paper could be used for all enterprises and infrastructures. What the paper really distills to is, is open mindedness for change and growth along with inclusiveness and collaboration of all parties involved to enhance productivity and business (and employee) success. Yes, sounds like buzz words, but, what I stated is the way it should be in the real world.

Corporate Growth and Success

What is your corporation looking for in your firm’s technology needs – Bleeding edge or Leading edge? Are you a CEO or President who is forward thinking?  Are you a CIO or CTO with solid long range thinking in mind?  Or are you a CFO or COO thinking in terms of long-term ROI and not counting pennies (hoarding) in the short-term in order to ensure a very decent growth curve…? If you are of the aforementioned, then this paper is for you.  Not to worry EVPs, SVPs, Directors, Managers and Supervisors – this paper is also for you too….

However, if you are only working in the ‘here and now’ for the immediate short-term rather than working on ‘now’ AND the future, you – yes, you – you need to reconsider your game plan and change your mindset.

  1. Are you looking for more productivity?
  • faster, larger CPUs in servers // faster bandwidth // faster storage & distribution points

2. Or are you looking for increased cost savings by maintaining –

  • slower, smaller CPUs in servers // slower storage devices // slower network connectivity)?

3. Or are you looking for innovation in going into the ‘Cloud’?

Keeping the above in mind, following are several areas you should take into consideration as you ponder moving to the cloud and virtualization. And, you do not have to be a large enterprise to embrace some of this content, you can be a small or medium sized business as well.

Knock Out Strategy (worth a shot to try – correct…?)

Is the organizations’ leadership looking at right now or 6 months from now (or 12 — 24 months)? This is the overarching goal for any executive. It is what is coming down the pike that needs deep and full consideration, not the temporary immediacy and the cost savings ‘right now’. Because if the leadership is only looking at the immediate time frame rather than looking at the long term gains and benefits, that organizations’ leadership is doing a disservice to everyone around them (employees, customers and even the shareholders).

Keep in mind that if you do not continue advancing your on-premises technology/infrastructure, and bringing your firm’s IT staff and all other employees along with those advances — you have already doomed your organization to a less than energizing future.

Being an organizational thought leader is always, always a desired role to be in. Remember, there is no end­game, it is always an on-going process. It is about strategically planning where you want your organization to go and how you want to do it. Do you want database and storage right at your fingertips? Do you want to have a large number of servers that can scale up, on demand rather your firm purchasing more over a several week period – and planning on what to put on each server and pulling Ethernet cabling to each one (and are you going with Cat5e or Cat 6 or even Cat 7 – maybe fiber…? Your strategic plan has got to cover what is in the pipeline right now (think about DARPA or MIT efforts) because you may likely see that product in the marketplace, soon, what with the way tech continues to take off. Keep Quantum Computing in mind…

Here is an example of one area of strategy, bandwidth. Look at what Verizon is doing, they are bringing gigabit (well, actually, it is near gigabit) speeds to customer’s homes. Is your firm advancing the external and internal bandwidth for future demand? And remember a couple of small areas: 5G is rapidly approaching, meaning that more people will be connecting to your business via a fast hand held computer (phone/tablet/watch). And do not forget this speedy growth in the IoT spectrum — that is an eye-opener, even as we watch it and the world of IoT is basically just ramping up. Imagine, someone in a Class 4 or Class 5 autonomous (fully) vehicle ordering one of your products, via the vehicles connected network (V2V or V2I), which in turn is connected to the ‘Net over a 5G connection via a holographic interface…

Buy-in and Path Forward

Even if you are the executive, do you need buy in at the executive level? You already know the answer is ‘Yes’ – you need to know which executives — CEO, COO, CTO, CIO, President, EVP and CFO {costs & ROI} to gain buy in from (which should be all of them).

What about managers & directors, supervisors & team members? The culture of the workplace and work force is a highly malleable entity and it needs gentle tweaking to help — accept, adapt and grow successfully.

Is the corporation looking to:

  • Improve work flow and productivity, which may mean reducing the workforce? But, retaining that workforce and helping them to grow in the ‘new’ enterprise/corporation will be even more beneficial — loyalty, better retention rates, retention of empirical knowledge…
  • Boost that work flow and productivity with the staff on hand as well as
    expanding those numbers?
  • Migrate to faster, more powerful servers (storage, network bandwidth) as
    technology grows (included in cloud vendor contract)

No matter what, the foundation of your strategist journey is:

  • Planning,
  • Balance – tough choices of staying on-site with hardware/software // moving to the Cloud // obtaining leading edge or bleeding edge technology and
  • ThinkingDEEP thinking, as you plan out the path forward and in balancing what your firm will need – as well as how much focus you will place on the firm’s employees to help them grow, to make the firm grow more successfully. You, and others, will have to think of all the multitudinous options that will face you as you go forward and choose what you believe is best for the firm, the employees and customers (and shareholders if applicable). All the while, keeping a plan B option in your hip pocket (see how well that plays for men AND women – hip pocket).

Brand & Reputation — Key Items

No matter what your level is within the corporation, you will have to maintain the corporate brand. The corporate reputation is at stake, whether it is for successfully moving to the cloud or not moving to the cloud and seeing success move, inexorably, to other firms who are in the cloud.

Brand and reputation is paramount to a firm’s continued existence but then, you likely already knew that…

Economy of Scales On-Demand

Elasticity in your infrastructure is one aspect of economies of scale. Can you rapidly spin up additional servers, rack space for those servers, cooling for those additional servers and possibly routers and switches to name a couple of additional items in your data center? Easily and inexpensively? What about in the event of disasters (fires, floods, theft, etc.), can you rapidly scale up to get back to business in a rapid timeframe?

Is your firm taking a long hard look at how you are going to expand — more hardware, more space and more IT staff? Or are you looking longingly into what the ‘Cloud’ can do for your firms’ business? This is not to say you are looking at the ‘Cloud’ with loving eyes but looking at it with a jaundiced and pragmatic eye as to practical means and outcomes. You are looking at the cloud to improve efficiencies, productivity and the well-used term — elasticity (for which I use for nimbleness and agility — some use the terms of scaling up and scaling out). You are looking at the cloud in order to scale your ‘cloud’ IT infrastructure on demand, when ‘you’ need to expand or contract.

Make no mistake about it; if and when you do move to the cloud, you still have to work on having a successful cloud infrastructure — successful to your customers and to your business.

Good Solid Cloud Contracts

Be sure you work out the best possible contract you can with your cloud vendor. You want to cover items such as (not all inclusive):

  • Length of contract
  • DR/BCP (see below)
  • Security — (if not AWS, Google, Microsoft) who can access the facility (vendor & your firm’s staff), rack access to servers, who SSH in
  • Encryption — who manages the keys (rotation, add, delete, store)
  • Updates — software/hardware — who is responsible, frequency of updates, mandatory updates
  • Hardware/software advancements — is that cloud vendor going to include that as cost of doing business with you or are they going to charge you for those advancements (i.e. moving to the latest Xeon CPU or the newer SSD drives)
  • Failure & Emergencies — vendor site — fires, power outage, failover/redundancy
  • Cost of services — tech support, access to storage, queries to databases, storage to content distribution networks, bandwidth speeds, monitoring tools & metrics, audit tools, etc.
  • Early contract cancellation — penalties
  • Migration costs to/from cloud vendor (lock in issue/incompatibilities)

Security To Bank On

MFA, SSO, self-help log in, tokens, biometric, etc. — these are all items to ensure better security. A significantly important component to any business is data encryption. Encryption in the cloud is not the scary beast that it used to be, it is now an easier process than ever in the past, especially when your cloud vendor can take of most of that effort for you (i.e. AWS and their different offerings of encryption management and control).

Data at Rest, Data in Transit — both are covered very adequately, as long as the organization’s work force (administrators, developers, engineers) ensure to enable and consciously use encryption capabilities in the cloud

Your firm can even turn on deletion protection in case someone inadvertently (hopefully) deletes files that should not have been deleted. But, as a fail-safe, for your data in the cloud — you could always turn on policies and/or access control lists (ACL). These ACLs and policies allow for permissions as to who can and cannot delete (or read or copy) specific data.

Monitoring — At All Times

If you are moving to the cloud, this is a must have, same as it is for your on-premises hardware. With the cloud environment, being more spread out — you want a monitoring capability to see what is being used and by who and what is not used and from where (what cloud region) in order to spin it down to reduce costs (this should be part of any auto-scaling methodology).

But you also want to have the capability to use/build dashboards, for the staff running the Cloud environment or for the executive team in order for them to develop strategies for the direction the company needs to move in.

Using a software tool like Splunk in addition to your provided cloud monitoring tool will help your firm pinpoint multiple areas of your cloud usages and problem areas.

Auditing – Ditto

Here is another significantly important must have. You will not want to skimp on this area — you want to obtain the best possible auditing capability you can gain. You want this capability to ensure that users are accessing properly authorized resources and in turn, to track who is unauthorized and attempting to access resources they are not allowed to access.

You will want this feature to audit costs, where resources are being heavily used, where those resources are being under-used and even more importantly, where resources are fraudulently being used.

Compliance – DITTO

Doing compliance (various local, state, federal and international laws, rules, regulations and acts) efforts in the hardware arena is a big challenge, but it can be an easier challenge in the cloud because of the simplicity of manipulating data via software controls/management.

You just have to ensure you have compensating security controls to protect all of the data involved and who has access to manipulate that data.

DR / BCP — Just in Case, Right

Where, when, who, cost, testing (monthly, bi-annually, annually). This is another extremely large area and suffice it to say that you will be required to duplicate most if not all of your current business functions and needs. This is an area that many corporations fall down on in not planning, very diligently, and then testing on a periodic (constant) basis — as well as UPDATING the DR / BCP plans.

Redundancy and high availability is a requirement for every firm, not just for day-to-day use but it must also be covered for DR/BCP as well. This is not a “when we can get to it” item.

The business world is moving more and more towards CaaS (Code as a Service) or as some call it, IaC (Infrastructure as Code) and it is not at a glacial pace either. Using code to build the organization’s desired infrastructure is what more and more (again) enterprises are moving towards. Why spend the massive CapEx outlay for customized (or not) hardware and software that is not being used to near full potential? Why not use that money for more dynamic OpEx instead?

You can use code to spin up the desired infrastructure and use those spun up servers, storage and network to near full potential and then scale up (or down or out) again as needed — on the fly.

For example, you can spin up additional servers at will, to:

  • Modify the size to larger servers in near real time
  • Scale up your storage volumes on the fly (those storage volumes attached to the servers, i.e. EBS)

Following are a few of the numerous cloud service offerings (primarily for the IaaS model) to highlight what is possible.

Severless (Lambda) or EC2

Is the organization going to maintain servers on site? Are you willing to continue buying blade servers with various CPUs (and upgrade and license them)? Are you willing to deal with data center heat island effects and mitigating that issue (cost of cooling is not inexpensive)?

With serverless, you do not need to have dedicated EC2 servers, your code will allocate/deallocate whenever those EC2 servers (and other resources) are required.

EC2 Reserved Instances would be an advantageous gain “if” the organization has a need to always have a demand for instances being available — this would be in addition to EC2 server instances on demand usage.

CloudFormation or Python

What type of coding are you going to do? Are you going to build your infrastructure in the cloud using code? What kind of code?

DynamoDB or Aurora

What specific kind of database usage does your business require, along with what kind of DB management levels you are willing to experience, pay for…? Simple or complex usage?

RedShift               What kind of analytics are you looking to do, warehouse level efforts or something less than this large software tool

Be aware that storage, instant retrieval or long-term retrieval costs are declining — not just due to competition (i.e. Google’s Reginal, Nearline & Coldline and AWS S3 & Glacier) but also in part due to advances in SSD & RAM storage density (i.e. Intel & Micron’s 3D Xpoint TLC) – why would you buy spinning HDD when you can have faster SSD and move to more & faster density when the technology advances (in near real-time), compared to outright owning those HDD.

Just as well, there are storage vendors out there that will attempt to persuade you that owning your own HDD/SSD on-site has become an economical advantage as costs come down (manufacturing, fabric to facilitate speed of data read/writes, lower latency). Even energy consumption and heat output has improved.  This is all well and good and may be true and advantageous for some firms. The point is, do your homework, put your eyeballs on data that backs up the various claims. You will want numbers – statistics, graphs, use cases/case studies to bear out taking on more hardware. You will need to look at the long-term depreciation costs. Basically, you will need to do your homework and weigh the pros & cons as it fits your organizational goals as to which to select – it could be a hybrid decision at the end…

Big Data

Do you have a need to work with data from all avenues (this is what Big Data is all about) — images, text, PDFs, application output, machine language, video — from social media, from other business partners (B2B), etc.

Machine Learning

What about working with large datasets to teach computers/robots to take over mundane (or extremely complex operations — brain surgery)

Complexities        Multiple clouds (IaaS, PaaS, SaaS) / hybrid cloud (on-prem & IaaS)
Commodity equipment on premises
Redundant Internet connectivity

Not as common today but you never know when a local telco or national or international IXC circuit will take a hit. For example, a fishing trawler severs an international ocean bottom cable/circuit — yes, it has happened in the past or someone is digging on public areas, just where a telco’s circuit happens to run through.

High bandwidth    To the desktop on site


The Future

Even though Hyperconvergence has already been put into play in various firms, it is still a growing area as the components of hardware is improving significantly (i.e. 3D Xpoint memory density and speed — SSDs, RAM —switching and throughput speeds still going up). You should keep an eye on this if your firm is looking to retain on-premise hardware.

Hyperconvergence is where code plays a key role; this code tightly couples the CPU, storage and networking on a commodity box, an inexpensive but powerful unit to handle the workloads.

Hyperconverged equipment however, takes some deep thoughts as how long does the corporation want to continue on purchasing/leasing on-site hardware? This is noticeably visible in the consumer market space where laptops and desktops are declining. Tablets and smart TVs are taking up a lot of slack as they become more powerful and view-friendly. Smart phones are a different matter, many of us cannot and do not want to do any extensive work on a form factor that small (keyboard and screen space).

Also, SDN/NFV, software defined networks/network function virtualization is continuing to make large strides to improve the networks (efficiencies & throughput) of the world.

This is an area where again, you (or your telco carrier) are using commodity equipment but offloading the network functionality from the equipment and over to code, which can be managed remotely (not just remoting in to the device and using UNIX or Cisco IOS commands). It is so much easier in working with modularized code instead of working with equipment that becomes obsolete nearly as fast as you open up the box the equipment arrived in…

Conclusion – or rather, a Never-ending Ending

As you can see, just from this small write up, there is a great deal to transitioning to and working in the cloud.

Reduction of Cost (CapEx) and revenue growth are going to be the overall winners in having a cloud solution bolstering, enhance or flat out replace your traditional data center. No one is likely to dispute that claim. This is a matter of ‘cost savings through efficiencies’ and yes —”right now”, some services and/or resources may seem expensive but when you factor in the costs of all traditional data center components, you will see some savings. And at times, you will see significant savings, primarily over time.

Beware though, you must ascertain the hidden costs (the intangible areas that you may not consider to have an associated cost) of moving to the cloud — getting resources out of your site, doing hybrid-cloud or going all in to the cloud.

This stems from what are you looking to do in the future in comparison to what you are doing right now? Having your own hardware components means keeping up with:

  • licensing,
  • under-usage of your CPUs/servers (i.e. somewhere around 20% for server utilization {from Gartner in 2015})
  • wear and tear of components,
  • cooling expenses (the more hardware components means more cooling),
  • space requirements,
  • power requirements,
  • security of some of those components,
  • cabling to and from the components (servers, routers, switches, blades in racks, storage),
  • troubleshooting hardware, copper and fiber connections,
  • weight of all the components in the data center,
  • upgrading components at the end of life or really, at the end of a depreciation cycle for CapEx

Buying commercial software will mean keeping up with:

  • licensing (per seat or per employee) for the current corporate staff and then future staff (and if you reduce staff will you be able to reduce the license count to save money),
  • adding in new features,
  • upgrading

Training your staff will mean keeping up with:

  • Changes in software and hardware features that keep growing as manufacturers, vendors continue innovating their products — training of staff to maintain changes and growth in technology — newer versions of Ethernet cabling, newer Top of Rack switches (leaf and spine), new routers, new enterprise switches,
  • Time off for employees to gain that training,
  • Travel expenses — meals, hotel, car rentals, time and/or mileage depending on where that training is (unless you bring the instructor to your facility which may save some money),
  • This is a given at any rate — you will want to train your firms’ teams to learn and to perform better. Space:
  • How much data center space do you need, want or require to do business?
  • How much do you want to lease? Will you be expanding in the future? Will you move to new facilities? What about downsizing?
  • Can your lease allow you to reduce data center space as you move to the cloud (if you move to the cloud)?

Bottom line, it is about the Total Cost of Ownership (TCO for those not familiar) for your on-site equipment and software AND your cloud expenditures. But you want to be able to tap into some of the latest leading edge, not necessarily the bleeding edge, technology (software and hardware) in order to harness and profit from that investment.

You want to be nimble in your strategy (and budgets) as demand for your services and products grow.

And should any incident (or catastrophe) happen to come your way, you most definitely want to be able to spin up your redundant resources in another availability zone (if need be, in another region) …

You want to have high availability in making your customers (and staff) experiences agreeable, enjoyable with cost added value while at the same time, grow profits for the firm.



  • CapEx Capital Expenditures
  • BCP      Business Continuity Process
  • DR       Disaster recovery
  • MFA      Multi-Factor Authentication
  • NFV      Network Function Virtualization
    OpEx    Operating Expenditures
  • SDN     Software Defined Network
    SSO      Single Sign On
  • V2V      Vehicle to Vehicle (i.e. communication from car to car, car to truck and on and on)
  • V2I       Vehicle to Infrastructure (i.e. communication from car to building or street lights)