Densify leverages patented, predictive analytics that perform advanced modeling of workload patterns to give precise, automatable results - not naive suggestions. This means we can truly align supply and demand, and perform sophisticated optimization scenarios for your cloud environments that others simply can’t.
This analysis starts with a learning step, where historical data is used to establish a baseline operational envelope for cloud workloads. Using historical workload data is very important, as cloud optimization recommendations must account for cyclical loads such as month end and quarter end peaks, as well as time-of-day patterns, such as start-of-day peaks or overnight batch jobs. Many cloud management products focus on the bill, not the actual workloads, and even if they do they work with averages that obscure the detail and result in the wrong answers being given. And that means more cost and risk for your cloud workloads.
Detailed Amazon EC2 optimization recommendations, highlighting cases where moving to a different instance class would save a lot of money
A critical part of workload pattern is normalizing the data using benchmarks, so the actual work being done can be analyzed against the service offerings of the cloud providers. Understanding the detailed configurations and performance characteristics of the hardware underlying the cloud offerings is also critical to this, and makes the difference between vague recommendations, like “consider making that instance smaller”, to specific, actionable optimization actions, like “make that c3.xlarge instance a c4.large”
The ability to normalize workload data between cloud offerings and instance classes is key to modernizing and just simply choosing the right class of instance for your workloads. And this is can be big driver of cost savings. Analysis shows that making adjustments within the same instance class typically yields a 15-20% reduction, but being able to precisely analyze across classes gives 35-40% savings. You cannot do this unless you normalize the data using benchmarks and other products cannot do this.
Optimization is not a one-time thing. Cloud environments are constantly changing and growing, and what is optimal for today may not be optimal for tomorrow. Densify provides the ability to look into the future and model what is going to happen, both by analyzing the trends in the data and modeling application growth, deployments and other planned changes. This means you can stay ahead of the game and keep your applications running smoothly. And if you use RIs it means you don’t lock into a set of services that can’t take you where you need to go.
Densify also enables pattern-based workload stacking, enabling you to take advantage of bare metal cloud offerings and container-based hosting models. This ability to "play Tetris™" with workloads and dove-tail workload patterns means that you can sweat your cloud assets while reducing resource contention, and can only be done by Densify.