Skip to content

Latest commit

 

History

History
146 lines (107 loc) · 6.24 KB

README.md

File metadata and controls

146 lines (107 loc) · 6.24 KB
Raw

说明

平常用到的脚本

flamesvg.sh

性能分析脚本。根据需要生成火焰图或者以文本的方式统计函数 CPU 占比，主要使用的命令是 perf 命令。

  • 火焰图： 着眼于模块，从函数调用的角度分析 cpu 占比
  • 函数占用文本： 着眼于函数本身，聚合之后的整个函数占用 cpu 的百分比

help

[root@localhost smaug]# sh flamesvg.sh 
error param
Usage: ./flamesvg.sh [command]

    record       using 'perf record' to generate perf.data
    svg          after using [record] command, than using [svg] to generate flamegraph perf.svg
    txt          after using [record] command, than using [txt] to generate cpu-percentage of symbol(function) perf_symbol.data
    all_svg      equal to [record] + [svg] command
    all_txt      equal to [record] + [txt] command

record

使用 record 命令可以生成 perf.data 文件，抓取 cpu 信息作为原始数据。

使用方式:

sh flamesvg.sh record [record_time]

注意上面的 record_time 值不应该太大， 不然采集到的 perf.data 数据文件会比较大。一般推荐在 10 一下：

sh flamesvg.sh record 10

命令结束之后会生成一个 perf.data 文件，这是 cpu 的原始数据文件；

svg

基于 record 命令生成的 perf.data 进行分析，生成火焰图。

sh flamesvg.sh svg

命令结束后会生成 perf.svg 火焰图文件；

txt

基于 record 命令生成的 perf.data 进行分析，生成以文本方式展示的分析结果，着眼点在于 函数聚合之后的 cpu 占比；

sh flamesvg.sh txt

命令结束后会生成 perf_symbol.data 文件，可以打开查看消耗 cpu 的函数排序；

all_svg

等于 record 和 svg 两个命令；

sh flamesvg.sh all_svg 10

all_txt

等于 record 和 txt 两个命令：

sh flamesvg.sh all_txt

nat_birth_attach.py

生日攻击算法的实现和验证，为 p2p 打洞（对称性）提供理论基础（端口预测）。

p: 0.507297 > rate:0.5 k: 23 success
when total_samples is 365, if request for rate greater than 0.5, then try at least 23 times is ok

p: 0.500722 > rate:0.5 k: 302 success
when total_samples is 65536, if request for rate greater than 0.5, then try at least 302 times is ok

p: 0.701186 > rate:0.7 k: 398 success
when total_samples is 65536, if request for rate greater than 0.7, then try at least 398 times is ok

p: 0.801039 > rate:0.8 k: 460 success
when total_samples is 65536, if request for rate greater than 0.8, then try at least 460 times is ok

p: 0.900755 > rate:0.9 k: 550 success
when total_samples is 65536, if request for rate greater than 0.9, then try at least 550 times is ok

p: 0.990025 > rate:0.99 k: 770 success
when total_samples is 64512, if request for rate greater than 0.99, then try at least 770 times is ok

打开文件直接修改 total_samples 和 rate 值就可以验证在总样本量为 total_samples 的场景下，当需要满足至少两个样本一样的概率大于等于rate值需要的尝试次数 。

iftop_remote_ip.py

调用 iftop 命令从 remote_ip 维度统计发送接收流量，脚本中限定了只统计本地 80 和 443 端口，实际场景下可以修改成其他过滤条件或者去掉。

remote_ip:124.160.217.1 send:1450 Kb/s recv:7 Kb/s
remote_ip:124.160.217.1 send:3160 Kb/s recv:20 Kb/s
remote_ip:124.160.217.1 send:2407 Kb/s recv:26 Kb/s
remote_ip:124.160.217.1 send:2696 Kb/s recv:25 Kb/s
remote_ip:124.160.217.2 send:2729 Kb/s recv:25 Kb/s
remote_ip:124.160.217.3 send:2852 Kb/s recv:26 Kb/s
remote_ip:124.160.217.4 send:2650 Kb/s recv:23 Kb/s
remote_ip:124.160.217.1 send:2539 Kb/s recv:25 Kb/s
remote_ip:124.160.217.3 send:2362 Kb/s recv:24 Kb/s
remote_ip:124.160.217.2 send:2741 Kb/s recv:24 Kb/s
remote_ip:124.160.217.4 send:2667 Kb/s recv:23 Kb/s
remote_ip:124.160.217.2 send:2679 Kb/s recv:21 Kb/s
remote_ip:124.160.217.1 send:2540 Kb/s recv:28 Kb/s

mykill

简单封装了下 kill 命令，本质上是 killall 命令的实现，只不过更加友好一点。

smaug@smaug-VirtualBox:~$ mykill  -h
error param
grep all program name  and kill all
Usage: mykill [program_name] [yes/no]

filscan.py

爬取 https://filscan.io/#/tipset/address-detail?address=f02301 网站的数据

From,Receipt,MessageID,Value,Height,To,Time,Method
f3qhvpg5byv25lcdch2z62ml7jp3rxnpt5ogg5jtokjwv7q6sqwzd2e4qw2sqh5kvh7nitnejhao6bc5xnqdna,OK,bafy2bzacedwhwgnhzjsgljm5shbiq5ni7je562ho7uus6fjno4gjgzitsek6a,0.0351423918839 FIL,147292,f02301,2020-10-15 09:26:00,ProveCommitSector
f3tckh5m4xq6rttaosbzh2xjk4wbwzzxjft4srbqup64tmywhofhfug55yuvybhtw7mjcxzztwook2xkuzkggq,OK,bafy2bzaceb5vuuafr4kxslnym5bymjujdh5ophpr6mxqlazldhyva75bgpine,0.0 FIL,147289,f02301,2020-10-15 09:24:30,DeclareFaultsRecovered
f3qhvpg5byv25lcdch2z62ml7jp3rxnpt5ogg5jtokjwv7q6sqwzd2e4qw2sqh5kvh7nitnejhao6bc5xnqdna,OK,bafy2bzacedxubn3patupbnvwiaon7nq6d7psiv2rgeqamm2roxd5kwg7i7gtk,0.0351766999492 FIL,147289,f02301,2020-10-15 09:24:30,ProveCommitSector
f3qhvpg5byv25lcdch2z62ml7jp3rxnpt5ogg5jtokjwv7q6sqwzd2e4qw2sqh5kvh7nitnejhao6bc5xnqdna,OK,bafy2bzacebmhkr6h5utipjqtf6bt2bbuer2l7z4ky7onlpma2cr22giyk565a,0.0351584198055 FIL,147289,f02301,2020-10-15 09:24:30,ProveCommitSector
f3qhvpg5byv25lcdch2z62ml7jp3rxnpt5ogg5jtokjwv7q6sqwzd2e4qw2sqh5kvh7nitnejhao6bc5xnqdna,OK,bafy2bzaceasyr2bt3tynwufadzxvhryybly4vfwa4psjw7j6rg66o2l66fedy,0.168976637674 FIL,147288,f02301,2020-10-15 09:24:00,PreCommitSector
f3qhvpg5byv25lcdch2z62ml7jp3rxnpt5ogg5jtokjwv7q6sqwzd2e4qw2sqh5kvh7nitnejhao6bc5xnqdna,OK,bafy2bzacedceqm2isw7rsuiroe4sninkql4msibjn4jt2spnyjelcoft7qc7s,0.168990631288 FIL,147287,f02301,2020-10-15 09:23:30,PreCommitSector
f3qhvpg5byv25lcdch2z62ml7jp3rxnpt5ogg5jtokjwv7q6sqwzd2e4qw2sqh5kvh7nitnejhao6bc5xnqdna,OK,bafy2bzaceboueg6uvwhi7nthrtlrs7hc6doiycurhqbywzktmmoera4uqgfis,0.0352552658287 FIL,147286,f02301,2020-10-15 09:23:00,ProveCommitSector
f3qhvpg5byv25lcdch2z62ml7jp3rxnpt5ogg5jtokjwv7q6sqwzd2e4qw2sqh5kvh7nitnejhao6bc5xnqdna,OK,bafy2bzacectjyk77l3t5tuoxkotjzo33sobp6mrdtbwousfl7tj5scj2rtp32,0.169004548953 FIL,147286,f02301,2020-10-15 09:23:00,PreCommitSector
f3qhvpg5byv25lcdch2z62ml7jp3rxnpt5ogg5jtokjwv7q6sqwzd2e4qw2sqh5kvh7nitnejhao6bc5xnqdna,OK,bafy2bzacecb7okvcovy42hqxr5jyfw2dlnftom4zan6tvasmer2efvnzoq774,0.0352578666153 FIL,147284,f02301,2020-10-15 09:22:00,ProveCommitSector