python实现skywalking的trace模块过滤和报警
skywalking本身的报警功能，用起来视乎不是特别好用，目前想实现对skywalking的trace中的错误接口进行过滤并报警通知管理员和开发。所以自己就用python对skywalking做了二次数据清洗实现。项目方在了自己了github（https://github.com/shygit-dev/skywalking-cli-python）上了，有兴趣的同学可以做二次改造，共同学习。下面简单列出了代码内容：
#!/usr/bin/env python
# _*_ coding: utf-8 _*_
# Tile:
# Author:shy
import requests
import time
import smtplib
from email.mime.text import MIMEText
import re
def interface_content_filter(trace_id):
'''
对详细日志内容（业务逻辑报错）进行过滤
:param trace_id:
:return: 【1|0】
'''
url = "http://172.16.53.232:50001/query"
params = {
"trace_id": trace_id
}
detail_trace_id_log = requests.request(method="GET",url=url,params=params)
detail_trace_id_log = detail_trace_id_log.text
print(detail_trace_id_log)
print(type(detail_trace_id_log))
with open("blackname_keyword_list","r",encoding="utf-8") as f:
for line in f:
print(line)
result = re.search(line.strip(),detail_trace_id_log)
print(result)
if result != None:
print("哥们匹配到日志黑名单关键字了：%s" % line)
return 0
print("提示：%s不在关键字黑名单中" % trace_id)
return 1
def interface_filter(endpointName):
"""
设置接口黑名单
:param endpointName:
:return: 【1|0】
"""
endpointName = re.sub("\(|\)",".",endpointName)
with open("blackname_list","r",encoding="utf-8") as f:
bn_list = f.read()
match_result = re.search(endpointName.strip(),bn_list)
if match_result == None:
print("提示：接口不存在黑名单中")
return 1
print("提示：接口在黑名单中")
return 0
def trace_erro_interface(start_time,end_time,sw_url,per_page_size,trace_detail_addr):
"""
skywalking trace功能对错误接口进行过滤，默认最大一次获取2000条数据，每分钟执行一次
:param start_time:
:param end_time:
:return:
"""
url = sw_url
data = {
"query": "query queryTraces($condition: TraceQueryCondition) {\n data: queryBasicTraces(condition: $condition) {\n traces {\n key: segmentId\n endpointNames\n duration\n start\n isError\n traceIds\n }\n total\n }}",
"variables": {
"condition": {
"queryDuration": {
"start": start_time, #"2021-12-07 1734"
"end": end_time,
"step": "MINUTE"
},
"traceState": "ERROR",
"paging": {
"pageNum": 1,
"pageSize": per_page_size,
"needTotal": "true"
},
"queryOrder": "BY_START_TIME"
# "traceId": "b669d0069be84fce82261901de412e7c.430.16388637511348105"
}
}
}
result = requests.request(method="post",url=url,json=data)
i = 0
# print(result.content)
# print(time.strftime("%Y-%m-%d %H:%M:%S",time.localtime(float("%s.%s" % (trace["start"][0:10],trace["start"][10:])))))
with open("mail.html","w",encoding="utf-8") as f:
f.write('<head><meta charset="UTF-8"><title>Title</title><style>.t {border-right: 2px solid black;border-bottom: 2px solid black;}.t th,td {border-top: 2px solid black;border-left: 2px solid black;font-size: 10px;}</style></head><body><div style="color:red;font-size=15px;">最近15分钟统计：</div><table class="t" border="0" cellspacing="0" cellpadding="10px"><thead><tr style="<th style="width: 100px;">时间</th><th>持续时长</th><th>接口名称</th><th>追踪ID</th></tr></thead><tbody>')
for trace in result.json()["data"]["data"]["traces"]:
# print(trace["endpointNames"])
print("时间：%s\n" % time.strftime("%Y-%m-%d %H:%M:%S",time.localtime(float("%s.%s" % (trace["start"][0:10],trace["start"][10:])))),
"持续时长：%s\n" % trace["duration"],
"接口名称：%s\n" % trace["endpointNames"][0],
"跟踪ID：%s" % trace["traceIds"][0])
# print(time.localtime(1638869640.194))
i+=1
print(i)
s_time = time.strftime("%Y-%m-%d %H:%M:%S",time.localtime(float("%s.%s" % (trace["start"][0:10],trace["start"][10:]))))
dur_time = trace["duration"]
endpointName = trace["endpointNames"][0]
trace_id = trace["traceIds"][0]
# 调用接口黑名单过滤功能
result = interface_filter(endpointName)
if result == 0:
print("哥们进入黑名单了！",endpointName)
continue
# 调用关键字黑名单过滤功能
keyword_result = interface_content_filter(trace_id)
if keyword_result == 0:
print("哥们进入关键字黑名单了！", trace_id)
continue
with open("mail.html","a",encoding="utf-8") as f:
f.write('<tr><td>%s</td><td>%s</td><td>%s</td><td><a href="http://%s/query?trace_id=%s">%s</a></td></tr>' %(s_time,dur_time,endpointName,trace_detail_addr,trace_id,trace_id))
with open("mail.html","a",encoding="utf-8") as f:
f.write('</tbody></table></body>')
def send_mail(receiver):
"""
发送报错接口邮件
:return:
"""
server = "mail.test.com"
sender = "sa@test.com"
sender_pwd = "1qaz@WSX"
send_addr = "sa@test.com"
receiver = receiver
with open("mail.html","r",encoding="utf-8") as f:
content = f.read()
if re.search("<td>",content) == None:
print("无报错接口！",content)
return 0
print("邮件前",content)
msg_mail = MIMEText(content,"html","utf-8")
msg_mail["Subject"] = "Skywalking报错接口统计"
msg_mail["From"] = sender
msg_mail["To"] = receiver
server_obj = smtplib.SMTP_SSL(server)
server_obj.connect(server,465)
server_obj.login(sender,sender_pwd)
server_obj.sendmail(send_addr,receiver,msg_mail.as_string())
if __name__ == "__main__":
# 设定查询时间间隔，默认900s（15min）
end_time = time.time()
start_time = end_time - 900
start_time=time.strftime("%Y-%m-%d %H%M",time.localtime(start_time))
end_time = time.strftime("%Y-%m-%d %H%M", time.localtime(end_time))
print(start_time)
print(end_time)
sw_url = "http://172.16.53.232:9412/graphql" # skywalking的前端服务的地址和端口
per_page_size = 5000 #指定一次获取endpoint接口的数目
trace_detail_addr = "127.0.0.1:5000" #指定查询指定trace_id详细日志
receiver = "shy@test.com" #报警邮件接收人地址
trace_erro_interface(start_time,end_time,sw_url,per_page_size,trace_detail_addr)
send_mail(receiver)
# interface_filter()
# interface_content_filter("3c4212dd2dd548d394ba312c4619405d.104.16390380592724487")
#!/usr/bin/env python
# _*_ coding: utf-8 _*_
# Tile:
# Author:shy
import requests
import time
from flask import Flask,request
app = Flask(__name__)
@app.route("/query",methods=["get"])
def trace_id_query():
"""
查询指定trace_id详细日志信息
:return: f.read()
"""
trace_id = request.args.get("trace_id")
url="http://172.16.53.232:9412/graphql"
# url="http://skywalking.roulw.com/graphql"
data = {
"query": "query queryTrace($traceId: ID!) {\n trace: queryTrace(traceId: $traceId) {\n spans {\n traceId\n segmentId\n spanId\n parentSpanId\n refs {\n traceId\n parentSegmentId\n parentSpanId\n type\n }\n serviceCode\n serviceInstanceName\n startTime\n endTime\n endpointName\n type\n peer\n component\n isError\n layer\n tags {\n key\n value\n }\n logs {\n time\n data {\n key\n value\n }\n }\n }\n }\n }",
"variables": {
"traceId": trace_id
}
}
result = requests.request(method="post",url=url,json=data)
with open("detail_log", "w", encoding="utf-8") as f:
f.write("<div style='color: red;font-size: 30px;'>生产Skywalking报错接口跟踪日志日志：<br /></div>")
for trace_id in result.json()["data"]["trace"]["spans"]:
if trace_id["isError"]:
# print(trace_id)
print("服务名称：%s\n" % trace_id["serviceCode"],
"开始时间：%s\n" % trace_id["startTime"],
"接口名称：%s\n" % trace_id["endpointName"],
"peer名称：%s\n" % trace_id["peer"],
"tags名称：%s\n" % trace_id["tags"],
"详细日志：%s" % trace_id["logs"])
content = "服务名称：%s<br />开始时间：%s<br />接口名称：%s<br />peer名称：%s<br />tags名称：%s" % (trace_id["serviceCode"],trace_id["startTime"],trace_id["endpointName"],trace_id["peer"],trace_id["tags"])
with open("detail_log","a",encoding="utf-8") as f:
f.write(content)
f.write("<br />********详细日志**********<br />")
for logs in trace_id["logs"]:
for log in logs["data"]:
if log["key"] == "message":
print(log["value"])
with open("detail_log","a",encoding="utf-8") as f:
f.write(log["value"])
# return log["value"]
elif log["key"] == "stack":
print(log["value"])
with open("detail_log","a",encoding="utf-8") as f:
f.write(log["value"])
with open("detail_log", "a", encoding="utf-8") as f:
f.write("<div style='color: red;font-size: 20px;'><br />========下一个接口信息=========<br /></div>")
with open("detail_log","r",encoding="utf-8") as f:
return f.read()
if __name__ == "__main__":
# trace_id = "14447ae7199c40a2b9862411daba180b.2142.16388920322367785"
# trace_id_query(trace_id)
app.run()
python实现skywalking的trace模块过滤和报警的更多相关文章
- Python开发之日志记录模块：logging
1 引言 最近在开发一个应用软件,为方便调试和后期维护,在代码中添加了日志,用的是Python内置的logging模块,看了许多博主的博文,颇有所得.不得不说,有许多博主大牛总结得确实很好.似乎我再写 ...
- python之旅：常用模块
一.time与datetime模块 在Python中,通常有这几种方式来表示时间 时间戳(timestamp):通常来说,时间戳表示的是从1970年1月1号00:00:00开始按照秒计算的偏移量.我们 ...
- python基础学习十 logging模块详细使用【转载】
很多程序都有记录日志的需求,并且日志中包含的信息既有正常的程序访问日志,还可能有错误.警告等信息输出,python的logging模块提供了标准的日志接口,你可以通过它存储各种格式的日志,主要用于输出 ...
- python基础（10）--模块
模块(module) 模块,用一坨代码实现了某个功能的代码集合 类似于函数式编程和面向过程编程,函数式编程则完成一个功能,其他代码用来调用即可,提供了代码的重用性和代码间的耦合.而对于一个复杂的功能来 ...
- Python之日志处理 logging模块
Python之日志处理(logging模块) 本节内容 日志相关概念 logging模块简介 使用logging提供的模块级别的函数记录日志 logging模块日志流处理流程 使用logging四 ...
- python中常⽤的excel模块库
python中常用的excel模块库&安装方法 openpyxl openpyxl是⼀个Python库,用于读取/写⼊Excel 2010 xlsx / xlsm / xltx / xltm⽂ ...
- Python全栈开发【模块】
Python全栈开发[模块] 本节内容: 模块介绍 time random os sys json & picle shelve XML hashlib ConfigParser loggin ...
- Python 学习笔记(6)--常用模块(2)
一.下载安装 下载安装有两种方式: yum\pip\apt-get 或者源码 下载源码 解压源码 进入目录 编译源码 python setup.py build 安装源码 python setup.p ...
- python学习笔记之常用模块（第五天）
参考老师的博客: 金角:http://www.cnblogs.com/alex3714/articles/5161349.html 银角:http://www.cnblogs.com/wupeiqi/ ...
随机推荐
- ajax的get方法获取豆瓣电影前10页的数据
# _*_ coding : utf-8 _*_ # @Time : 2021/11/2 11:45 # @Author : 秋泊酱 # 1页数据 电影条数20 # https://movie.dou ...
- JS中bind、call和apply的作用以及在TS装饰器中的用法
目录 1,前言 1,call 1.1,例子 1.2,直接调用 1.3,将this指向另一个对象 1.4,传递参数 2,apply 2.1,例子 2.2,直接调用 2.3,将this指向另一个对象 2. ...
- 测试平台系列(79) 编写Redis配置功能(下)
大家好~我是米洛! 我正在从0到1打造一个开源的接口测试平台, 也在编写一套与之对应的完整教程,希望大家多多支持. 欢迎关注我的公众号测试开发坑货,获取最新文章教程! 回顾 上一节我们提出了优化Dao ...
- es基本概念
电商实时数据分析平台需要学些什么? [运营指标,流量指标,销售转化指标,客户价值指标,商品指标,营销指标,风险控制指标,市场竞争指标]等电商指标数据. 检索,数据更新,排序,分词,query等缓存机 ...
- python网络自动化运维之telnetlib实验（EVE-NG+pycharm）
运行环境: 物理机:win10 1903 网络设备:EVE-NG模拟器上运行思科三层路由器 网络设备OS版本:cisco ios(versions 15.6) python环境:pycharm 3.3 ...
- 洛谷 P4062 - [Code+#1]Yazid 的新生舞会（权值线段树）
题面传送门 题意: 给出一个序列 \(a\),求 \(a\) 有多少个子区间 \([l,r]\),满足这个区间中出现次数最多的数出现次数 \(>\dfrac{r-l+1}{2}\) \(1 \l ...
- Markdown—.md文件是什么？怎么打开？
md全称markdown,markdown也是一种标记语言. md文件其实可以用常用的文本编辑器都可以打开. 用记事本打开,把markdown文件拖到记事本图标上就可以打开 . 用 subli ...
- 44-Count and Say
Count and Say My Submissions QuestionEditorial Solution Total Accepted: 79863 Total Submissions: 275 ...
- SQL-关联表查询（连表查询）
0.例如:select * from T1,T2 where T1.a=T2.a 1.连表查询 <=> join(inner join)内连接查询 数据源: Persion表: ...
- kubernetes部署 flannel网络组件
创建 flannel 证书和私钥flannel 从 etcd 集群存取网段分配信息,而 etcd 集群启用了双向 x509 证书认证,所以需要为 flanneld 生成证书和私钥. cat > ...