StormForge

Tired of your JVM being a memory hog? Rightsizing JVM workloads on Kubernetes requires knowing what the appropriate heap size should be and then sizing the container around that heap size. By analyzing Java metrics using machine learning, we generate tailored recommendations for heap size, as well as Kubernetes resource requests and limits. These recommendations can be reviewed and applied through a user-friendly interface, or they can be deployed automatically, for continuous performance optimization. Drop us a comment or message about accessing our limited availability Java Workload Optimization.

"Just set your requests and limits" - Obi Wan Kubernetes "If simple it was, accurate they would always be." - YAMLoda Whether you're a Kubernetes Padawan or Master, setting accurate requests and limits is not magic and cannot be done with a wave of a hand. StormForge is as close to magic as you'll get. 🕵 Deploy an agent 📈 StormForge will discover and analyze your workload usage patterns 🤖 Machine learning will generate predictive request and limit values 📄 Policy-based automation will apply those recommendations based on your operational patterns 🎉 All workloads are rightsized and if you're running #karpenter or #eksautomode, your cluster capacity is automatically managed to match workload resource requirements. If you're an Amazon Web Services (AWS) customer, you can try StormForge for free for 45-days, no limits, by clicking the *Try for free* button on #AWSMarketplace - https://lnkd.in/gNDkWPEn

We are excited to share that we are Amazon EKS Ready! Achieving this designation confirms our expertise providing a world-class solution for customers to rightsize workloads on Amazon Elastic Kubernetes Service (Amazon EKS). This means that Optimize Live was reviewed by AWS Solution Architects and tested using the Amazon EKS Anywhere (EKS-A) Conformance and Validation Framework. It ensures architectural soundness, adherence to AWS best practices, and compatibility with EKS deployment models. Read the press release for more info ➡️ http://bit.ly/3QcpvOe #EKSready #amazonwebservices #amazonEKS AWS Partners #kubernetes

To reach peak cost-efficiency using Amazon EKS Auto Mode, you will need to carefully configure resource requests for each of the workloads deployed onto your cluster. Accurate workload-level resource requests are critical to provisioning the optimal EC2 instance types in order to keep your costs under control and applications at peak reliability. To get started maximizing, you just need to deploy the StormForge Agent onto your EKS Auto mode cluster (it's free). In this video, you can see how to use the AWS Marketplace Add-on to deploy the Agent. #EKSautomode #EKS #AWS #Kubernetes Amazon Web Services (AWS) AWS Partners

We're working on a lot of educational resources around Amazon EKS Auto Mode, and we're curious to know how much people know about it. Take a second to respond to our poll to help us determine what level of information would be most beneficial. 👇

We've seen a lot of interest from the Kubernetes community about this topic, so we've dedicated this edition of the Kubernetes Optimization Digest to it. You'll learn: ✴️ How to use and apply the kubectl command with the scale subcommand to manually execute horizontal scaling. ✴️ How and when to use kubectl to scale deployment to 0, and the limitations of this action. ✴️ How Kubernetes Event-Driven Autoscaling (KEDA) can fill in the gaps left by the Kubernetes Horizontal Pod Autoscaler (HPA). 

As an Amazon Web Services (AWS) EKS-Ready partner, we are excited to see how EKS Auto Mode will redefine expectations for cost and operational efficiency on #kubernetes. But once Auto Mode is enabled on your clusters, what comes next? That must be all, right? Just sit back, relax, and...look for tickets in your Jira backlog to grab? Actually, once Auto Mode is enabled, you'll likely see an even greater need for automated workload resource rightsizing. Those workload requests are what EKS Auto Mode's intelligent capacity management feature (#karpenter) uses to determine what size and shape EC2 instance to provision for your workloads. Under or over-provisioned workloads mean that even though your cluster compute capacity is optimized, you can still experience wasted resources and/or application performance risks. If you're interested in a brief primer explaining why workload rightsizing is so important once Auto Mode is enabled, you're in luck! We even include some step-by-step instructions for deploying the StormForge agent using EKS Add-ons to keep everything under the #AmazonEKS operational umbrella! https://lnkd.in/gPawYVce #k8s #platformengineering #automode #resourceoptimization #sre #costoptimization

This is a neat little peek behind the curtain here at StormForge by our Director of Product, Nicholas Walker. As a SaaS platform - our resource usage is a direct reflection of the way our customers use our software. When our customers optimize more workloads or generate more frequent rightsizing values - of course our services must scale to meet demand while still meeting our SLAs. But the cookie cutter approach doesn't work for us. We cannot sit by while some tenants are overprovisioned and others are underprovisioned. The operational and reputational risks would be astounding. While we benefit from autoscaling technologies like the HPA, KEDA, and Karpenter/EKS Auto Mode - there is still a significant challenge to rightsizing the microservices that support each individual customer tenant. We have to account for every customers' scale and desired operational patterns. Foisting this responsibility onto our own platform engineering team would be detrimental to our innovation initiatives. So it's a good thing we're using StormForge Optimize Live to analyze the application resource patterns for each of our customers - ensuring that each customer "fingerprint" is understood - their seasonality (if present) is accounted for - and that every one of our microservices are continuously optimized without human intervention. At StormForge, innovation doesn't wait, especially for pod rightsizing:) #platformengineering #kubernetes #k8s #resourceoptimization #finops #devops #sre

✨ NOSSO PATROCINADOR GOLD ✨ 🎉 Estamos extremamente contentes em anunciar nosso próximo parceiro, StormForge! Com uma plataforma para gerenciamento e otimização de recursos em Kubernetes, a StormForge vem inovando as iniciativas de FinOps e de eficiência em ambientes Cloud Native. Usando Machine Learning, dimensione automaticamente contêineres superprovisionados com escalonamento automático vertical + horizontal, aumentando a densidade de pods e nodes, liberando capacidade do cluster e economizando de 50% a 70% nos custos de nuvem! Muito obrigado por estarem conosco nessa jornada, StormForge! Cloud Native Computing Foundation (CNCF)

I read a few articles about #kubernetes OOM kills today and they bother me a little bit because they describe what an OOM kill is and what causes them but then they go in to the solution and the solution is always “set your requests and limits correctly.” Which is fine and true that we should all set them correctly but jeez it’s quite a bit of work to review the usage metrics and set your requests/limits for every single one of your 100s of apps running in #k8s right? We don’t need a bunch of blog posts and articles telling us that we must set requests/limits correctly what we need is something to do it for us. If you agree, then reach out and let’s talk. #platformengineering #devops #aks #aws #eks #gke #sre