By Louise Dilley, Regional Services Director, Virtual Instruments
Cloud migration is a strategic necessity in Enterprise organizations – but do those of us who support and innovate in the on-premise space have the tools and skills that a smooth transition from on-premise to cloud demands?
Gartner’s infrastructure team MVP, David J. Cappuccio, recently stated that his planning assumption is that by 2025, 80% of enterprises will have shut down their traditional data center, versus 10% today.
The first wave of low-criticality application migration to public cloud was relatively easy, as performance is not the No 1 priority for these applications. The next wave will be much more difficult – migrating top-tier applications that are critical to the success of the business, are highly latency sensitive and put competitive advantage at risk if performance isn’t maintained.
Infrastructure as a Service (IaaS) needs thorough and expert planning because it should ensure application performance, rather than just availability. Lifting and shifting existing applications to new infrastructure of the same spec doesn’t take into account changing workload behavior and often results in disappointing performance and, potentially, spiraling costs. The primary driver of failure is lack of planning.
If you simply decide to replicate your on-premise environment at a hosted site then the costs and performance profile will be similar to today. But, if you are seeking to host it in the public cloud, new questions arise. The planning process must include the question ‘How much will it cost to run my application in the cloud and will it meet our performance SLAs?’
And to know what service you expect from the cloud provider you need to start with knowing how your applications are performing now – this is a very different question to ‘how are my servers/fabric/storage/ performing now?’
The other fundamental is ‘Which Cloud Service Provider is the best choice for my application workload?’ Very few cloud providers even come close to offering a Service Level Agreement (SLA) for the performance of your application in their environment – yet this is vital to your decision process. Multiple factors will affect the performance of specific applications in the cloud. Lack of understanding of these factors and specific workload characteristics increases the risk of business impacting failure post migration.
The starting point, therefore, is to analyze the workload behavior and performance characteristics of every application you plan to migrate. In addition to individual application performance you need to understand dependencies between the application and other applications and infrastructure components. If the application is accessing other resources that are not going to be hosted in the cloud, or are managed by a different provider, this can drastically affect performance and cost.
Each cloud provider has strengths and weaknesses when it comes to hosting your application. Critical workloads support your core business so it makes sense to avoid any risk of service deterioration in the cloud. Cloud providers will make performance claims but how do you ensure these are trustworthy? Ideally each top-tier application workload needs to be tested against the new environment to ensure you are getting equal or better service than you have on-premise.
In summary, don’t rush to the cloud without answering these three questions:
Virtual Instruments has over ten years’ experience in workload performance analysis, migration and service assurance. Its Cloud Migration Readiness (CMR) service enables validated and seamless migration to the cloud, delivering assurance to your business that their critical applications will maintain or even improve performance and service in the cloud.
Click here to learn more about the CMR service, and drop us a line to see how Virtual Instruments can help maximize the ROI for your cloud migration projects. And as always, don’t forget to follow us on Twitter, LinkedIn and Facebook to stay up-to-date on the latest and greatest in app-centric infrastructure performance monitoring.