Python脚本收集腾讯云CDN日志，并入ELK日志分析

负责搭建公司日志分析，一直想把CDN日志也放入到日志分析，前些日志终于达成所愿，现在贴出具体做法：

1、收集日志

　　腾讯云CDN日志一般一小时刷新一次，也就是说当前只能下载一小时之前的日志数据，但据本人观察，有时前一小时的并下载不到，所以为了保险起见，可以下载两小时之前的日志数据。下载日志可以通过腾讯云的API获取日志列表，然后下载。

　　腾讯云日志下载API 链接：https://www.qcloud.com/document/product/228/8087

日志采集脚本：

[root@BJVM-2-181 bin]# cat get_cdn_log.py
#!/usr/bin/env python
# coding=utf-8

import hashlib
import requests
import hmac
import random
import time
import base64
import json
import gzip
import os
import sys
from datetime import datetime, timedelta

class Sign(object):

    def __init__(self, secretId, secretKey):
        self.secretId = secretId
        self.secretKey = secretKey

    # 生成签名串
    def make(self, requestHost, requestUri, params, method='GET'):
        srcStr = method.upper() + requestHost + requestUri + '?' + "&".join(k.replace("_",".") + "=" + str(params[k]) for k in sorted(params.keys()))
        hashed = hmac.new(self.secretKey, srcStr, hashlib.sha1)
        return base64.b64encode(hashed.digest())

class CdnHelper(object):
    SecretId='AKIDLsldjflsdjflsdjflsdjfpGSO5XoGiY9'
    SecretKey='SeaHjSDFLJSLDFJQIuFJ7rMiz0lGV'
    requestHost='cdn.api.qcloud.com'
    requestUri='/v2/index.php'

    def __init__(self, host, startDate, endDate):
        self.host = host
        self.startDate = startDate
        self.endDate = endDate
        self.params = {
            'Timestamp': int(time.time()),
            'Action': 'GetCdnLogList',
            'SecretId': CdnHelper.SecretId,
            'Nonce': random.randint(10000000,99999999),
            'host': self.host,
            'startDate': self.startDate,
            'endDate': self.endDate
        }
        self.params['Signature'] =  Sign(CdnHelper.SecretId, CdnHelper.SecretKey).make(CdnHelper.requestHost, CdnHelper.requestUri, self.params)
        self.url = 'https://%s%s' % (CdnHelper.requestHost, CdnHelper.requestUri)

    def GetCdnLogList(self):
        ret = requests.get(self.url, params=self.params)
        return ret.json()

class GZipTool(object):
    """
    压缩与解压gzip
    """
    def __init__(self, bufSize = 1024*8):
        self.bufSize = bufSize
        self.fin = None
        self.fout = None
    def compress(self, src, dst):
        self.fin = open(src, 'rb')
        self.fout = gzip.open(dst, 'wb')
        self.__in2out()
    def decompress(self, gzFile, dst):
        self.fin = gzip.open(gzFile, 'rb')
        self.fout = open(dst, 'wb')
        self.__in2out()
    def __in2out(self,):
        while True:
            buf = self.fin.read(self.bufSize)
            if len(buf) < 1:
                break
            self.fout.write(buf)
        self.fin.close()
        self.fout.close()

def download(link, name):
    try:
        r = requests.get(link)
        with open(name, 'wb') as f:
            f.write(r.content)
        return True
    except:
        return False

def writelog(src, dst):
    # 保存为以天命名日志
    dst = dst.split('-')[0][:-2] + '-' + dst.split('-')[1]
    with open(src, 'r') as f1:
       with open(dst, 'a+') as f2:
        for line in f1:
            f2.write(line)

if __name__ == '__main__':
    #startDate = "2017-02-23 12:00:00"
    #endDate = "2017-02-23 12:00:00"

    # 前一小时
    # startDate = endDate = time.strftime('%Y-%m-%d ', time.localtime()) + str(time.localtime().tm_hour-1) + ":00:00"
    tm = datetime.now() + timedelta(hours=-2)
    startDate = endDate = tm.strftime("%Y-%m-%d %H:00:00")

    #hosts = ['userface.51img1.com']
    hosts = [
        'pfcdn.xxx.com',
        'pecdn.xxx.com',
        'pdcdn.xxx.com',
        'pccdn.xxx.com',
        'pbcdn.xxx.com',
        'pacdn.xxx.com',
        'p9cdn.xxx.com',
        'p8cdn.xxx.com',
        'p7cdn.xxx.com',
        ]                

    for host in hosts:
        try:
            obj = CdnHelper(host, startDate,endDate)
            ret = obj.GetCdnLogList()

            link = ret['data']['list'][0]['link']
            name = ret['data']['list'][0]['name']

            # 下载链接保存的文件名
            gzip_name = '/data/logs/cdn/cdn_log_temp/' + name + '.gz'
            # 解压后的文件名
            local_name = '/data/logs/cdn/cdn_log_temp/' + name + '.log'
            # 追加的文件名
            real_path = '/data/logs/cdn/' + name + '.log'
            print local_name, real_path

            status = download(link, gzip_name)
            if status:
                try:
                    GZipTool().decompress(gzip_name, local_name)
                    writelog(local_name, real_path)
    #                os.remove(gzip_name)
                    os.remove(local_name)
                except:
                    continue
        except Exception ,e:
            print e
            continue

get_cdn_log.py

放到定时任务，每小时执行一次

# cdn日志
 */ * * * /usr/bin/python /root/bin/get_cdn_log.py &> /dev/null

此图解压后的日志，每个域名保存为一个文件，按天分割。

　　

2、filebeat配置（具体含义查看官方文档）

[root@BJ-- bin]# cat /usr/local/app/filebeat-1.2.-x86_64/nginx-php.yml
filebeat:
  prospectors:
    -
      paths:
        - /data/logs/cdn/*.log
      document_type: cdn-log
      input_type: log
      #tail_files: true
      multiline:
        negate: true
        match: after
output:
  logstash:
    hosts: ["10.80.2.181:5048", "10.80.2.182:5048"]
shipper:
logging:
  files:

3、logstash配置

日志格式：

 61.135.234.125 cdn.xxx.com /game////57037f7fc1a0dde9091d4fe6502a6c53.jpg     http://www.xxx.com/ 5 "Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; Trident/4.0; SLCC2; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR 3.0.30729; Media Center PC 6.0; NetworkBench/7.0.0.282-5004888-124025)" "(null)" GET HTTP/1.1 hit

日志内容依次包括：请求时间、访问域名的客户端IP、被访问域名、文件请求路径、本次访问字节数大小、省份、运营商、http返回码、referer信息、request-time(毫秒)、User-Agent、range、HTTP Method、HTTP协议标识、缓存Hit/Miss。

配置文件

# /usr/local/app/logstash-2.3.4/conf.d/logstash.conf
input {
    beats {
        port =>
        host => "0.0.0.0"
    }
}

filter {

    .....(省略)

    else if [type] == "cdn-log" {
        grok {
            patterns_dir => ["./patterns"]
            match => { "message" => "%{DATESTAMP_EVENTLOG:timestamp} %{IPORHOST:client_ip} %{IPORHOST:server_name} %{NOTSPACE:request} %{NUMBER:bytes} %{NUMBER:province} %{NUMBER:operator} %{NUMBER:status} (?:%{URI:referrer}|%{WORD:referrer}) %{NUMBER:request_time} %{QS:agent} \"\(%{WORD:range}\)\" %{WORD:method} HTTP/%{NUMBER:protocol} %{WORD:cache}" }
        }
        date {
            match => [ "timestamp", "yyyyMMddHHmmss"]
            target => "@timestamp"
        }
        alter {
            condrewrite => [
                "province", "", "北京",
                "province", "", "内蒙古",
                "province", "", "山西",
                "province", "", "河北",
                "province", "", "天津",
                "province", "", "宁夏",
                "province", "", "陕西",
                "province", "", "甘肃",
                "province", "", "青海",
                "province", "", "新疆",
                "province", "", "黑龙江",
                "province", "", "吉林",
                "province", "", "辽宁",
                "province", "", "福建",
                "province", "", "江苏",
                "province", "", "安徽",
                "province", "", "山东",
                "province", "", "上海",
                "province", "", "浙江",
                "province", "", "河南",
                "province", "", "湖北",
                "province", "", "江西",
                "province", "", "湖南",
                "province", "", "贵州",
                "province", "", "云南",
                "province", "", "重庆",
                "province", "", "四川",
                "province", "", "西藏",
                "province", "", "广东",
                "province", "", "广西",
                "province", "", "海南",
                "province", "", "其他",
                "province", "", "港澳台",
                "province", "", "海外",
                "operator", "", "中国电信",
                "operator", "", "中国联通",
                "operator", "", "教育网",
                "operator", "", "长城宽带",
                "operator", "", "中国移动",
                "operator", "", "中国铁通",
                "operator", "-1", "海外运营商",
                "operator", "", "其他运营商"
            ]
        }
    }

} # filter

output {

    if "_grokparsefailure" in [tags] {
        file { path => "/var/log/logstash/grokparsefailure-%{[type]}-%{+YYYY.MM.dd}.log" }
    }
    ......(省略)

    else if [type] == "cdn-log"{
        elasticsearch {
            hosts => ["10.80.2.13:9200","10.80.2.14:9200","10.80.2.15:9200","10.80.2.16:9200"]
            sniffing => true
            manage_template => true
            template_overwrite => true
            template_name => "cdn"
            template => "/usr/local/app/logstash-2.3.4/templates/cdn.json"
            index => "%{[type]}-%{+YYYY.MM.dd}"
            document_type => "%{[type]}"
        }
    }
    ......(省略)

} # output

4 效果图（一小时数据）

cdn使用量效果图

cdn访问情况统计

状态码统计