In its simplest terms, Turbonomic is a tool that provides "Application Resource Management". With its powerful artificial intelligence technology in its infrastructure, it can manage resources in an application-based and dynamic manner. It analyzes the data provided and lists the actions that need to be taken as a result. If we wish, we can take these actions manually or leave them to Turbonomic through an automated process.
Along with this, optimizing application resource consumption in data centers or Cloud environments plays a significant role in improving a company's long-term energy consumption profile. Turbonomic helps us use these resources efficiently without compromising performance.
It collects data from various environments such as virtualization, Kubernetes, network, disk, to which the application is connected, through APIs. It combines the data collected within these technologies on a single platform, analyzes it with machine learning, and detects the root cause analysis of a problem. In environments where Turbonomic is not used, analyses may involve a lengthy and intense process as a result of the examination by many different teams. Turbonomic, on the other hand, is able to make these analyses and decisions continuously and from many perspectives.
Turbonomic fundamentally provides two separate use case scenarios. The first of these use case scenarios is the ability to perform dynamic resource management by taking actions on application resources. The second is; when looking at the Cloud perspective, it is to provide optimization and cost savings. The details of these two use cases are given below:
A) Application Resource Management
Turbonomic is able to perform resource management on an application basis. Before diving into the details of this, we can examine a complex application stack in the image below:
As seen, when we examine the layers of the application stack, there are many layers ranging from the physical layer to the application layer. Problems related to performance can occur at any of these layers. Therefore, a bottleneck at any layer can directly affect application performance. Since each layer can be managed by a different team, inter-team cooperation may not be fast and easy. Turbonomic, however, analyzes these data on a single platform and produces a solution to this problem.
B) Cloud Resource Savings
On cloud platforms, when we need any resources, we can scale these resources and it is quite important for us to use the flexibility provided by this scalability efficiently. So, how efficiently are we using the resources we use on cloud platforms? Are we using unnecessary resources? Will there be a decrease in application performance if we use a lower-resource virtual machine? In cloud environments like Amazon Web Services (AWS), Microsoft Azure Cloud, Google Cloud Platform (GCP), Turbonomic analyzes these and similar questions for us and provides a list of actions to be taken.
Besides, On-prem provides solutions on issues such as virtual servers used, disks, their performance, whether there is a resource problem in meeting the requests from services.
Continuous Optimization with Turbonomic
One of the working principles of Turbonomic is to be able to perform API-Driven discovery. It communicates using API services from APMs to Kubernetes, from Storages to Multicloud environments. It combines this data it gets by doing the target integration into a single data set. Turbonomic automatically extracts the application and infrastructure map from this data set.
Business Transaction, services, application components, containers, container specs, database servers, disks, volumes, etc. It presents the whole structure to us. As can be seen, it presents the risks both by coloring and numerically. Turbonomic makes this coloring through "2 main risks". These risks are divided into critical and minor risks. Critical risk shows situations that require action at the highest level and directly affect performance, minor risks show risks that do not directly affect performance but can affect capacity.
It manages performance with optimum scaling by dynamically and continuously meeting resource needs. It can do these scaling operations in cases of instant resource demands, instant CPU and memory increases in containers or virtual machines. It can also take actions such as starting/stopping services automatically.
Turbonomic lists the actions to be taken in the "actions" part. If we want, we can take these listed actions manually or we can ensure that actions are taken automatically within the time frame we determine. We can include it by creating a pipeline and workflow or we can ensure that these actions are taken in real time.
If you wish, you can reach the Turbonomic interface from "https://try.turbonomic.io" and examine it.
When we log into the Turbonomic interface, widgets such as expected actions, Top Business Applications and Transactions, Top Services, and risk avoided are presented. If we wish, it also allows us to customize the dashboard according to our needs. Actions are divided into categories such as compliance, efficiency, prevention, performance, savings, transactions.
In the image above, five possible scenarios and recommended actions to be taken are given. In the first two scenarios, critical risks directly affecting performance related to memory and CPU are seen. If we want to look at the details, we can find answers to questions like how long the CPU usage increased and at what time frame it happened or where was the peak point. In the third and fourth scenarios, instance types and their hourly rates are given in virtual machines and databases. When we switch to the recommended instance, the amount of savings is presented comparatively. In the last scenario, actions related to disks are included. It also presents us how much savings we can make with the change in IOPS values among disks with the same capacity by comparing.
Turbonomic Sustainability Calculator
As is known, the significant increase in carbon emissions has become a major threat to our planet. Now, countries and communities of nations aim to combat this threat with long-term and short-term environmental action plans. Efficient utilization of data centers will play a substantial role in achieving the targets set in these environmental plans.
It has been calculated that data centers globally consume approximately 200 terawatt-hours (TWh) of electricity each year. In other words, if all the data centers in the world were a country, with this statistic, it would rank 23rd on the list of countries with the highest electricity consumption. This is equivalent to about 40% of Germany's electricity consumption. In this case, if you want to see your consumption and the associated carbon emissions, you can use the Turbonomic Sustainability Calculator (https://www.turbonomic.com/sustainability-calculator).
The “Turbonomic Sustainability Calculator” allows you to measure the energy consumed in your Data Center and/or Cloud environments by providing more tangible comparisons; showing equivalents such as cars, electricity, barrels of oil, smartphone charges, etc. By using Turbonomic to use your environments more efficiently, we can contribute to the ecosystem and leave a greener environment for future generations.