Files
geedge-jira/md/OMPUB-1023.md
2025-09-14 22:27:11 +00:00

154 lines
4.7 KiB
Markdown
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

# 【XJ-NPM】CN标签月报数据异常
| ID | Creation Date | Assignee | Status |
|----|----------------|----------|--------|
| OMPUB-1023 | 2023-09-13T11:17:26.000+0800 | 杨威 | 已解决 |
---
CN 8月标签月报存在以下问题需排查
1.IDC整体流量 16号-18号平均延迟过高
2.IDC业务流量bytedance 17号alibaba18号固宽平均延迟过高
初步排查原因为个别会话日志延迟数据异常导致,到达上亿秒,导致平均值过大,已添加附件和评论。**sunjiajia** commented on *2023-09-13T15:30:19.859+0800*:
如附件截图所示:
bytedance 17号存在两条异常数据 common_establish_latency_ms 达到 1.69227E+12
Alibaba 18号存在一条异常数据 common_establish_latency_ms 达到 1.69233E+12
---
**liuyang** commented on *2023-09-14T10:56:04.924+0800*:
[~yangwei] 麻烦看下日志中common_establish_latency_ms 达到 1.69227E+12的问题
---
**yangwei** commented on *2023-10-24T12:02:50.435+0800*:
[~sunjiajia] 帮忙在新疆省口的会话日志中查询过去7天出现这种异常common_establish_latency_ms的日志量并且导出
---
**sunjiajia** commented on *2023-10-24T13:05:32.712+0800*:
查看过去7天日志延迟<200ms的日志条数有42,215,875,949条延迟在200-1000ms日志条数有3,214,283,160条延迟大于1000ms的有742,398,421条延迟大于1000,000ms的有37条延迟最高为23,410,417ms延迟高于1000,000的37条日志详情见附件latency_logs.csv
---
**yangwei** commented on *2023-10-24T16:43:54.391+0800*:
截至2023年10月24日common_establish_latency_ms数值异常分两种情况
问题1数值量级在1.69227E+1217号和18号分别出现2次和1次{*}数值量级接近时间戳{*}推测计算创建延迟时取出的当前时间为0或者较小的数导致计算出该异常值
问题210.23查询过去7天省口的会话日志不存在数值高于1.69E+12的记录分析附件中>1000,000ms的37条延迟较大的日志信息如下
* 所有的会话均为单向流
* 25条日志由10.111.192.161输出5条由10.111.192.161输出
** 10.111.192.161的25条日志记录的对应会话开始和结束的时段均在10.18 12:45-13:15之间
** 所有会话的持续时间跟common_establish_latency_ms值接近
** 部分会话存在重传、丢包的情况
* 推测原因:{*}功能端当前计算TCP会话建立延迟的方式从SYN开始到第一个传输带负载的数据包在单向流的情况下存在较大的误差{*}
---
**zhengchao** commented on *2023-10-26T17:54:04.504+0800*:
定位到bug代码了吗
---
**yangwei** commented on *2023-10-27T11:11:30.254+0800*:
* 问题2
** 原因为计算逻辑不合理。目前实现判断单向流建立延迟需要看到第一个TCP payload才认为连接建立可能存在较大数百-数千秒级别)的延迟值。
* 问题1
** 怀疑为时钟同步造成的异常。从代码逻辑上不会出现问题1级别的延迟值怀疑是时钟同步导致时间戳计算错误请孙佳佳查询了最近7天的全疆日志未复现问题1待持续观察。
 
 
---
**sunjiajia** commented on *2023-12-01T16:56:22.569+0800*:
CN出入口网络质量报告11月28日 Tencent Tcp会话延迟异常如图CN日报-Tencent延迟异常
查询当天日志推测为一条高达17万亿ms的延迟拉高了均值对应tsg日志见附件tencent_latency_logs.csv且该延迟数值量级接近其时间戳。
---
**zhengchao** commented on *2024-11-19T16:53:25.882+0800*:
Issue closed due to no activity
---
# Attachments
Attachment: Alibaba_18.txt
[Alibaba_18.txt](https://gfwleak.exec.li/admin/geedge-jira/raw/branch/master/attachment/44589/Alibaba_18.txt)
Attachment: Bytedance_17.txt
[Bytedance_17.txt](https://gfwleak.exec.li/admin/geedge-jira/raw/branch/master/attachment/44590/Bytedance_17.txt)
Attachment: CN日报-Tencent延迟异常.png
![CN日报-Tencent延迟异常.png](https://gfwleak.exec.li/admin/geedge-jira/raw/branch/master/attachment/47611/CN日报-Tencent延迟异常.png)
Attachment: latency_logs.csv
[latency_logs.csv](https://gfwleak.exec.li/admin/geedge-jira/raw/branch/master/attachment/46284/latency_logs.csv)
Attachment: screenshot-1.png
![screenshot-1.png](https://gfwleak.exec.li/admin/geedge-jira/raw/branch/master/attachment/44593/screenshot-1.png)
Attachment: screenshot-2.png
![screenshot-2.png](https://gfwleak.exec.li/admin/geedge-jira/raw/branch/master/attachment/44594/screenshot-2.png)
Attachment: tencent_latency_logs.csv
[tencent_latency_logs.csv](https://gfwleak.exec.li/admin/geedge-jira/raw/branch/master/attachment/47612/tencent_latency_logs.csv)