ITOA Will Optimize Workload Patterns for Public Cloud Applications
Businesses move to the public cloud to avoid upfront capital expenditure, eliminate the expense associated with administering physical servers, and to scale up and down according to their workload requirements. However, the vast majority of businesses are significantly over-provisioned in their use of public cloud resources, resulting in a high monthly expense that undermines the original business case. The reasons for being over-provisioned are two-fold: Firstly, some applications are not architected to take advantage of horizontally-scalable micro-services that lend themselves to auto-scaling; the second reason stems from under-estimating the complexity of measuring and analyzing the multiple dimensions of capacity utilization (CPU, memory and I/O by resource type) and variations in pricing contracts (reserved, on-demand, spot instances.)
The public cloud platforms such as Amazon Web Services (AWS) have the APIs necessary to allow analytics service providers help right-size public cloud resources with confidence. A new generation of IT Operations Analytics (ITOA) will optimize contract types against workload patterns, and detect the need for increase or decrease capacity in real-time based on performance analysis.