Organisations have been increasingly adopting and migrating workloads to cloud-based technologies, mainly SaaS and IaaS. It is estimated that adoption of, and expenditure on SaaS and IaaS alone will experience growth rates of between 100% and 200% within the next two years. Migration of applications and services to the cloud is very appealing and seemingly easy to justify. One of the main drivers for this surge in investment in cloud computing services is the perceived lower-cost benefit that comes with it, which is a catch!
Most organisations on the “cloud journey” soon experience the Cloud Shock phenomenon, whereby they are spending more than they budgeted for. In some worst-case scenarios, this has resulted in some business entities totally abandoning their adoption of cloud-based services and returning to on-premises data centre services.
The utilisation of cloud computing services can be more costly than initially intended if not contained and not well managed. Thus, it is essential that organisations get an understanding of how they can reduce, optimise and manage the costs of using cloud services.
What can cause cloud over-expenditure?
- Oversizing, because of a lack of capacity planning and management
Just like in traditional compute environments, an organisation can throw infrastructure and resources to its capacity problems, instead of first embarking on a cost-effective capacity-planning process and rightsizing the environment before, during and after migration to the cloud.
Without capacity planning, the enterprise finds itself having one or more of the following:
- Increment of capacity resources way before thresholds are reached, just to be on the safe side.
- Having no accurate sizing of workloads before they migrated to the cloud. This is most likely as a result of the absence of effective testing strategies, which are divorced from capacity planning.
- Business and compute demand forecasting is thumbsucked or is non-existent.
- Resources are increased even without understanding or identifying the source of the performance bottlenecks; for example, temporarily adding capacity resources to mask badly performing applications.
- Badly written application or software code
- Non-optimised infrastructure
Often IT departments do a lift and shift, without optimising the infrastructure because of time constraints. Capacity issues, such as the existence of oversized servers with wasted CPUs and storage migrated across to the cloud, or the running of non-critical workloads on expensive cloud services, are a result of such bad practices.
- Existence of legacy systems that are not cloud-ready (inelastic)
How to avoid and manage the Cloud Shock
Across most boardrooms, CIOs, CFOs and IT managers are increasingly receiving shocking monthly bills related to cloud services spend. As soon as the shocking bill or over-expenditure becomes a reality, it is these key business decision-makers who are brought to account. What can one do to avoid the pitfalls of the Cloud Shock?
Monitor and measure utilisation
In order to identify wastage, monitor and measure utilisation and costs using cloud-monitoring tools such as Microsoft Azure Monitor or Cloudability.
Azure Monitor, courtesy of www.microsoft.com
These tools enable visibility of use, costs and performance all on one portal, are easy to access and allow for better decision making, no matter how large your enterprise is. MentPro can guide you in the adoption and usage of the right tools to automate the monitoring and gathering of cloud services analytical data for better decision-making.
Select candidate areas for optimisation for cost-saving opportunities
An optimised cloud environment means you do not waste financial resources on unnecessary resources. You pay for what you need; nothing less, nothing more. Only when this has been achieved will you be able to truly realise the cost benefits of cloud migration.
Plan and optimise
You should plan for optimising your cloud environment before and after migration into the cloud. Optimisation involves rightsizing, infrastructure and application optimisation, application performance improvements, and improving elastic application footprints.
Rightsizing involves the many aspects of capacity-planning metrics and dimensions: CPU, Memory, Disc, Network and IOPS. It corrects (“rights”) the wrongs of over-provisioning or under-provisioning by making sure that the resources are efficiently utilised and that you only spend on what you need. However, increasing or reducing allocations, subscriptions etc cannot be done in isolation, as the exercise also requires a business risk and impact analysis and an understanding of the business demand and your workload characteristics. Armed with the right tools and a strategic business discovery process, rightsizing can assist in fast decision-making and recommendations in order to realise savings, and lower and optimise your cloud costs before and after migrating to the cloud.
Rightsizing for the cloud – the right way
MentPro can greatly assist your organisation with a “cloud rightsizing done right” exercise through many years of capacity planning and management expertise and experience, historical data and benchmarks.
Infrastructure and application resource optimisation
In order to optimise infrastructure and applications, one must fully understand the business demand in the environment through business discovery, and translate this into computing resource requirements. This, through capacity modelling to ascertain accurate demand quantities.
Application performance improvements
There are several powerful Application Performance Monitoring (APM) tools that can greatly assist in ascertaining application performance and help to identify bottlenecks and badly written code, which results in inefficiencies. Inefficient application systems can have a big bearing on the cloud costs.
Increase footprint of elastic applications
Elastic applications can adapt easily to autoscaling, which means you can also enable autoscaling in your environment. However, some applications do not easily support autoscaling, and so it presents a problem. You cannot take full advantage of the cloud’s elasticity benefits when you have such applications. The extent to which an application is elastic determines how fit it is as a candidate for migration to the cloud. The more elastic applications you have, the more you can realise savings and control costs through the autoscaling of resources such as databases and caches for cloud services.
In conclusion, cloud computing adoption promises a number of benefits, with the main driver being lower costs and realisation of huge savings. However, if there is no proper planning and strategy to control, reduce, optimise and manage the utilisation of cloud services and resources, you can easily overspend without realising it.
You need right cloud capacity and monitoring tools, capacity planning and performance processes for visibility and continuous improvements, so as to take advantage of the beneficial characteristics of the cloud model before or after your migration. MentPro can assist you with cloud cost optimisation, capacity planning and cloud migration plans.