Ely Ronen, Solution Architect, Cloudport.
Experienced with Cloud cost reduction, Cloud Infrastructure and AWS services
Introduction
For organizations, delivering high performance usually comes with big cluster sizes, high 3rd party costs, and excess management needs, just to get some needed code changes effected. All that effort just to provide the needed throughput and maintenance of a proper response time makes the process. However, most organizations will try to optimize their performance and leverage on cost balance. That is, they will use specific instance types, aggressive and dynamic scaling policies, leveraging Spot instances, etc.
As we have mentioned in previous blog posts, here at Cloudport we always thrive to find the next solution that will dramatically reduce your costs while improving your workload performance, which is the reason we offer you Granulate, which is a new approach to performance optimization.
Granulate provides a real-time continuous performance optimization that allows you to handle compute workloads with 60% fewer servers while improving performance by 40%, with no code changes.
Yes, you read it right, no need for code changes!
Granulate is past initiative agnostic, as there is no other solution that is optimizing on the kernel level. It’s entirely driven by AI and helps businesses overcome complex optimization challenges.
Granulate for Cost Reduction
Concerning cost reduction, Granulate results will always be an add-on for any other cost reduction initiative you may have already applied to your infrastructure. For example, an organization will use Granulate to conduct to work on dynamic scaling metrics, fine-tuning solutions, instance lifecycle (e.g. spot instances usage), etc.
As most of you already have an auto-scaling policy configured, a reduction in average CPU utilization for every instance will be immediately and automatically translated to cost reduction as your cluster scales fewer machines at any given moment.
For Java/scala/python (and many more - as can be seen in the photo below) applications, Granulate is a game-changer at performance vs. cost balance as their technology allows you to shove twice more workload on the same amount of machines.
Using Granulates Agent (GAgent) Granulate’s AI continuously learns your application-specific resource usage, Data flow patterns in each one of your servers and regulates the OS-level scheduling to improve performance and enable cost reduction while no PII Data or user data is being collected.
GAgent installation is extremely simple and can be done through any infrastructure provisioning tool as well
The installation process of the GAgent is extremely simple and quick (estimated time of 15 minutes). However, for the AI to autonomously study your application resource requirements and data patterns, we suggest providing at least a week of evaluation to maximize the benefits you will gain from using Granulate’s AI.
After a week of evaluations, deployment of the agent across your environment will take an additional 15 minutes. From that moment onwards, the agent will automatically enhance and adjust according to changes in your application.
Besides the great GAgnet, Granulate recently released their newest product - Granulate Profiler (G-Profiler)
The G-profiler is an open-source, plug-and-play tool with immediate visibility, wide coverage, and minimal performance overhead. The G-profiler will help you identify performance bottlenecks by analyzing code performance continuously on your entire environment. It will also help you optimize your resource-consuming code, and improve performance while lowering your total cost of ownership.
Looks great! How do I get started?
All you have to do is to submit your details here and get access to a FREE trial with Granulate.
For more information, check out these:
For further information about Granulate autonomous cloud application-driven optimizer:
Further information about Granulate Profiler:
Some Relevant Granulate Blog posts:
To learn more about Compute connect with us.