📚 Complete Guide
🕐 25分钟
📊 Intermediate
✓ Updated 2026-02-28
Cron表达式完全指南 | 从入门到精通
全面深入学习Cron表达式语法，包含10个实际案例、调试技巧、最佳实践和常见问题解答。掌握Linux定时任务调度必备知识。
Learn by doing
Practice with the while reading - free, no registration required
Real-World Examples
Case Study 1: 自动化数据库备份
Problem
一家初创公司需要每天可靠地备份PostgreSQL数据库，但最初几次尝试都失败了，cron日志显示"command not found"错误。
Solution
问题在于pg_dump需要环境变量，且脚本中使用了相对路径。解决方案是在脚本顶部设置环境变量，使用绝对路径，添加日志记录和错误处理：
#!/bin/bash
# 数据库备份脚本
# 环境变量
export PGPASSWORD='your_password'
export PATH=/usr/local/bin:/usr/bin:/bin
# 配置
DB_NAME="mydb"
BACKUP_DIR="/backups"
DATE=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/${DB_NAME}_${DATE}.sql"
# 创建备份
/usr/bin/pg_dump -h localhost -U postgres "$DB_NAME" > "$BACKUP_FILE" 2>&1
# 检查成功
if [ $? -eq 0 ]; then
echo "备份成功: $BACKUP_FILE" >> /var/log/db_backup.log
# 保留最近7天的备份
find "$BACKUP_DIR" -name "${DB_NAME}_*.sql" -mtime +7 -delete
else
echo "备份失败，检查错误信息" >> /var/log/db_backup.log
exit 1
fi
Case Study 2: 微服务健康检查
Problem
一个部署在Kubernetes上的微服务需要定期检查健康状态。服务偶尔会卡住，需要自动重启。
Solution
创建Kubernetes CronJob资源，定期调用健康检查端点：
apiVersion: batch/v1
kind: CronJob
metadata:
name: health-check
spec:
schedule: "*/5 * * * *" # 每5分钟
successfulJobsHistoryLimit: 3
failedJobsHistoryLimit: 3
jobTemplate:
spec:
template:
spec:
containers:
- name: health-check
image: curlimages/curl:latest
imagePullPolicy: Always
command:
- /bin/sh
- -c
- |-
response=$(curl -s -o /dev/null -w "%{http_code}" http://service:8080/health)
if [ $response -ne 200 ]; then
echo "健康检查失败，HTTP状态码: $response"
# 触发重启逻辑
kubectl rollout restart deployment/my-service
fi
restartPolicy: OnFailure
Case Study 3: 日志轮转和归档
Problem
Web服务器日志文件快速增长，磁盘空间经常不足。需要自动轮转、压缩和清理日志。
Solution
使用logrotate配合cron实现日志管理：
/var/log/nginx/*.log {
daily
missingok
rotate 14
compress
delaycompress
notifempty
create 0640 www-data adm
sharedscripts
postrotate
[ -f /var/run/nginx.pid ] && kill -USR1 $(cat /var/run/nginx.pid)
endscript
lastaction
# 将7天前的压缩日志移动到归档目录
find /var/log/nginx -name "*.gz" -mtime +7 -exec mv {} /archive/nginx/ \;
endaction
}
# Cron任务（logrotate自动创建）
0 0 * * * /usr/sbin/logrotate -f /etc/logrotate.conf
Case Study 4: 定时邮件报告
Problem
财务部门需要每天早上8点收到前一天的营收报告邮件。
Solution
创建生成报告并发送邮件的脚本：
#!/usr/bin/env python3
import smtplib
from email.mime.text import MIMEText
from email.mime.multipart import MIMEMultipart
import subprocess
from datetime import datetime, timedelta
def generate_report():
"""生成营收报告"""
yesterday = datetime.now() - timedelta(days=1)
date_str = yesterday.strftime('%Y-%m-%d')
# 执行SQL查询
cmd = f'''
mysql -u reporter -ppassword -h db.internal -e "
SELECT product, SUM(amount) as revenue
FROM sales
WHERE DATE(created_at) = '{date_str}'
GROUP BY product
"
'''
result = subprocess.run(cmd, shell=True, capture_output=True, text=True)
return result.stdout
def send_email(report):
"""发送邮件"""
msg = MIMEMultipart()
msg['From'] = '[email protected]'
msg['To'] = '[email protected]'
msg['Subject'] = f'营收报告 - {datetime.now().strftime("%Y-%m-%d")}'
body = f'''
营收报告
{'='*40}
{report}
{'='*40}
此邮件由系统自动生成，请勿回复。
'''
msg.attach(MIMEText(body, 'plain'))
with smtplib.SMTP('smtp.company.com', 587) as server:
server.starttls()
server.login('[email protected]', 'password')
server.send_message(msg)
if __name__ == '__main__':
report = generate_report()
send_email(report)
Case Study 5: 分布式定时任务同步
Problem
在多台服务器上运行的定时任务需要确保同一时间只有一台执行，避免重复处理。
Solution
使用分布式锁确保任务互斥：
#!/bin/bash
# 分布式数据处理任务
LOCK_FILE="/tmp/data_processing.lock"
LOCK_TIMEOUT=3600 # 1小时超时
REDIS_HOST="redis.internal"
# 尝试获取锁
acquire_lock() {
redis-cli -h "$REDIS_HOST" SETNX "$LOCK_FILE" "$HOSTNAME:$$" > /dev/null
return $?
}
# 释放锁
release_lock() {
redis-cli -h "$REDIS_HOST" DEL "$LOCK_FILE" > /dev/null
}
# 主逻辑
main() {
echo "开始数据处理..."
# 检查锁是否超时
lock_time=$(redis-cli -h "$REDIS_HOST" GET "$LOCK_FILE.ttl")
if [ -n "$lock_time" ]; then
current_time=$(date +%s)
if [ $((current_time - lock_time)) -gt $LOCK_TIMEOUT ]; then
echo "锁已超时，强制获取"
release_lock
fi
fi
# 尝试获取锁
if ! acquire_lock; then
owner=$(redis-cli -h "$REDIS_HOST" GET "$LOCK_FILE")
echo "任务正在运行: $owner"
exit 0
fi
# 设置锁超时
redis-cli -h "$REDIS_HOST" SETEX "$LOCK_FILE.ttl" "$LOCK_TIMEOUT" "$(date +%s)" > /dev/null
# 执行任务
python3 /scripts/process_data.py
# 释放锁
release_lock
echo "数据处理完成"
}
# 执行
main
Case Study 6: 渐进式部署脚本
Problem
需要在不影响服务的情况下滚动更新所有服务器，每批5台，间隔2分钟。
Solution
使用Cron触发的渐进式部署系统：
#!/bin/bash
# 渐进式部署脚本
DEPLOYMENT_CONFIG="/etc/deployment/config.json"
CURRENT_BATCH=$(cat /var/run/deployment_batch 2>/dev/null || echo 0)
BATCH_SIZE=5
BATCH_DELAY=120 # 2分钟
# 获取服务器列表
servers=$(jq -r '.servers[]' "$DEPLOYMENT_CONFIG")
total_servers=$(echo "$servers" | wc -l)
# 计算当前批次
start=$((CURRENT_BATCH * BATCH_SIZE))
end=$((start + BATCH_SIZE))
echo "开始部署批次 $((CURRENT_BATCH + 1))，服务器 $start-$end"
# 部署当前批次
echo "$servers" | sed -n "${start},${end}p" | while read server; do
echo "部署到 $server ..."
ssh "$server" 'bash -s' < /scripts/update_service.sh
done
# 记录进度
if [ $end -lt $total_servers ]; then
echo $((CURRENT_BATCH + 1)) > /var/run/deployment_batch
echo "批次完成，等待下一批..."
else
echo "所有批次部署完成"
rm -f /var/run/deployment_batch
fi
Case Study 7: 证书自动续期
Problem
SSL证书需要每90天续期，手动操作容易忘记导致服务中断。
Solution
使用Let's Encrypt自动续期：
#!/bin/bash
# SSL证书自动续期
DOMAINS="example.com www.example.com api.example.com"
EMAIL="[email protected]"
CERT_DIR="/etc/letsencrypt/live"
LOG_FILE="/var/log/cert_renewal.log"
log() {
echo "[$(date '+%Y-%m-%d %H:%M:%S')] $1" >> "$LOG_FILE"
}
# 检查证书有效期
check_cert_expiry() {
local domain=$1
local cert_file="$CERT_DIR/$domain/cert.pem"
if [ ! -f "$cert_file" ]; then
log "证书不存在: $domain"
return 1
fi
# 计算剩余天数
expiry_date=$(openssl x509 -enddate -noout -in "$cert_file" | cut -d= -f2)
expiry_epoch=$(date -d "$expiry_date" +%s)
current_epoch=$(date +%s)
days_left=$(( (expiry_epoch - current_epoch) / 86400 ))
echo $days_left
}
# 续期证书
renew_cert() {
log "开始续期证书: $DOMAINS"
certbot certonly --non-interactive --agree-tos --email "$EMAIL" --webroot -w /var/www/html -d $DOMAINS >> "$LOG_FILE" 2>&1
if [ $? -eq 0 ]; then
log "证书续期成功"
# 重载nginx
systemctl reload nginx
# 发送通知
echo "证书续期成功" | mail -s "证书续期通知" "$EMAIL"
else
log "证书续期失败"
echo "证书续期失败，请检查日志" | mail -s "证书续期失败" "$EMAIL"
exit 1
fi
}
# 主逻辑
for domain in $DOMAINS; do
days_left=$(check_cert_expiry "$domain")
if [ $days_left -lt 30 ]; then
log "证书将在 $days_left 天后过期，开始续期"
renew_cert
else
log "证书还有 $days_left 天，无需续期"
fi
done
Case Study 8: 缓存预热任务
Problem
电商网站在流量高峰期缓存命中率低，需要提前预热缓存。
Solution
在高峰前自动预热热门商品缓存：
#!/usr/bin/env python3
import requests
import json
from datetime import datetime
API_BASE = "https://api.example.com"
CACHE_ENDPOINT = f"{API_BASE}/cache/warmup"
def get_popular_products():
"""获取热门商品列表"""
response = requests.get(f"{API_BASE}/analytics/popular", params={
'limit': 1000,
'time_range': '24h'
})
return response.json()['products']
def warmup_cache(product_ids):
"""预热缓存"""
chunks = [product_ids[i:i+50] for i in range(0, len(product_ids), 50)]
for chunk in chunks:
try:
response = requests.post(CACHE_ENDPOINT, json={
'product_ids': chunk,
'ttl': 3600
}, timeout=30)
if response.status_code == 200:
print(f"预热成功: {len(chunk)} 个商品")
else:
print(f"预热失败: {response.text}")
except Exception as e:
print(f"预热错误: {e}")
def main():
print(f"[{datetime.now()}] 开始缓存预热")
try:
# 获取热门商品
products = get_popular_products()
product_ids = [p['id'] for p in products]
print(f"获取到 {len(product_ids)} 个热门商品")
# 预热缓存
warmup_cache(product_ids)
print(f"[{datetime.now()}] 缓存预热完成")
except Exception as e:
print(f"预热失败: {e}")
exit(1)
if __name__ == '__main__':
main()
Case Study 9: 数据同步和清理
Problem
需要将生产数据库的脱敏数据同步到测试环境，并保留最近30天的数据。
Solution
创建数据同步和清理脚本：
#!/bin/bash
# 数据库同步到测试环境
PROD_DB="production"
TEST_DB="test_staging"
RETENTION_DAYS=30
log() {
echo "[$(date '+%Y-%m-%d %H:%M:%S')] $1"
}
# 导出生产数据（脱敏）
dump_data() {
log "开始导出生产数据..."
mysqldump -h prod-db.internal -u exporter -ppassword \
--single-transaction \
--quick \
--lock-tables=false \
--where="DATE(created_at) >= DATE(NOW()) - INTERVAL $RETENTION_DAYS DAY" \
"$PROD_DB" \
users orders products \
| gzip > /tmp/dump.sql.gz
# 脱敏处理
zcat /tmp/dump.sql.gz | sed \
-e 's/"\(email\|phone\|ssn\)":"[^"]*"/"\1":"***"/g' \
-e 's/"\(credit_card\|password\)":"[^"]*"/"\1":"***"/g' \
| gzip > /tmp/dump_anon.sql.gz
log "数据导出和脱敏完成"
}
# 导入到测试环境
import_data() {
log "开始导入到测试环境..."
gunzip < /tmp/dump_anon.sql.gz | mysql -h test-db.internal -u importer -ppassword "$TEST_DB"
log "数据导入完成"
}
# 清理临时文件
cleanup() {
log "清理临时文件..."
rm -f /tmp/dump*.sql.gz
}
# 清理测试环境旧数据
cleanup_old_data() {
log "清理测试环境旧数据..."
mysql -h test-db.internal -u admin -ppassword "$TEST_DB" <<EOF
DELETE FROM orders WHERE DATE(created_at) < DATE(NOW()) - INTERVAL $RETENTION_DAYS DAY;
DELETE FROM audit_logs WHERE DATE(created_at) < DATE(NOW()) - INTERVAL $RETENTION_DAYS DAY;
OPTIMIZE TABLE users, orders, products;
EOF
log "旧数据清理完成"
}
# 主流程
main() {
log "===== 数据同步开始 ====="
dump_data
import_data
cleanup_old_data
cleanup
log "===== 数据同步完成 ====="
}
main
Case Study 10: 监控告警集成
Problem
定时任务失败时需要立即通知运维团队，集成现有监控系统。
Solution
使用Prometheus和Alertmanager集成告警：
#!/bin/bash
# 带监控集成的任务包装器
TASK_NAME="data_import"
TASK_COMMAND="/usr/bin/python3 /scripts/import.py"
LOG_FILE="/var/log/tasks/${TASK_NAME}.log"
METRICS_FILE="/var/lib/node_exporter/textfile_collector/${TASK_NAME}.prom"
# 初始化
START_TIME=$(date +%s)
echo "# HELP task_last_success_timestamp Last success time" > "$METRICS_FILE"
echo "# TYPE task_last_success_timestamp gauge" >> "$METRICS_FILE"
# 执行任务
run_task() {
echo "[$(date)] 开始执行任务: $TASK_NAME" >> "$LOG_FILE"
$TASK_COMMAND >> "$LOG_FILE" 2>&1
EXIT_CODE=$?
END_TIME=$(date +%s)
DURATION=$((END_TIME - START_TIME))
if [ $EXIT_CODE -eq 0 ]; then
echo "[$(date)] 任务成功，耗时: ${DURATION}s" >> "$LOG_FILE"
# 记录成功指标
echo "task_last_success_timestamp{task="$TASK_NAME"} $END_TIME" >> "$METRICS_FILE"
echo "task_duration_seconds{task="$TASK_NAME",status="success"} $DURATION" >> "$METRICS_FILE"
else
echo "[$(date)] 任务失败，退出码: $EXIT_CODE" >> "$LOG_FILE"
# 记录失败指标
echo "task_last_success_timestamp{task="$TASK_NAME"} 0" >> "$METRICS_FILE"
echo "task_duration_seconds{task="$TASK_NAME",status="failed"} $DURATION" >> "$METRICS_FILE"
# 发送告警
curl -X POST http://alertmanager:9093/api/v1/alerts -d '[
{
"labels": {
"alertname": "TaskFailed",
"task": "'$TASK_NAME'",
"severity": "critical"
},
"annotations": {
"summary": "任务失败: '$TASK_NAME'",
"description": "退出码: '$EXIT_CODE'，查看日志: '$LOG_FILE'"
}
}
]'
exit 1
fi
}
run_task