skills.rest
decocms
decocmsOFFICIAL
Explore all skills by this author →
Website𝕏@deco_cmsGitHub

deco-site-scaling-tuning

Analyzes Prometheus metrics to identify CPU inflection points and recommend scaling targets for Deco sites.

5|2|Updated Feb 26, 2026
Search Tags:#performance#metrics#kubernetes#latency#prometheus#autoscaling
One-click install
npx skills add https://github.com/decocms/deco-start --skill deco-site-scaling-tuning
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: deco-site-scaling-tuning
Source: https://github.com/decocms/deco-start/tree/main/.cursor/skills/deco-site-scaling-tuning
Command: npx skills add https://github.com/decocms/deco-start --skill deco-site-scaling-tuning

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates the discovery of optimal autoscaling parameters for Deco sites by analyzing Prometheus metrics to identify the CPU inflection point where latency begins to degrade and to decide between CPU-based and concurrency-based scaling strategies.

Core Features & Use Cases

  • Prometheus-driven analysis to locate CPU inflection points and guide scaling targets
  • Guidance on choosing between CPU-based and concurrency-based scaling for different workloads
  • End-to-end workflow from data collection to applying recommended scaling configuration

Quick Start

Run the 12-hour Prometheus data collection and analysis workflow to locate the latency-inflection CPU target and apply the recommended scaling configuration.

Frequently Asked Questions about deco-site-scaling-tuning

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find the optimal Kubernetes autoscaling target for my site using Prometheus metrics?▼

To find the optimal autoscaling target, analyze Prometheus metrics like CPU usage and concurrency to identify the CPU inflection point where latency degrades, which determines your recommended scaling value.

When should I use CPU-based autoscaling versus concurrency-based scaling?▼

Choose CPU-based versus concurrency-based scaling by analyzing Prometheus metrics to see which workload pattern correlates better with latency degradation, guiding your selection for common deployment configurations.

How do I configure Deco site autoscaling based on latency inflection points?▼

Configure Deco site autoscaling by running a 12-hour Prometheus data collection workflow to locate the latency-inflection CPU target, which outputs an installation-ready state snippet for your site configuration.

Do I need Prometheus access to analyze CPU usage and pod counts for autoscaling?▼

Yes, you need Prometheus access along with a target namespace and site context to analyze CPU usage, concurrency, pod counts, and latency data for autoscaling recommendations.

What metrics are required to identify the CPU inflection point for Kubernetes pod scaling?▼

Identifying the CPU inflection point requires metrics covering CPU usage, concurrency, pod counts, and latency where available, analyzed together to locate where latency begins to degrade.

Related Skills

Explore More Data & Analytics Skills→
redhat-etredhat-et
1

check-metrics

Query Prometheus metrics with PromQL to diagnose resource usage and platform performance.

Official
Intermediate
raphaelmansuyraphaelmansuy
2

knative-serving

Deploy AI agents as Knative Services with autoscaling and traffic splitting.

Community
Advanced
pluginagentmarketplacepluginagentmarketplace
1

model-serving

Deploy ML models for production inference with BentoML and Triton.

Community
Advanced
zangxin75zangxin75

k8s-autoscaling

Configure HPA, VPA, and KEDA autoscaling for Kubernetes deployments.

Community
Advanced