PDO批量并发执行SSH工具介绍

时间 2019-11-10

标签 pdo 批量并发执行 ssh 工具介绍栏目 PHP 繁體版

原文原文链接

批量并发执行工具PDO,主要是解决批量执行的繁锁,更安全便捷的操做工具.
自己是解决公司内部的一些问题,而且有不少特定环境的一些使用,如今抽离出其中均可以使用的部分.linux

工具名称: pdo(parallel do something) https://github.com/ohlinux/pdo

安装

依赖

须要有一个中控机与被管理机器创建了无密码的密钥关系.
须要有go语言的环境,进行编译安装.这里没有提供bin文件.
本身所测试的环境有,centos macos.

编译

先获取依赖的第三方库:git

go get github.com/cihub/seelog
go get github.com/robfig/config

安装go 环境.github

go build pdo.go

配置文件目录,默认在~/.pdo.若是不在此处指定.web

pdo 结构

pdo 处理对象来源

获取机器列表和相对应的路径有三种途径.(这里去掉了数据库这种特定的)redis

-f 文件,host的列表文件,能够是一列,也能够是两列有相关的目录依赖.后面有例子.
-a app名字;-p 产品名;-a支持多app采用 app1,app2逗号分隔. (这个是数据库的来源,由于是特定环境的因此再也不有)
标准输入 cat 1.host | pdo
-R当使用的时候,能够自动生成失败的列表.详细查看例子"Retry功能"

pdo 列表过滤

若是列表名称是这样的结构,xxx.yyy 那么过滤的就是yyy,若是没有这个须要,能够忽略.数据库

-i yf01,dbl01,cq02 过滤物理机房名称,多个可用逗号隔开.
-I JX/TC 过滤逻辑机房,配置在-c configure file 中

配置文件中:macos

[IDC]
    JX:yf01,cq01,dbl01,ai01
    TC:cq02,tc,m1,db01

日志记录

日志主要记录使用者,使用过的命令,保证多人操做的时候能够查看到.centos

日志配置文件查看github.com/cihub/seelog
主配置文件格式查看github.com/robfig/config缓存

pdo 功能

会有主机和命令和单台执行确认.安全

-r 数字 ,并发量,默认为1
-C configure , 配置文件,默认为~/.pdo/pdo.conf
-o dir , 输出目录,默认为空,会打印在屏幕,若是添加则只会打印到文件中.
-cmd 命令, 命令缩写,在配置文件中.
-t 超时结束,超时时间默认5min钟.
-y 不用输入确认.
-c copy file/dir ,复制文件/或者目录,要求无故的目录必需存在.
-e script 执行脚本.
-T 执行间隔时间,默认为0,比命令中加sleep效果好.
-R retry ,失败以后retry功能,会记录上次失败的列表和ctrl+c未执行的列表.
-temp template模板名字,在配置中.
-b build-in 在使用template的模板时候能够嵌入脚本.
-V 查看version
-show 查看显示方式"row" 行显示方式 .
-match 在行显示的时候能够进行match字符串.高亮红色显示.
-rule 在行显示的模式下,可使用conf中的rule,来定义不一样match的字符串的动做.(暂时无)

pdo 未完成功能

去重与多实例,产品线重启并发优化.
web页面展现功能.

注意事项

操做的命令,被引号括起来的,必定要保持在命令最后.

pdo使用用举例

配置文件

[PDO]
 logconf:/home/work/.pdo/log.xml

 [IDC]
 JX:yf01,cq01,dbl01,ai01,jx,cp01
 TC:cq02,tc,m1,db01,st01

 [TEMPLATE]
 container : /home/work/.pdo/template/container.sh
 startbykill : /home/work/.pdo/template/startbykill.sh

 [CMD]
 restart: bash bin/xxxControl.sh N%%N%%N%%restart
 findLog: find xxx00* -name "debug" -type d
 findCount: ls log | wc -l

host列表文件

第一列必定是host,hostname或者ip均可以,第二列可选是命令工做的路径.

cat godir/1.list
yf-xxx-app01.yf01 /home/work/xxx001
yf-xxx-app02.yf01 /home/work/xxx004
yf-xxx-app03.yf01 /home/work/xxx002

使用管道方式

cat 1.list | pdo -r 2 "pwd"
>>>> Welcome ajian...
yf-xxx-pre01.vm          -/home/work/xxx001    yf-xxx-app01.yf01        -/home/work/xxx001
yf-xxx-app02.yf01        -/home/work/xxx001    yf-xxx-app03.yf01        -/home/work/xxx001
yf-xxx-app04.yf01        -/home/work/xxx001    1-xxx-app17.m1           -/home/work/xxx001
m1-xxx-app25.m1          -/home/work/xxx001    m1-xxx-app0220.m1        -/home/work/xxx001
m1-xxx-app0154.m1        -/home/work/xxx004    cq01-xxx-app0242.cq01    -/home/work/xxx003
ai-xxx-app01.ai01        -/home/work/xxx004    db-xxx-app17.db01        -/home/work/xxx003
db-xxx-app63.db01        -/home/work/xxx001

#--Total--#  13
#---CMD---#  pwd
//每一次确认
Continue (y/n):y
go on ...
[1/13] yf-xxx-app01.yf01  [SUCCESS].
/home/work/xxx001

Continue (y/n):[1/13] yf-xxx-pre01.vm  [SUCCESS].
/home/work/xxx001
//单台执行完 第二次确认
Continue (y/n):
//后面就是按2并发执行.

使用简写命令

这个主要是解决一些重复执行的繁锁的单行命令.看下面的一个重启命令非常麻烦,但经过转换以后就输入很方便了.

-cmd为缩写命令= bash bin/xxxControl.sh N%%N%%N%%restart

$ pdo -f 1.list  -cmd restart
>>>> Welcome ajian...
yf-xxx-app01.yf01        -/home/work/xxx001    yf-xxx-app02.yf01        -/home/work/xxx001
yf-xxx-app03.yf01        -/home/work/xxx001    yf-xxx-app04.yf01        -/home/work/xxx001
yf-xxx-app00.yf01        -/home/work/xxx001    yf-xxx-app0148.yf01      -/home/work/xxx004
dbl-xxx-app0109.dbl01    -/home/work/xxx003    m1-xxx-app17.m1          -/home/work/xxx001
m1-xxx-app25.m1          -/home/work/xxx001    m1-xxx-app0220.m1        -/home/work/xxx001
m1-xxx-app0154.m1        -/home/work/xxx004    cq01-xxx-app0242.cq01    -/home/work/xxx003
cq01-xxx-app0179.cq01    -/home/work/xxx001    cq01-xxx-app0131.cq01    -/home/work/xxx005
st01-xxx-app03.st01      -/home/work/xxx001    st01-xxx-app04.st01      -/home/work/xxx001
st01-xxx-app02.st01      -/home/work/xxx001    st01-xxx-app00.st01      -/home/work/xxx001
st01-xxx-app05.st01      -/home/work/xxx001    cq02-xxx-app0258.cq02    -/home/work/xxx001
cq02-xxx-app0287.cq02    -/home/work/xxx001    jx-xxx-app17.jx          -/home/work/xxx001
ai-xxx-app10.ai01        -/home/work/xxx001    db-xxx-app17.db01        -/home/work/xxx003


#--Total--#  24
#---CMD---#  bash bin/xxxControl.sh N%%N%%N%%restart
Continue (y/n):

-o输入与屏幕输出

使用带-o 指定输出目录,将不会再打印在屏幕上,主要是对grep日志这种需求使用.速度要比屏幕打印快不少,是实时写入.

$ cat 1.list | pdo -o xxxout "pwd"
>>>> Welcome ajian...
yf-xxx-app01.yf01        -/home/work/xxx001    yf-xxx-app02.yf01        -/home/work/xxx001
yf-xxx-app03.yf01        -/home/work/xxx001    yf-xxx-app04.yf01        -/home/work/xxx001
yf-xxx-app00.yf01        -/home/work/xxx001    yf-xxx-app0148.yf01      -/home/work/xxx004
dbl-xxx-app0109.dbl01    -/home/work/xxx003    m1-xxx-app17.m1          -/home/work/xxx001
m1-xxx-app25.m1          -/home/work/xxx001    m1-xxx-app0220.m1        -/home/work/xxx001
m1-xxx-app0154.m1        -/home/work/xxx004    cq01-xxx-app0242.cq01    -/home/work/xxx003
cq01-xxx-app0179.cq01    -/home/work/xxx001    cq01-xxx-app0131.cq01    -/home/work/xxx005
st01-xxx-app03.st01      -/home/work/xxx001    st01-xxx-app04.st01      -/home/work/xxx001
st01-xxx-app02.st01      -/home/work/xxx001    st01-xxx-app00.st01      -/home/work/xxx001
st01-xxx-app05.st01      -/home/work/xxx001    cq02-xxx-app0258.cq02    -/home/work/xxx001
cq02-xxx-app0287.cq02    -/home/work/xxx001    cq02-xxx-app0212.cq02    -/home/work/xxx001
jx-xxx-app17.jx          -/home/work/xxx001    ai-xxx-app10.ai01        -/home/work/xxx001
db-xxx-app17.db01        -/home/work/xxx003

#--Total--#  25
#---CMD---#  pwd
Continue (y/n):y
go on ...
[1/25] yf-xxx-app01.yf01  [SUCCESS].
Continue (y/n):y
go on ...
[2/25] yf-xxx-app02.yf01  [SUCCESS].
[3/25] yf-xxx-app03.yf01  [SUCCESS].
[4/25] yf-xxx-app04.yf01  [SUCCESS].
[5/25] yf-xxx-app00.yf01  [SUCCESS].
[6/25] yf-xxx-app0148.yf01  [SUCCESS].
[7/25] dbl-xxx-app0109.dbl01  [SUCCESS].
[8/25] m1-xxx-app17.m1  [SUCCESS].
[9/25] m1-xxx-app25.m1  [SUCCESS].
[10/25] m1-xxx-app0220.m1  [SUCCESS].
[11/25] m1-xxx-app0154.m1  [SUCCESS].
[12/25] cq01-xxx-app0242.cq01  [SUCCESS].
[13/25] cq01-xxx-app0179.cq01  [SUCCESS].
[14/25] cq01-xxx-app0131.cq01  [SUCCESS].
[15/25] st01-xxx-app03.st01  [SUCCESS].
[16/25] st01-xxx-app04.st01  [SUCCESS].
[17/25] st01-xxx-app02.st01  [SUCCESS].
[18/25] st01-xxx-app00.st01  [SUCCESS].
[19/25] st01-xxx-app05.st01  [SUCCESS].
[20/25] cq02-xxx-app0258.cq02  [SUCCESS].
[21/25] cq02-xxx-app0287.cq02  [SUCCESS].
[22/25] cq02-xxx-app0212.cq02  [SUCCESS].
[23/25] jx-xxx-app17.jx  [SUCCESS].
[24/25] ai-xxx-app10.ai01  [SUCCESS].
[25/25] db-xxx-app17.db01  [SUCCESS].

超时killed进程

时间都带单位,如1秒 1s , 1分钟 1m , 1小时 1h .

这里的1.log是一个大文件.

$ pdo -f 1.list -t 1s -o out/ -r 3 "cat 1.log"
>>>> Welcome ajian...
yf-xxx-app01.yf01        -/home/work/xxx001    yf-xxx-app02.yf01        -/home/work/xxx001
yf-xxx-app03.yf01        -/home/work/xxx001    yf-xxx-app04.yf01        -/home/work/xxx001
yf-xxx-app00.yf01        -/home/work/xxx001    yf-xxx-app0148.yf01      -/home/work/xxx004
dbl-xxx-app0109.dbl01    -/home/work/xxx003    m1-xxx-app17.m1          -/home/work/xxx001
m1-xxx-app25.m1          -/home/work/xxx001    m1-xxx-app0220.m1        -/home/work/xxx001
m1-xxx-app0154.m1        -/home/work/xxx004    cq01-xxx-app0242.cq01    -/home/work/xxx003
cq01-xxx-app0179.cq01    -/home/work/xxx001    cq01-xxx-app0131.cq01    -/home/work/xxx005
st01-xxx-app03.st01      -/home/work/xxx001    st01-xxx-app04.st01      -/home/work/xxx001
st01-xxx-app02.st01      -/home/work/xxx001    st01-xxx-app00.st01      -/home/work/xxx001
st01-xxx-app05.st01      -/home/work/xxx001    cq02-xxx-app0258.cq02    -/home/work/xxx001
cq02-xxx-app0287.cq02    -/home/work/xxx001    cq02-xxx-app0211.cq02    -/home/work/xxx001
cq02-xxx-app0212.cq02    -/home/work/xxx001    jx-xxx-app17.jx          -/home/work/xxx001
ai-xxx-app10.ai01        -/home/work/xxx001    db-xxx-app17.db01        -/home/work/xxx003


#--Total--#  26
#---CMD---#  cat log/ral-zoo.log
Continue (y/n):y
go on ...
[1/26] yf-xxx-app01.yf01  [Time Over KILLED].
Continue (y/n):y
go on ...
[2/26] yf-xxx-app04.yf01  [Time Over KILLED].
[3/26] yf-xxx-app03.yf01  [Time Over KILLED].
[4/26] yf-xxx-app02.yf01  [Time Over KILLED].
[5/26] yf-xxx-app0148.yf01  [SUCCESS].
[6/26] dbl-xxx-app0109.dbl01  [SUCCESS].
[7/26] yf-xxx-app00.yf01  [Time Over KILLED].
[8/26] m1-xxx-app0220.m1  [SUCCESS].
[9/26] m1-xxx-app25.m1  [Time Over KILLED].
[10/26] m1-xxx-app17.m1  [Time Over KILLED].
[11/26] m1-xxx-app0154.m1  [SUCCESS].
[12/26] cq01-xxx-app0242.cq01  [SUCCESS].
[13/26] cq01-xxx-app0179.cq01  [Time Over KILLED].
[14/26] cq01-xxx-app0131.cq01  [SUCCESS].
[15/26] st01-xxx-app03.st01  [Time Over KILLED].
[16/26] st01-xxx-app04.st01  [Time Over KILLED].
[17/26] st01-xxx-app02.st01  [Time Over KILLED].
[18/26] st01-xxx-app00.st01  [Time Over KILLED].
[19/26] st01-xxx-app05.st01  [Time Over KILLED].
[20/26] cq02-xxx-app0211.cq02  [SUCCESS].
[21/26] cq02-xxx-app0258.cq02  [SUCCESS].
[22/26] cq02-xxx-app0287.cq02  [SUCCESS].
[23/26] ai-xxx-app10.ai01  [SUCCESS].
[24/26] cq02-xxx-app0212.cq02  [SUCCESS].
[25/26] jx-xxx-app17.jx  [Time Over KILLED].
[26/26] db-xxx-app17.db01  [SUCCESS].

-c copy文件

copy文件实际上是能够copy目录的,只要远端的目录是存在的就不会报错.

$ cat 1.host  | pdo -c get.sh /tmp/
>>>> Welcome ajian...
yf-xxx-upload05.yf01     -/home/work           yf-xxx-upload01.yf01     -/home/work
yf-xxx-upload02.yf01     -/home/work

#--Total--#  3
#---CMD---#  get.sh --> /tmp/
Continue (y/n):y
go on ...
[1/3] yf-xxx-upload05.yf01  [SUCCESS].

Continue (y/n):y
go on ...
[2/3] yf-xxx-upload01.yf01  [SUCCESS].

[3/3] yf-xxx-upload02.yf01  [SUCCESS].

//检查下文件
$ cat 1.host | pdo "ls /tmp/get.sh"
>>>> Welcome ajian...
yf-xxx-upload05.yf01     -/home/work           yf-xxx-upload01.yf01     -/home/work
yf-xxx-upload02.yf01     -/home/work

#--Total--#  3
#---CMD---#  ls /tmp/get.sh
Continue (y/n):y
go on ...
[1/3] yf-xxx-upload05.yf01  [SUCCESS].
/tmp/get.sh

Continue (y/n):y
go on ...
[2/3] yf-xxx-upload01.yf01  [SUCCESS].
/tmp/get.sh

[3/3] yf-xxx-upload02.yf01  [SUCCESS].
/tmp/get.sh

Retry功能

-R 就是至关于第四种列表来源,当执行错误,或者ctrl+c的时候就可使用上,避免列表反复执行某些命令.

此次多加两台服务器,有两台是没有这个上面脚本文件的.因此新加的服务器会报错.

$ cat 2.list | pdo "ls /tmp/get.sh"
>>>> Welcome ajian...
yf-xxx-upload05.yf01     -/home/work           yf-xxx-upload01.yf01     -/home/work
yf-xxx-upload02.yf01     -/home/work           yf-xxx-upload03.yf01     -/home/work
yf-xxx-upload04.yf01     -/home/work

#--Total--#  5
#---CMD---#  ls /tmp/get.sh
Continue (y/n):y
go on ...
[1/5] yf-xxx-upload05.yf01  [SUCCESS].
/tmp/get.sh

Continue (y/n):y
go on ...
[2/5] yf-xxx-upload01.yf01  [SUCCESS].
/tmp/get.sh

[3/5] yf-xxx-upload02.yf01  [SUCCESS].
/tmp/get.sh

[4/5] yf-xxx-upload03.yf01  [FAILED].
ls: /tmp/get.sh: No such file or directory

[5/5] yf-xxx-upload04.yf01  [FAILED].
ls: /tmp/get.sh: No such file or directory

//使用-R 就能够直接拿到上一次执行失败的列表.
$pdo -R "ls /tmp/get.sh"
>>>> Welcome ajian...
yf-xxx-upload03.yf01     -/home/work           yf-xxx-upload04.yf01     -/home/work


#--Total--#  2
#---CMD---#  ls /tmp/get.sh
Continue (y/n):y
go on ...
[1/2] yf-xxx-upload03.yf01  [FAILED].
ls: /tmp/get.sh: No such file or directory

//若是是使用的ctrl+C中断了列表,-R会记录未执行完(包括已经执行但失败的列表)
$ cat 1.host | pdo  -T 10s "ls /tmp/get.sh"
>>>> Welcome ajian...
yf-xxx-upload05.yf01     -/home/work           yf-xxx-upload01.yf01     -/home/work
yf-xxx-upload02.yf01     -/home/work           yf-xxx-upload03.yf01     -/home/work
yf-xxx-upload04.yf01     -/home/work

#--Total--#  5
#---CMD---#  ls /tmp/get.sh
Continue (y/n):y
go on ...
[1/5] yf-xxx-upload05.yf01  [SUCCESS].
/tmp/get.sh

Continue (y/n):y
go on ...
[2/5] yf-xxx-upload01.yf01  [SUCCESS].
/tmp/get.sh

^C$ pdo -R "ls /tmp/get.sh"
>>>> Welcome ajian...
yf-xxx-upload02.yf01     -/home/work           yf-xxx-upload03.yf01     -/home/work
yf-xxx-upload04.yf01     -/home/work

#--Total--#  3
#---CMD---#  ls /tmp/get.sh
Continue (y/n):y
go on ...
[1/3] yf-xxx-upload02.yf01  [SUCCESS].
/tmp/get.sh

Continue (y/n):

### -e脚本执行功能

$ cat t.sh
#!/bin/bash

cd /tmp/ && pwd
echo "test"
touch /tmp/t.log

$ cat 1.host | pdo -e t.sh
>>>> Welcome ajian...
yf-xxx-upload05.yf01     -/home/work           yf-xxx-upload01.yf01     -/home/work
yf-xxx-upload02.yf01     -/home/work

#--Total--#  3
#---CMD---#  Script: t.sh
Continue (y/n):y
go on ...
[1/3] yf-xxx-upload05.yf01  [SUCCESS].
/tmp
test

模板功能

模板功能主要是解决重复的脚本修改动做,能够固化成一些模板,直接使用.

配置中能够本身添加模板

$ cat ~/.pdo/pdo.conf
[TEMPLATE]
container : /home/work/.pdo/template/container.sh

模板内容,这个模版主要是在一台服务器上的xxxxxx目录里面进行操做. {{.CMD}} 就是会被替换的位置.

$ cat /home/work/.pdo/template/container.sh
#!/bin/bash
grep -l "^appName:" /home/work/xxx[0-9][0-9][0-9]/xxx.conf | while read file  ; do
eval $(awk '{if($1 ~ /xxxPath/){printf "apppath=%s\n",$2};if($1 ~ /appName/){printf "appName=%s",$2}}' $file)
echo $appName
if [  -d "$apppath" ];then
        cd $apppath

            {{.CMD}}

fi
done

使用嵌入命令

$ pdo -a xxxtest -temp container "pwd"
    >>>> Welcome ajian...
    yf-xxx-app02.yf01        -/home/work/xxx001    yf-xxx-app03.yf01        -/home/work/xxx001
    yf-xxx-app00.yf01        -/home/work/xxx001    yf-xxx-app0148.yf01      -/home/work/xxx004
    dbl-xxx-app0109.dbl01    -/home/work/xxx003    m1-xxx-app0220.m1        -/home/work/xxx001
    m1-xxx-app0154.m1        -/home/work/xxx004    cq01-xxx-app0242.cq01    -/home/work/xxx003
    cq01-xxx-app0179.cq01    -/home/work/xxx001    cq02-xxx-app0258.cq02    -/home/work/xxx001
    cq02-xxx-app0287.cq02    -/home/work/xxx001    cq02-xxx-app0211.cq02    -/home/work/xxx001
    jx-xxx-app17.jx          -/home/work/xxx001    db-xxx-app17.db01        -/home/work/xxx003


    #--Total--#  14
    #---CMD---#  pwd
    Continue (y/n):y
    go on ...
    [1/14] yf-xxx-app02.yf01  [SUCCESS].
    xxxtest
    /home/work/xxx001
    jingyan
    /home/work/xxx002
    pc_anti
    /home/work/xxx003
    bakan
    /home/work/xxx004
    smallapp
    /home/work/xxx006
    appui
    /home/work/xxx008

    Continue (y/n):n
    exit ...

还能够嵌入脚本

//脚本内容
$ cat 1.sh

echo "1.sh"
pwd

//嵌入脚本使用-b
$ pdo -a xxxtest -temp container -b 1.sh
>>>> Welcome ajian...
yf-xxx-app02.yf01        -/home/work/xxx001    yf-xxx-app03.yf01        -/home/work/xxx001
yf-xxx-app00.yf01        -/home/work/xxx001    yf-xxx-app0148.yf01      -/home/work/xxx004
dbl-xxx-app0109.dbl01    -/home/work/xxx003    m1-xxx-app0220.m1        -/home/work/xxx001
m1-xxx-app0154.m1        -/home/work/xxx004    cq01-xxx-app0242.cq01    -/home/work/xxx003
cq01-xxx-app0179.cq01    -/home/work/xxx001    cq02-xxx-app0258.cq02    -/home/work/xxx001
cq02-xxx-app0287.cq02    -/home/work/xxx001    cq02-xxx-app0211.cq02    -/home/work/xxx001
jx-xxx-app17.jx          -/home/work/xxx001    db-xxx-app17.db01        -/home/work/xxx003


#--Total--#  14
#---CMD---#
Continue (y/n):y
go on ...
[1/14] yf-xxx-app02.yf01  [SUCCESS].
xxxtest
1.sh
/home/work/xxx001
jingyan
1.sh
/home/work/xxx002
pc_anti
1.sh
/home/work/xxx003
bakan
1.sh
/home/work/xxx004
smallapp
1.sh
/home/work/xxx006
appui
1.sh
/home/work/xxx008

行显示与匹配

这个功能有两种使用场景:

有点相似multi tail 能够实现同时tail多个日志,显示在一个屏幕内,并且能够对match的字符串进行高亮显示.
若是输出是单行输出,没有状态显示会显示得加的美观和可参考性.

因此这种显示方式取决于时间的前后顺序,交错输出.

拿redis的迁移过程为例子:

redis迁移至少有原来的一主一从,新主和新从.在迁移的过程当中须要同时观察四台服务器的变化.若是是每次ssh四台服务器tail 日志是很麻烦并且容易出错.

如今使用pdo命令:

//操做的主机列表1.list
    tc-yyy-redis40.tc /home/yyy/redis-shard3   //old master
    cq02-yyy-redis80.cq02 /home/yyy/redis-shard3 //new master
    yf-yyy-redis40.yf01 /home/yyy/redis-shard3 //old slave
    jx-yyy-redis80.jx /home/yyy/redis-shard3  //new slave
    第一步操做:  yf-yyy-redis40.yf01为主 --> cq02-yyy-redis80.cq02

    #命令
    #cat 1.list | pdo -r 5 -y -show row  -match "success" "tail -f log/redis.log"
    > yf-yyy-redis40.yf01      >> [11523] 06 Jan 13:56:51 * Slave ask for new-synchronization  //被要求同步
    > cq02-yyy-redis80.cq02    >> [14752] 06 Jan 13:56:58 * (non critical): Master does not understand REPLCONF listening-port: Reading from master: Connection timed out
    > yf-yyy-redis40.yf01      >> [11523] 06 Jan 13:56:58 * Slave ask for synchronization
    > yf-yyy-redis40.yf01      >> [11523] 06 Jan 13:56:58 * Starting BGSAVE for SYNC
    > yf-yyy-redis40.yf01      >> [11523] 06 Jan 13:56:58 * Background saving started by pid 22855
    > yf-yyy-redis40.yf01      >> [22855] 06 Jan 13:58:31 * DB saved on disk   //dump到磁盘
    > yf-yyy-redis40.yf01      >> [11523] 06 Jan 13:58:31 * Background saving terminated with success
    > cq02-yyy-redis80.cq02    >> [14752] 06 Jan 13:58:31 * MASTER <-> SLAVE sync: receiving 1868940396 bytes from master  //从接收到主的文件
    > cq02-yyy-redis80.cq02    >> [14752] 06 Jan 13:58:47 * MASTER <-> SLAVE sync: Loading DB in memory //将接收到的文件加载到内存
    > yf-yyy-redis40.yf01      >> [11523] 06 Jan 13:58:47 * Synchronization with slave succeeded  //文件同步成功
    > cq02-yyy-redis80.cq02    >> [14752] 06 Jan 14:01:21 # Update masterstarttime[1382324097] after loading db
    > cq02-yyy-redis80.cq02    >> [14752] 06 Jan 14:01:21 * AA: see masterstarttime: ip[10.36.114.56], port[9973], timestamp[1382324097]
    > cq02-yyy-redis80.cq02    >> [14752] 06 Jan 14:01:21 * Write aof_global_offset[92961804447] to new aof_file[46] success
    > cq02-yyy-redis80.cq02    >> [14752] 06 Jan 14:01:21 * MASTER <-> SLAVE sync: Finished with success //slave完成主从同步,说明第一步已经结束.

说明:

由于是tail -f 是不会主动退出命令,因此须要使用-y 和使用-r 来增长并发量,否则会先进行单台显示 ,而不会显示后面的.
match是匹配字符串,暂时不支持正则,会进行高亮显示.红色显示.
-show如今只支持row这一种方式,默认方式仍是原来的缓存输出方式.

如下是一个测试脚本:随机打印数字 1.sh

#!/bin/bash
    for x in `seq 1 10` ; do
        echo $x
        sleep $[ ( $RANDOM % 4 )  + 1 ]s
    done

     //可使用以下命令:
   # cat 1.list | pdo -r 5 -y -show row  -match "5" -e 1.sh

还有更多的组合哦.