prometheus-cardinality-troubleshooter

Diagnose Prometheus cardinality issues via TSDB status and churn analysis.

Updated Mar 18, 2026
One-click install
npx skills add https://github.com/Canepro/codex-skills --skill prometheus-cardinality-troubleshooter
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: prometheus-cardinality-troubleshooter
Source: https://github.com/Canepro/codex-skills/tree/main/skills/prometheus-cardinality-troubleshooter
Command: npx skills add https://github.com/Canepro/codex-skills --skill prometheus-cardinality-troubleshooter

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prometheus cardinality fires can cause memory pressure, slow queries, and inflated costs when a few labels or metrics explode in value space. This guide provides a practical triage workflow to quickly identify the root offenders (metrics and labels) and coordinate a safe remediation path.

Core Features & Use Cases

  • Active-series triage: measure total active series, inspect TSDB status, and surface top offenders by metric and label to focus investigation.
  • Per-metric drill-down: count distinct label values for a metric to spot unbounded values and identify churn drivers.
  • Remediation guidance: follow a decision tree to decide between dropping entire metrics at scrape, applying Adaptive Metrics for post-ingest cost control, or fixing the source instrumentation.

Quick Start

Identify the metric driving cardinality and initiate the triage workflow to isolate the offending series.

Frequently Asked Questions about prometheus-cardinality-troubleshooter

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

FAQPage Schema
How do I triage Prometheus cardinality issues during a production incident?

To triage Prometheus cardinality issues, inspect the TSDB status to measure total active series, then drill down into specific metrics to count distinct label values and identify unbounded churn drivers.

What causes high cardinality in Prometheus metrics and how do I find the offending labels?

High cardinality in Prometheus metrics is caused by labels exploding in value space. You can find offending labels by drilling down into a metric to count distinct label values and spot unbounded dimensions.

What is the best way to reduce Prometheus memory pressure from high-cardinality metrics?

The best way to reduce memory pressure is to follow a remediation decision tree to either drop entire metrics at scrape, apply Adaptive Metrics for post-ingest cost control, or fix the source instrumentation.

When should I drop metrics at scrape versus using Adaptive Metrics to control Prometheus costs?

Drop metrics at scrape when entire metrics are unneeded, and use Adaptive Metrics for post-ingest cost control when you need to manage cardinality without modifying the original source instrumentation.

How do I analyze TSDB status to locate top cardinality offenders in Prometheus?

Analyze TSDB status by measuring total active series and surfacing top offenders by metric and label, allowing you to focus investigation on the specific dimensions driving memory growth and slow queries.