Home Docs Demos Blogs
GitHub Slack
Explore Kruize demos and see optimization in action!

Interactive Demos

Experience Kruize's powerful optimization capabilities through interactive demos, tutorials, and real-world scenarios

Kruize UI Sandbox

Watch a walkthrough of the Kruize UI showing how to create experiments, view recommendations, and optimize your Kubernetes resources

Real-World Optimization Scenarios

🔵 Idle Workload Detection

Problem: Container running but consuming minimal resources

Solution: Kruize identifies idle workloads and suggests significant downscaling

↓90%

CPU Request

↓85%

Memory Request

🟡 Over-Provisioned Resources

Problem: Resources allocated far exceed actual usage

Solution: Right-size recommendations to reduce waste and costs

↓60%

CPU Savings

↓45%

Memory Savings

🔴 Under-Provisioned Resources

Problem: Containers hitting resource limits causing performance issues

Solution: Increase allocations to prevent throttling and OOM kills

↑40%

CPU Increase

↑35%

Memory Increase

🎮 GPU Optimization

Problem: Inefficient GPU utilization with expensive accelerators

Solution: MIG partitioning recommendations for Nvidia GPUs

3x

GPU Density

↓65%

GPU Costs

📦 Namespace Recommendations

Problem: Managing resource quotas across multiple containers

Solution: Namespace-level CPU limits and memory quotas

Multi

Container Support

Auto

Quota Sync

📊 Box Plot Analysis

Problem: Understanding usage patterns and variance

Solution: Statistical visualization of resource consumption

P95

Percentile Analysis

24/7

Monitoring

⚙️ Runtime Recommendations

Problem: Optimizing Java runtime parameters for better performance and efficiency

Solution: Kruize analyzes runtime metrics and suggests optimal JVM settings, heap sizes, and GC configurations

Java

JVM Tuning

Auto

GC Optimization

Video Tutorials

🎬 Getting Started with Kruize

Learn how to set up Kruize and create your first experiment in under 5 minutes

🎥 Kruize UI Walkthrough

Explore the Kruize UI interface and see how to navigate experiments, view recommendations, and analyze resource optimization

Quick Start Guide

🔧 Kruize Operator Demo

Automate Kruize deployment using the Kubernetes Operator for declarative resource optimization:

# Clone the operator repository git clone https://github.com/kruize/kruize-operator.git cd kruize-operator # Install CRDs make install # Deploy on KIND make deploy-kind IMG=<registry>/kruize-operator:tag # Create a Kruize instance kubectl apply -f config/samples/v1alpha1_kruize.yaml -n monitoring

What it does: Uses Custom Resource Definitions (CRDs) to automate Kruize Autotune deployment and provide resource optimization recommendations for Kubernetes workloads

Supports: Kubernetes v1.23.0+, OpenShift v4.10+

⚡ Kruize Optimizer Demo

Lightweight Quarkus-based microservice with REST APIs for automated configuration and optimization:

# Clone the optimizer repository (mvp_demo branch) git clone -b mvp_demo https://github.com/kruize/kruize-optimizer.git cd kruize-optimizer # Run in development mode (enables live coding) ./mvnw quarkus:dev # Or build and run as JAR ./mvnw package java -jar target/quarkus-app/quarkus-run.jar # Build native executable in container (no GraalVM required) ./mvnw package -Dnative -Dquarkus.native.container-build=true ./target/kruize-optimizer-1.0.0-SNAPSHOT-runner

What it does: Tag It | Forget It | Optimize It - Automates experiment creation, profile validation, and health orchestration for hands-off optimization once configured

Features: Quarkus-based microservice, REST APIs, Kubernetes Client integration, Prometheus monitoring, Scheduler support