How To Select The Right Cloud Service
This autumn and winter have been rich with opportunities to discuss cloud strategies with many different audiences; from CIO’s to CFO’s and LCA’s. I have experienced a big demand for clarifications on the question(s): What can I put in the cloud and what´s in it for me? I put together a recent blog post on cloudadvisor.se which I’d like to share with you. I will start with the “What” part…
When you are considering Infrastructure as a Service there are some clear areas that are suitable:
- Seasonal peaks in IT demand makes a perfect business case for IaaS. This is the easiest way to also understand the “What´s in it for me” part of the question.
- Evaluation of software solutions can be demanding and consume a lot of HW resources. Every piece of SW have specific demands and the needs are clearly periodical. Since these kinds of test are of paramount importance for every organization, IaaS can be used to perform the testing with all capacity needs.
- System testing is even more demanding in shorter periods of time, than SW evaluations. perfect IaaS candidate.
This scenario holds many benefits that are crystal clear: From the full lifecycle platforms to specific technology enabling platforms. The full lifecycle platforms typically include (and integrate):
- A workflow engine
- A set of development tools
- Testing environments for your application development projects
- Possibility to integrate databases
- A rich range of third party tools and/or services
In most cases the initial development part comes with no costs. So, clearly this is a very attractive possibility to have a full development environment with necessary underlying data center power. But, be cautious when planning to deploy your solution in the PaaS solution. You need to understand the total costs and the possibility to move your logic out of the PaaS solution (code/logic portability).
There are two general “truths” on SaaS:
- The more independent the SaaS is from your in-house applications, the easier to achieve fast effects.
- Massively scalable applications are excellent candidates to move from your datacenter to the cloud (e-mail as the best example).
Now, to the second part: What´s in it for me?
TCO (Total Cost Of Ownership) today, tomorrow and in the future
To be able to make any kind of informed decision you first need to understand your TCO per application/workload that you currently have. You probably have an economic model in place, but if that isn´t the case, you need to understand the cost associated with each application/workload. In the model you need to map costs for server, storage, networking, communications, disaster recovery, datacenter costs, platform costs, packaging, provisioning, AAA processes, maintenance, help-desk, support, end user education, management and much more. By adding all costs per application/workload, you now have a pretty good TCO figure annually. Now, you need to do the math per application/workload that you see as a cloud candidate. The effects that you need to understand are sometimes clear and sometimes blurry.
Here are only three areas you need to consider, beside the financial model (and there are at least a dozen more):
- Server utilization: A specific application that is easy to virtualize is probably more expensive to run in the cloud.
- Plattform considerations: Have you entered SW Maintenance Agreements? Are those agreements per server, processor, user or a combination of other factors? Do you have some kind of transaction engine, message bus or middleware solution in place to support your applications platforms? How tied are those?
- Storage costs: An application that is light on network based storage resources is probably harder to find an economic benefit for in the cloud.
I have provided some brief examples to show you that a cloud transition is a thought, planning and strategy process. When you embark on the process, you must understand that during some transitional phases, you will actually end up having a higher TCO. But, I guess that this is the sheer beauty of the clouds?
- 15 Cloud Data Performance Monitoring Companies - June 21, 2016
- Cross-Site Scripting – Why Is It A Serious Security Threat For Big Data Applications? - June 21, 2016
- What’s Ahead For Predictive Analytics And Why It Matters For Businesses - June 17, 2016
- Do Not Rely On Passwords To Protect Your Online Information - June 15, 2016
- Despite Record Breaches, Secure Third Party Access Still Not An IT Priority - June 14, 2016