Cloud Strategy and the things you need to consider
Cloud …. That thing everyone’s been banging on about for the last decade. The thing that would change your life, both at the consumer level and in the Enterprise. There’s no doubt Cloud technology is something that has changed the way in which we can benefit from Internet technology…being able to access information and data from the Cloud rather than through a local network has made everything from remote and mobile working feasible, to leveraging big data and physical hardware like home assistants. But what about in organisations? What are the considerations there when it comes to Cloud?
Twenty years ago, you would hear senior IT people talk about capacity or redundancy. Today you hear those same people talk about containers and zone resiliency. Previously discussions about uptime and ROI filled the board room, now we talk about security, Governance and cost optimisation.
So, whilst Cloud has made it easier to manage all this stuff required to make technology work, the underlying best practices are exactly the same. I see that it comes down to capacity planning, security, workforce skills, governance and management. When I first started getting into virtualisation in the early 90’s, thinking I was the ‘shit’ running VMWare and removing bare metal like it was banned, I thought it’d do away with lots of server management roles and so did my IT peers. We were all spending the money we’d be saving already! All it’d actually done is move the needle into different skillsets and give us slightly different problems to manage later down the line.
Sprawling
One of the problems I suffer from personally and one which is only just now starting to hit companies is ‘Cloud sprawl’. Just like server and service sprawl in the 90’s, I’ve recently had a clean-up of all my old unused (but still paying for) images, containers, and micro-services that I’ve just left. From little test beds on Amazon’s AWS, to that map polling database I had running in Azure, to my sexual health test app R&D in Kubernetes. Paying for CPU cycles, paying for RAM, paying for bandwidth. Can you imagine the cost of this at Enterprise scale? That’s what I’m seeing now.
What can businesses do to realise the benefits of Cloud without falling into a management hell?
Cloud first is fine – I hear it all the time. Everyone wants Cloud first. Of course! But organisations have to consider their existing policies and procedures ensuring they are actually ready for the Cloud. Firms need to invest in people skills and not forget new technology means a new set of skills.
Cost – it’s so complex when it comes to Cloud. When you then bake in multi-cloud solutions and burstable service with unfathomable pricing tiers that no one understands, you have a potential for commercial chaos. Think about how Cloud services should be itemised within your business, project chargeback on a per-project basis. You need to carefully calculate Cloud costs to understand the true ROI for your business.
Governance & Security – Currently the most opaque area especially as VAR’s and Brokers are making it up as they go along with no clear standards, Governance & Security represent as far as I’m concerned an organisations biggest risk. Get this right, from where your data is stored and who it belongs to, to how you manage and track your ROI against the SLA of your vendor. Using multiple vendors? Consider using a brokerage service who’ll manage this for you as often this is a full-time job all of its own.
What I see in the future, when we talk about Multi Cloud, is that it’ll end up so democratised at Enterprise level, it’ll come down to merely cost and you’ll switch services month on month just as you may do at home with your energy supplier. Azure not doing it for you in Q2? No problem, move over to Google Cloud Services. Sick and tired of AWS’s customer service? Off you pop over to OpenShift. The differentiating factor being merely price.
What’s going to be able you to leverage this? Ensuring your application migration strategy is considered in such a way to accommodate this future.
83% Of Enterprise Workloads Will Be In The Cloud By 2020
The way organisations build out their applications now, need a really heavy re-think. So your Cloud strategy needs to be more about not what Cloud service you ultimately consume, but how you’ll staff and train up your internal IT teams to develop applications and services to accommodate this multi-cloud hot-swappable future. Don’t under-estimate the costs there, otherwise you’re just increasing your overall spend by purchasing Cloud sometimes, just for the sake of it.
Let’s consider ‘data gravity’ too
and here best put by Dave McCrory, inventor of the concept says;
Data gravity describes the effect that as data accumulates, there is a greater likelihood that additional services and applications will be attracted to this data, essentially having the same effect gravity has on objects around a planet. Although services and applications have their own gravity, data is the most massive and dense, meaning it has the most gravity. If data becomes large enough it can become virtually impossible to move.
So data accumulates over time and becomes dense while increasing in mass. As density or mass accumulates, the data’s gravitational pull increases. Services and applications also have their own mass and gravity. This causes other applications and services to be gravitated towards data depending on their density.
Think about how you’ll structure out your applications based on their data requirements and subsequent gravity.
The other thing I see being a huge problem is vendor lock in. Let’s remember it’s in vendors best interests to keep their customers dependant on their services much in the same way a drug dealer keeps their customers dependant on crack – (probably not the most elegant analogy but you see my point, and hey, if you’ve made it this far through the article, hopefully you’re smiling too!).
The concept of data gravity above is a key factor to consider when thinking about Vendor lock in, but don’t forget about;
• Incompatible proprietary technology
• Inefficient processes
• Contract constraints
Watch out for those data fee’s. Getting your data in, no problem, getting it out and watch your egress charges escalate to the eye watering level you don’t want to be showing your CFO. This is where the benefit of a Cloud broker comes in, and ensuring your contracts are up to par and benchmarked accordingly against the market.
To be sure you are selecting the right service provider within your organisation, IT leaders should at least consider;
• Ingress and Egress charges and processes
• Make sure your provider is prepared and set up to support emerging standards
• Be clear how your provider will deal with moving large amounts of data
In Summary
The Clouds most probably mature enough now for you not to fear it, but you should definitely consider some of these more detailed nuances when factoring in the detail of your Cloud strategy. Do your homework, consider broker services and watch out for complexity and sprawl.
Get in touch
If you’d like to talk about Cloud or implementing your own Cloud strategy get in touch with me on Twitter @mariodc and continue the conversation. Use the hashtag #mariocloudtalk
References
https://mind-core.com/3-risks-of-a-cloud-migration/
https://www.forbes.com/sites/louiscolumbus/2018/01/07/83-of-enterprise-workloads-will-be-in-the-cloud-by-2020/#3a97d86e6261
