compaction

Diagnose StarRocks compaction backlog and version-explosion failures using metrics, SQL, and logs.

73|12|Updated Apr 24, 2026
One-click install
npx skills add https://github.com/StarRocks/starrocks-debug-skills --skill compaction-starrocks
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: compaction
Source: https://github.com/StarRocks/starrocks-debug-skills/tree/main/compaction
Command: npx skills add https://github.com/StarRocks/starrocks-debug-skills --skill compaction-starrocks

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps investigate StarRocks compaction backlog, too many versions import failures, and slow MERGE time by tracing the issue from cluster-level symptoms to a specific tablet and root cause.

Core Features & Use Cases

  • Metric-driven triage: Interprets FE and BE compaction scores, rowset counts, and version thresholds to distinguish cumulative, base, and PK update compaction problems.
  • Log and SQL fallback paths: Uses be.INFO, fe.audit.log, SHOW TABLET, and information_schema queries when metrics are unavailable.
  • Actionable recovery guidance: Recommends manual compaction, configuration tuning, workload throttling, or DROP TABLE FORCE depending on whether the issue is import pressure, DELETE explosion, or large-tablet base compaction delay.

Quick Start

Use the compaction skill to diagnose the given StarRocks tablet or cluster symptom and return the most likely root cause with the next recovery step.

Frequently Asked Questions about compaction

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

FAQPage Schema
How do I troubleshoot StarRocks compaction backlog and too many versions import failures?

To troubleshoot StarRocks compaction backlog, trace symptoms from cluster-level metrics to specific tablets, validating rowset counts, compaction scores, and version thresholds to identify root causes like import pressure or DELETE bursts.

What is the difference between cumulative, base, and update compaction in StarRocks?

StarRocks compaction types differ by scope: cumulative compaction merges smaller rowsets, base compaction handles large-tablet delays by merging into the base rowset, and update compaction manages primary key merge operations.

How do I diagnose StarRocks version explosion when metrics are unavailable?

Diagnose StarRocks version explosion without metrics by using fallback paths: analyze be.INFO and fe.audit.log files, run SHOW TABLET commands, and query information_schema to check tablet versions.

Does compaction troubleshooting apply to shared-nothing StarRocks clusters?

Yes, StarRocks compaction troubleshooting applies to shared-nothing cluster diagnostics, specifically targeting cumulative, base, and update compaction issues caused by import pressure, batch DELETE bursts, and large-tablet delays.

What is the best way to fix slow MERGE time and compaction delays in StarRocks?

Fix slow StarRocks MERGE time by applying actionable recovery steps: trigger manual compaction, tune cluster configurations, throttle import workloads, or execute DROP TABLE FORCE for severe large-tablet base compaction delays.