VP Marketing & Product (UX)
11 April, 2023
3 min read
"You know what the happiest animal on Earth is? It's a goldfish. You know why? It's got a 10-second memory. Be a goldfish, Sam," said TV’s famous football coach Ted Lasso to Sam Obisanya, a player in his team.
And isn't that the dream for those managing cloud infrastructure? To have tools and software that can be truly autonomous, requiring no further attention or memory once they've been configured. This is Avesha's vision and journey - building a framework for an autonomous infrastructure - and I'm excited to be part of it.
Let's see how Avesha is working towards its vision.
The first product, KubeSlice, simplifies the work of platform teams in achieving true multi-cluster, multi-region application deployments. It achieves this by creating a unified, secure application network between all clusters - we call it a Slice. No matter where the physical clusters are located, the Slice binds all applications together seamlessly. The Slice virtual cluster transcends geographic boundaries and incorporates all characteristics of a Kubernetes cluster, enabling easy configurations for service discovery, resource quotas, namespace isolation, QoS bandwidth limits, and other crucial elements, with simple YAML files.
A neat little tool in Avesha’s arsenal is the KubeSlice Manager UI. Pick your clouds, pick your clusters, pick your namespaces like toppings on an ice cream. Then add them to a Slice, and forget the rest, just like a goldfish! The KubeSlice UI generates the YAML files, reducing the complexity of manually building them and eliminating the potential for fat-finger errors. The Slice is GitOps friendly and the generated YAMLs can be driven through your CI/CD pipeline. As a result, the Slice continuously honors your YAML definitions, enforcing them and avoiding config drift.
It's one type of autonomous infrastructure or what Gartner terms "hyper-automation".
Next up in Avesha's autonomous infrastructure toolkit is SmartScaler. This AI-powered tool works autonomously to make your HPA (Horizontal Pod Autoscaling) work at the most optimal levels, guaranteeing the SLA for applications and keeping your cloud costs in check. The Smart Scaler AI model is trained on your historical application and infrastructure metrics and when applied to your deployments, the Reinforcement Learning uses learned traffic patterns to scale pods up or down to meet the desired SLA state. DevOps can once again sit back, be a goldfish and focus on other tasks without having to guess and pray that the autoscaler threshold settings were correct.
That’s another type of autonomous infrastructure.
Avesha's KubeSlice and SmartScaler are paving the way for a new era of autonomous infrastructure management. These tools automate a lot of the work platform teams have to do and leverage intelligent systems to proactively manage infrastructure based on the unique needs of each application. With these tools at hand, businesses can achieve high-performing application deployments and free up their teams to tackle other responsibilities.
March towards autonomy. Be a goldfish. Watch Ted Lasso in the time that's freed up. What else is happiness?