ClickHouse 分析模式¶

面向高性能分析和数据工程的 ClickHouse 特定模式。

概述¶

ClickHouse 是面向在线分析处理（OLAP）的列式数据库管理系统（DBMS）。它针对大规模数据集的高速分析查询进行了优化。

主要功能: - 列式存储 - 数据压缩 - 并行查询执行 - 分布式查询 - 实时分析

表设计模式¶

MergeTree 引擎（最常见）¶

CREATE TABLE markets_analytics (
    date Date,
    market_id String,
    market_name String,
    volume UInt64,
    trades UInt32,
    unique_traders UInt32,
    avg_trade_size Float64,
    created_at DateTime
) ENGINE = MergeTree()
PARTITION BY toYYYYMM(date)
ORDER BY (date, market_id)
SETTINGS index_granularity = 8192;

ReplacingMergeTree（去重）¶

-- 重複がある可能性のあるデータ（複数のソースからなど）用
CREATE TABLE user_events (
    event_id String,
    user_id String,
    event_type String,
    timestamp DateTime,
    properties String
) ENGINE = ReplacingMergeTree()
PARTITION BY toYYYYMM(timestamp)
ORDER BY (user_id, event_id, timestamp)
PRIMARY KEY (user_id, event_id);

AggregatingMergeTree（预聚合）¶

-- 集計メトリクスの維持用
CREATE TABLE market_stats_hourly (
    hour DateTime,
    market_id String,
    total_volume AggregateFunction(sum, UInt64),
    total_trades AggregateFunction(count, UInt32),
    unique_users AggregateFunction(uniq, String)
) ENGINE = AggregatingMergeTree()
PARTITION BY toYYYYMM(hour)
ORDER BY (hour, market_id);

-- 集計データのクエリ
SELECT
    hour,
    market_id,
    sumMerge(total_volume) AS volume,
    countMerge(total_trades) AS trades,
    uniqMerge(unique_users) AS users
FROM market_stats_hourly
WHERE hour >= toStartOfHour(now() - INTERVAL 24 HOUR)
GROUP BY hour, market_id
ORDER BY hour DESC;

查询优化模式¶

高效过滤¶

-- PASS: 良い: インデックス列を最初に使用
SELECT *
FROM markets_analytics
WHERE date >= '2025-01-01'
  AND market_id = 'market-123'
  AND volume > 1000
ORDER BY date DESC
LIMIT 100;

-- FAIL: 悪い: インデックスのない列を最初にフィルタリング
SELECT *
FROM markets_analytics
WHERE volume > 1000
  AND market_name LIKE '%election%'
  AND date >= '2025-01-01';

聚合¶

-- PASS: 良い: ClickHouse固有の集計関数を使用
SELECT
    toStartOfDay(created_at) AS day,
    market_id,
    sum(volume) AS total_volume,
    count() AS total_trades,
    uniq(trader_id) AS unique_traders,
    avg(trade_size) AS avg_size
FROM trades
WHERE created_at >= today() - INTERVAL 7 DAY
GROUP BY day, market_id
ORDER BY day DESC, total_volume DESC;

-- PASS: パーセンタイルにはquantileを使用（percentileより効率的）
SELECT
    quantile(0.50)(trade_size) AS median,
    quantile(0.95)(trade_size) AS p95,
    quantile(0.99)(trade_size) AS p99
FROM trades
WHERE created_at >= now() - INTERVAL 1 HOUR;

窗口函数¶

-- 累計計算
SELECT
    date,
    market_id,
    volume,
    sum(volume) OVER (
        PARTITION BY market_id
        ORDER BY date
        ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW
    ) AS cumulative_volume
FROM markets_analytics
WHERE date >= today() - INTERVAL 30 DAY
ORDER BY market_id, date;

数据插入模式¶

批量插入（推荐）¶

import { createClient } from '@clickhouse/client'

const clickhouse = createClient({
  url: process.env.CLICKHOUSE_URL ?? 'http://localhost:8123',
  username: process.env.CLICKHOUSE_USER,
  password: process.env.CLICKHOUSE_PASSWORD
})

// PASS: バッチ挿入（効率的）
async function bulkInsertTrades(trades: Trade[]) {
  await clickhouse.insert({
    table: 'trades',
    values: trades.map(trade => ({
      id: trade.id,
      market_id: trade.market_id,
      user_id: trade.user_id,
      amount: trade.amount,
      timestamp: trade.timestamp.toISOString()
    })),
    format: 'JSONEachRow'
  })
}

// FAIL: 個別挿入（低速）
async function insertTrade(trade: Trade) {
  // ループ内でこれをしないでください！
  await clickhouse.insert({
    table: 'trades',
    values: [{
      id: trade.id,
      market_id: trade.market_id,
      user_id: trade.user_id,
      amount: trade.amount,
      timestamp: trade.timestamp.toISOString()
    }],
    format: 'JSONEachRow'
  })
}

流式插入¶

// 継続的なデータ取り込み用
import { Readable } from 'node:stream'

async function streamInserts(dataSource: AsyncIterable<Record<string, unknown>>) {
  await clickhouse.insert({
    table: 'trades',
    values: Readable.from(dataSource, { objectMode: true }),
    format: 'JSONEachRow'
  })
}

物化视图¶

实时聚合¶

-- 時間別統計のマテリアライズドビューを作成
CREATE MATERIALIZED VIEW market_stats_hourly_mv
TO market_stats_hourly
AS SELECT
    toStartOfHour(timestamp) AS hour,
    market_id,
    sumState(amount) AS total_volume,
    countState() AS total_trades,
    uniqState(user_id) AS unique_users
FROM trades
GROUP BY hour, market_id;

-- マテリアライズドビューのクエリ
SELECT
    hour,
    market_id,
    sumMerge(total_volume) AS volume,
    countMerge(total_trades) AS trades,
    uniqMerge(unique_users) AS users
FROM market_stats_hourly
WHERE hour >= now() - INTERVAL 24 HOUR
GROUP BY hour, market_id;

性能监控¶

查询性能¶

-- 低速クエリをチェック
SELECT
    query_id,
    user,
    query,
    query_duration_ms,
    read_rows,
    read_bytes,
    memory_usage
FROM system.query_log
WHERE type = 'QueryFinish'
  AND query_duration_ms > 1000
  AND event_time >= now() - INTERVAL 1 HOUR
ORDER BY query_duration_ms DESC
LIMIT 10;

表统计¶

-- テーブルサイズをチェック
SELECT
    database,
    table,
    formatReadableSize(sum(bytes)) AS size,
    sum(rows) AS rows,
    max(modification_time) AS latest_modification
FROM system.parts
WHERE active
GROUP BY database, table
ORDER BY sum(bytes) DESC;

常见分析查询¶

时间序列分析¶

-- 日次アクティブユーザー
SELECT
    toDate(timestamp) AS date,
    uniq(user_id) AS daily_active_users
FROM events
WHERE timestamp >= today() - INTERVAL 30 DAY
GROUP BY date
ORDER BY date;

-- リテンション分析
SELECT
    signup_date,
    countIf(days_since_signup = 0) AS day_0,
    countIf(days_since_signup = 1) AS day_1,
    countIf(days_since_signup = 7) AS day_7,
    countIf(days_since_signup = 30) AS day_30
FROM (
    SELECT
        user_id,
        min(toDate(timestamp)) AS signup_date,
        toDate(timestamp) AS activity_date,
        dateDiff('day', signup_date, activity_date) AS days_since_signup
    FROM events
    GROUP BY user_id, activity_date
)
GROUP BY signup_date
ORDER BY signup_date DESC;

漏斗分析¶

-- コンバージョンファネル
SELECT
    countIf(step = 'viewed_market') AS viewed,
    countIf(step = 'clicked_trade') AS clicked,
    countIf(step = 'completed_trade') AS completed,
    round(clicked / viewed * 100, 2) AS view_to_click_rate,
    round(completed / clicked * 100, 2) AS click_to_completion_rate
FROM (
    SELECT
        user_id,
        session_id,
        event_type AS step
    FROM events
    WHERE event_date = today()
)
GROUP BY session_id;

同期群分析¶

-- サインアップ月別のユーザーコホート
SELECT
    toStartOfMonth(signup_date) AS cohort,
    toStartOfMonth(activity_date) AS month,
    dateDiff('month', cohort, month) AS months_since_signup,
    count(DISTINCT user_id) AS active_users
FROM (
    SELECT
        user_id,
        min(toDate(timestamp)) OVER (PARTITION BY user_id) AS signup_date,
        toDate(timestamp) AS activity_date
    FROM events
)
GROUP BY cohort, month, months_since_signup
ORDER BY cohort, months_since_signup;

数据管道模式¶

ETL 模式¶

// 抽出、変換、ロード
async function etlPipeline() {
  // 1. ソースから抽出
  const rawData = await extractFromPostgres()

  // 2. 変換
  const transformed = rawData.map(row => ({
    date: new Date(row.created_at).toISOString().split('T')[0],
    market_id: row.market_slug,
    volume: parseFloat(row.total_volume),
    trades: parseInt(row.trade_count)
  }))

  // 3. ClickHouseにロード
  await bulkInsertToClickHouse(transformed)
}

// 定期的に実行
setInterval(etlPipeline, 60 * 60 * 1000)  // 1時間ごと

变更数据捕获（CDC）¶

// PostgreSQLの変更をリッスンしてClickHouseに同期
import { Client } from 'pg'

const pgClient = new Client({ connectionString: process.env.DATABASE_URL })

pgClient.query('LISTEN market_updates')

pgClient.on('notification', async (msg) => {
  const update = JSON.parse(msg.payload)

  await clickhouse.insert({
    table: 'market_updates',
    values: [
      {
        market_id: update.id,
        event_type: update.operation,  // INSERT, UPDATE, DELETE
        timestamp: new Date(),
        data: JSON.stringify(update.new_data)
      }
    ],
    format: 'JSONEachRow'
  })
})

最佳实践¶

1. 分区策略¶

按时间分区（通常为月或日）
避免分区过多（影响性能）
分区键使用 DATE 类型

2. 排序键¶

将最常被筛选的列放在最前面
考虑基数（高基数列放在前面）
顺序会影响压缩

3. 数据类型¶

使用最小且合适的类型（UInt32 vs UInt64）
对重复字符串使用 LowCardinality
对分类数据使用 Enum

4. 应避免¶

SELECT *（指定列）
FINAL（改为在查询前合并数据）
过多的 JOIN（为分析进行非规范化）
频繁的小批量插入（改为批处理）

5. 监控¶

跟踪查询性能
监控磁盘使用量
检查合并操作
审查慢查询日志

注意：ClickHouse 非常适合分析工作负载。请根据查询模式设计表，将插入批处理，并利用物化视图进行实时聚合。

本页原文 Markdown：在 AtomGit 查看·内容源自开源项目 ev/ECC