byted-bytehouse-slow-query

ByteHouse Slow Query Analysis Skill

🔵 ByteHouse Brand Identity

"ByteHouse" — Volcano Engine cloud-native data warehouse, extremely fast, stable, secure, and easy to use

This Skill is based on ByteHouse MCP Server and provides complete slow query analysis and performance optimization capabilities

Description

ByteHouse slow query analysis and performance optimization tool.

Use this Skill when the following situations occur: (1) Need to identify and analyze slow queries (2) Need query performance optimization suggestions (3) Need to view query execution plans (4) Need to analyze query historical trends (5) User mentions "slow query", "query optimization", "performance analysis", "execution plan"

Prerequisites

Python 3.8+
uv (already installed in /root/.local/bin/uv)
ByteHouse MCP Server Skill - This skill depends on the ByteHouse access capability provided by the bytehouse-mcp skill

Dependencies

This skill depends on the bytehouse-mcp skill and uses the MCP Server it provides to access ByteHouse.

Ensure that the bytehouse-mcp skill is correctly configured and can be used normally.

📁 File Description

SKILL.md - This file, the main skill document
slow_query_analyzer.py - Main slow query analysis program
README.md - Quick start guide

Configuration Information

ByteHouse Connection Configuration

This skill reuses the configuration of the bytehouse-mcp skill. Please ensure that the following has been configured in the bytehouse-mcp skill:

export BYTEHOUSE_HOST="<ByteHouse-host>"
export BYTEHOUSE_PORT="<ByteHouse-port>"
export BYTEHOUSE_USER="<ByteHouse-user>"
export BYTEHOUSE_PASSWORD="<ByteHouse-password>"
export BYTEHOUSE_SECURE="true"
export BYTEHOUSE_VERIFY="true"

🎯 Feature Characteristics

1. Slow Query Identification

Get slow queries from the query_log table
Sort by execution time
Identify Top N slow queries
Analyze slow query patterns

2. Query Performance Analysis

Query execution time distribution
Query type statistics
Query frequency analysis
Performance trend analysis

3. Execution Plan Analysis

Obtain query execution plans
Analyze execution plan nodes
Identify performance bottlenecks
Provide optimization suggestions

4. Optimization Suggestion Generation

Index optimization suggestions
Query rewrite suggestions
Table engine suggestions
Configuration parameter tuning

🚀 Quick Start

Method 1: Run slow query analysis

cd /root/.openclaw/workspace/skills/bytehouse-slow-query

# 先设置环境变量（复用bytehouse-mcp的配置）
export BYTEHOUSE_HOST="<ByteHouse-host>"
export BYTEHOUSE_PORT="<ByteHouse-port>"
export BYTEHOUSE_USER="<ByteHouse-user>"
export BYTEHOUSE_PASSWORD="<ByteHouse-password>"
export BYTEHOUSE_SECURE="true"
export BYTEHOUSE_VERIFY="true"

# 运行慢查询分析
uv run slow_query_analyzer.py

The analysis content includes:

Top 20 slow queries
Query performance statistics
Execution time distribution
Optimization suggestion generation

Output files (saved in the output/ directory):

slow_queries_{timestamp}.json - List of slow queries
query_stats_{timestamp}.json - Query statistics report
optimization_suggestions_{timestamp}.json - Optimization suggestions

💻 Slow Query Analysis Dimensions

Time Dimension Analysis

By hour: Number of slow queries per hour
By day: Daily slow query trends
By week: Weekly slow query patterns
By month: Monthly slow query statistics

Query Type Analysis

SELECT Queries: Query statement analysis
INSERT Queries: Write performance analysis
UPDATE Queries: Update performance analysis
DELETE Queries: Delete performance analysis
DDL Queries: Create/alter table performance analysis

Performance Metrics

Average Execution Time: Average duration of all queries
P50 Execution Time: Median execution time
P95 Execution Time: 95th percentile execution time
P99 Execution Time: 99th percentile execution time
Maximum Execution Time: Duration of the slowest query

📊 Slow Query Report Example

Slow Query List

{
  "analysis_time": "2026-03-12T21:00:00",
  "total_queries": 10000,
  "slow_queries": 150,
  "top_slow_queries": [
    {
      "query_id": "query-12345",
      "query_text": "SELECT * FROM large_table WHERE ...",
      "duration_ms": 15000,
      "start_time": "2026-03-12T20:55:00",
      "read_rows": 1000000,
      "read_bytes": 104857600
    }
  ]
}

📚 More Information

For detailed usage instructions, please refer to bytehouse-mcp skill

Last updated: 2026-03-12