Home Invite QR code

Smart Scaler

What is Smart Scaler? 

Smart Scaler is an innovative solution designed to dynamically scale pods within cloud-native environments based on traffic predictions and microservice pod capacity estimations. Built on predictive analytics and Reinforcement Learning (RL), it revolutionizes the management of resources, enabling organizations to efficiently meet fluctuating demands while minimizing waste and operational costs.


Why you should implement Smart Scaler?

Escalating cloud costs is a challenge for organizations. This is because of lack of predictive capabilities in current autoscaling solutions to manage unplanned demand.

The best way to address these challenges is to predict demand ahead of time and 'precisely' scale up or down, infrastructure and application resources.

Core Features

49% containers use less than 30% requested CPU and memory
69% of requested CPU cores and 18% of requested memory were unused


Product Link: Smart Scaler

Smart Scaler Implements Lights On/Lights On with Intelligent and Predictive autoscaling powered by Generative AI.


Resources:


Whitepaper: Revolutionizing Pod Scaling and Resource Efficiency


Slides: Introduction to Smart Scaler


Use Cases 


Battle Card: AutoScaler Comparison


Product Sheet


Got questions? Read the FAQ.

Frequently Asked Questions

Compare cost of running the app workload with and without AI/R based optimization for a period covering 2-3cycles of traffic variations. (eg. if traffic varies weekly, time period is 2-3weeks)
Any app that has variations in traffic patterns (daily, weekly, monthly etc.)
2-3 cycles of the traffic pattern variations are needed.
Yes; but, by default, the model changes are disabled in production. Aerts provided by Smart Scaler indicate when the model needs to be updated and the user can choose to update the model at that time.
With Smart Scaler, we create models for a desired combination of SLA and/or cost. Typically, target for SLAs is 1% failure rate.
The time series training data is extracted from Prometheus on the application cluster. Dataset includes System [cpu, memory utilization] and Application data (rps, failure rates, pod counts, latency).
Smart Scaler is agnostic to the managed K8s provider orchestration engine.


About Avesha 


Avesha is committed to revolutionizing the way companies consume and manage their cloud application infrastructure, by offering a comprehensive suite of tools that enable secure application networking and improved application performance. Our products come together to solve deployment, performance, and cost challenges for multi-cluster, hybrid, multi-cloud, and edge. Our mission is to enable fully autonomous infrastructure that performs optimally to meet the unique needs of distributed applications. Check us out at http://avesha.io/ 



KubeSlice
Smart Scaler


Let's talk