sar -A -t -f /tmp/sa11 >/tmp/sar11

https://github.com/ryran/rsar

When dealing with sysstat sar data in a sosreport, it's almost always easier to parse through the plain-text sar data files than it is to get the binary sa files onto a system where you can use sar to parse through them for exactly what you're looking for.

The goal behind rsar is to make this process a little bit easier. rsar is like the sar command, but for plain-text sar files instead of sa files. It supports almost all of the same data-selection options sar uses.

INSTALLATION

Two choices:

  1. Manual install: Get the very latest version directly by downloading the main rsar file from raw.github.com/ryran/rsar/master/rsar. Updating will need to be done manually by re-downloading it. Explicitly, you could run the following as root:

    curl -Lo /usr/local/bin/rsar raw.github.com/ryran/rsar/master/rsar
    
chmod +x /usr/local/bin/rsar

  2. RPM/Yum repo (recommended): Configure access to the Fedora/RHEL6+ yum repository @ people.redhat.com/rsawhill/rpms and install the rsar RPM from there. Updating can happen automatically along with the rest of your system. Two-step instructions @ How to install xsos or rsar via yum?

Once rsar is installed by one of the above methods, run rsar -h as a normal user to see the help page and get started.

USAGE

rsar has an extensive help page. Use it. Feel free to make it better. Here's what it looks like in v0.1.0rc6:

Usage: rsar SARFILE... [--12hr] [--hn] [--noxh] [-t TIME] [SAROPTS] [DRILLDOWN]

 SAROPTS is any mix of: [-AbBcdHInqrRSuvwWy]

 DRILLDOWN is any mix of: [-P CPU] [-D DISK] [-N NIC]

 Note that getopt is used for arg-parsing (so argument order is arbitrary)

rsar requires at least 1 plaintext sar data file (optionally-compressed)

  Spaces in SARFILE name or path will break things (complain if you care)

  Supported file extensions for compressed sar files:

    z, gz, gzip, b, b2, bz, bz2, bzip, bzip2, x, xz, lzma

Opts Specific to rsar:

  --12hr         Switch input format from default 24-hr to 12-hr AM/PM

  --hn           Show hostname/kernel version lines from each sar file

  --noxh         Hide extra column header lines (usually indicative of reboots)

  -t <TIME>      Where TIME is a regex for time period to display

                   Ex: '^1[2-5]' or ':[23]0:' or '^(0.:40|2.:40)' or '^12.*AM'

  -x <PATTERN>   Where PATTERN is used to select a section

                   Ex: 'svctm', 'iowait', 'runq', 'proc|cswch'

                   Recommended to stick with below sar opts to avoid collisions

  -z             Disable showing 'Average:' lines

                   Hint: To show _only_ 'Average:' lines, use -t Av

Standard sar Selection Opts:

  -A { 5 | 6 | 7 }

          Like sar -A: alias for all selection opts (excl. -d & -n)

            The required number corresponds with a major RHEL version

  -b      I/O and transfer rate statistics

  -B      Paging statistics

  -c      Process creation statistics

  -d      Block device statistics

  -H      Hugepages utilization statistics

  -I { SUM }

          Interrupts statistics

  -n { <KEYWORD> [,...] | ALL }

          Network statistics

            KEYWORD may be any of the following (case-insensitive):

            DEV    Network interfaces

            EDEV   Network interfaces (errors)

            NFS    NFS client

            NFSD   NFS server

            SOCK   Sockets      (v4)

            IP     IP traffic   (v4)

            EIP    IP traffic   (v4) (errors)

            ICMP   ICMP traffic (v4)

            EICMP  ICMP traffic (v4) (errors)

            TCP    TCP traffic  (v4)

            ETCP   TCP traffic  (v4) (errors)

            UDP    UDP traffic  (v4)

            SOCK6  Sockets      (v6)

            IP6    IP traffic   (v6)

            EIP6   IP traffic   (v6) (errors)

            ICMP6  ICMP traffic (v6)

            EICMP6 ICMP traffic (v6) (errors)

            UDP6   UDP traffic  (v6)

  -q      Queue length and load average statistics

  -r      Memory utilization statistics

  -R      Memory statistics

  -S      Swap space utilization statistics

  -u      CPU utilization statistics

  -v      Kernel table statistics

  -w      System switching statistics

  -W      Swapping statistics

  -y      TTY device statistics

  Note that args passed to -I & -n & -P are cumulative and case-insensitive

    Ex: -n Dev,eDEV works, as does -n Dev -n eDEV -n icmp as does -nALL

        -P all works, as does -P 0,4,7, as does -P0 -P 4 -P 3. -P '[78]|22'

Drill-Down Opts:

  -P { <CPU> [,...] | ALL }

              Where CPU can be comma-separated list of CPU-numbers (like sar)

                Else, if 'ALL' is used, then everything will be shown

                Unlike with sar, CPU can also be a regex

                Ex: '1,4,6' or '1|4|6' or '0,[1-3]0'

  -D <DISK>   Where DISK is a regex for which block devices to select

                Ex: 'dev8-16' or '7.$' or 'sda2|rootlv'

  -N <NIC>    Where NIC is a regex for which network interfaces to select

                Ex: 'eth0' or 'eth[12]' or 'bond[1-3]|em1'

Examples:

  rsar sar03 -qut ^2

  rsar sar03 -P all -t Average

  rsar -t^09:30 -n edev,sock -N eth3 -I sum sar19.bz2

  rsar -x iowait -P [0-3] -t^15:30 /tmp/sar26.gz /tmp/sar27.xz -RB

Version info: rsar v0.1.0rc6 last mod 2013/02/16

See <github.com/ryran/xsos> to report bugs or suggestions

EXAMPLES IN ACTION

rsar doesn't read binary files; it's meant to read sar-generated plain-text sar files. Reading a single file with no options, rsar behaves the same as sar would, i.e.: shows the combined cpu-utilization.

[rsaw:sa]$ rsar sa22

rsar: Unable to read file: 'sa22'

Need a plain-text sar data file to continue

[rsaw:sa]$ rsar sar22

00:00:01          CPU     %user     %nice   %system   %iowait    %steal     %idle

00:10:01          all      0.10      0.00      0.08      0.00      0.00     99.82

00:20:01          all      0.09      0.00      0.08      0.00      0.00     99.82

00:30:01          all      0.10      0.00      0.08      0.00      0.00     99.81

00:40:01          all      0.11      0.00      0.09      0.01      0.00     99.80

00:50:01          all      0.10      0.00      0.09      0.01      0.00     99.81

01:00:01          all      0.10      0.00      0.09      0.00      0.00     99.80

01:10:01          all      0.11      0.00      0.11      0.00      0.00     99.78

01:20:02          all      0.09      0.00      0.09      0.00      0.00     99.82

01:30:01          all      0.10      0.00      0.08      0.00      0.00     99.82

01:40:01          all      0.11      0.00      0.08      0.00      0.00     99.80

01:50:01          all      0.09      0.00      0.08      0.00      0.00     99.83

02:00:01          all      0.10      0.00      0.08      0.00      0.00     99.82

02:10:01          all      0.10      0.00      0.08      0.00      0.00     99.82

(Note: Output truncated. That would have shown all available cpu utilization data, including any new header-lines caused by reboots and plus any "Average" lines.)

Time-selection is NOT like the -e and -s options of sar; instead, rsar has a -t option that takes a regular expression which is matched against the time field in the sar data. Be as specific or as general as you want.

[rsaw:sa]$ rsar sar22 -t ^21

00:00:01          CPU     %user     %nice   %system   %iowait    %steal     %idle

21:00:01          all      0.10      0.00      0.08      0.00      0.00     99.82

21:10:01          all      0.10      0.00      0.08      0.00      0.00     99.82

21:20:01          all      0.09      0.00      0.08      0.00      0.00     99.83

21:30:01          all      0.10      0.00      0.08      0.00      0.00     99.82

21:40:01          all      0.10      0.00      0.08      0.00      0.00     99.82

21:50:01          all      0.09      0.00      0.08      0.00      0.00     99.83

Average:          all      0.10      0.00      0.08      0.00      0.00     99.82

rsar supports many sar-specific data-selection options like -q for queue/loadavg or -r for memory utilization. So here's that, plus a very specific time selection.

[rsaw:sa]$ rsar -qt '^12:10|16:40' sar22 -r

00:00:01      runq-sz  plist-sz   ldavg-1   ldavg-5  ldavg-15

12:10:01            0      1221      1.14      1.14      1.12

16:40:01            2      1228      1.22      1.25      1.20

Average:            0      1221      1.21      1.21      1.18

00:00:01    kbmemfree kbmemused  %memused kbbuffers  kbcached kbswpfree kbswpused  %swpused  kbswpcad

12:10:01     77117704  54967744     41.62    502620  47472612  33551744         0      0.00         0

16:40:01     77067032  55018416     41.65    502904  47516168  33551744         0      0.00         0

Average:     77145109  54940339     41.59    502192  47449826  33551744         0      0.00         0

For those unfamiliar with sar options, regex searching for headers with -x is supported. You can also use -P in the same way as with the sar command, namely, to specify which CPUs to display info for (comma-separated numbers or else ALL to see each CPU + combined avg). With this example output, we can see how easy rsar makes it to spot in the data when a system was rebooted -- no scrolling halfway through the file necessary.

[rsaw:sa]$ rsar sar22 -t'^1(3:5|4:[0-5])' -x nice -P 0,9

00:00:01          CPU     %user     %nice   %system   %iowait    %steal     %idle

13:50:01            0      0.05      0.00      0.12      0.00      0.00     99.82

13:50:01            9      0.02      0.00      0.01      0.00      0.00     99.97

14:00:01            0      5.38      0.00      0.22      0.00      0.00     94.40

14:00:01            9      0.05      0.00      0.10      0.00      0.00     99.84

14:10:01            0      0.04      0.00      0.05      0.00      0.00     99.91

14:10:01            9      0.01      0.00      0.02      0.00      0.00     99.97

14:20:01            0      0.04      0.00      0.13      0.00      0.00     99.83

14:20:01            9      0.01      0.00      0.00      0.00      0.00     99.98

Average:            0      0.13      0.00      0.07      0.00      0.00     99.80

Average:            9      0.02      0.00      0.05      0.00      0.00     99.92

14:40:01          CPU     %user     %nice   %system   %iowait    %steal     %idle

14:50:01            0      0.22      0.00      0.19      0.01      0.00     99.58

14:50:01            9      0.03      0.00      0.01      0.00      0.00     99.96

Average:            0      0.08      0.00      0.08      0.00      0.00     99.84

Average:            9      0.03      0.00      0.05      0.00      0.00     99.91

[rsaw:sa]$ grep -i restart sar22

14:31:23          LINUX RESTART

Paging & page info:

[rsaw:sa]$ rsar sar22 -t2:30 -BR

00:00:01     pgpgin/s pgpgout/s   fault/s  majflt/s

02:30:01         0.59    340.28  13167.06      0.00

12:30:01         0.11    339.28  13130.51      0.00

22:30:01         0.11    350.35  13141.76      0.00

Average:        53.62    379.68  13026.00      0.01

00:00:01      frmpg/s   bufpg/s   campg/s

02:30:01         0.78      0.01     -0.05

12:30:01        -2.04      0.00      0.19

22:30:01        -1.42      0.00      0.19

Average:        -0.80      0.01      0.70

Interface-specific selection (-N) of network stats from multiple compressed files:

[rsaw:sa]$ rsar -n DEV -N eth4 sar24.gz -t^19:[12] sar26.bz2

------------------------ sar24.gz ------------------------

00:00:01        IFACE   rxpck/s   txpck/s   rxbyt/s   txbyt/s   rxcmp/s   txcmp/s  rxmcst/s

19:10:01         eth4      2.42      2.56    203.47    225.20      0.00      0.00      0.00

19:20:01         eth4      2.41      2.57    210.17    224.31      0.00      0.00      0.00

Average:         eth4      2.42      2.56    205.99    224.09      0.00      0.00      0.00

------------------------ sar26.bz2 ------------------------

00:00:01        IFACE   rxpck/s   txpck/s   rxbyt/s   txbyt/s   rxcmp/s   txcmp/s  rxmcst/s

19:10:01         eth4      2.43      2.55    204.02    224.23      0.00      0.00      0.00

19:20:01         eth4      2.41      2.57    205.94    225.02      0.00      0.00      0.00

Average:         eth4      2.42      2.56    206.07    224.04      0.00      0.00      0.00

Disk stats via regex (could also use the sar option: -d).

[rsaw:sa]$ rsar -x await sar19 -t^04:[0-4]

00:00:01          DEV       tps  rd_sec/s  wr_sec/s  avgrq-sz  avgqu-sz     await     svctm     %util

04:00:01       dev8-0      0.14      0.00      1.79     12.92      0.00      1.37      0.81      0.01

04:00:01      dev8-16      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

04:10:01       dev8-0      0.22      6.28      2.28     39.51      0.00      1.74      1.04      0.02

04:10:01      dev8-16      0.14      0.00      2.07     15.33      0.00      2.83      0.88      0.01

04:20:01       dev8-0      0.19      0.00      2.40     12.74      0.00      1.78      0.80      0.02

04:20:01      dev8-16      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

04:30:01       dev8-0      0.19      0.00      2.35     12.46      0.00      1.42      0.65      0.01

04:30:01      dev8-16      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

04:40:01       dev8-0      0.19      0.00      2.48     13.05      0.00      1.61      0.75      0.01

04:40:01      dev8-16      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

Average:       dev8-0      0.23      1.00      2.37     14.56      0.00      1.60      0.88      0.02

Average:      dev8-16      0.07      0.01     14.89    218.58      0.00      5.12      1.44      0.01

And just like with network interfaces, there's an option to specify specific block devices. Here I did that (-D) along with the -z option to hide the "Average" lines, plus threw in network errs (-n edev).

[rsaw:sa]$ rsar -x await sar19 -t^04:[0-4] -D 16 -zn edev -N eth0

00:00:01          DEV       tps  rd_sec/s  wr_sec/s  avgrq-sz  avgqu-sz     await     svctm     %util

04:00:01      dev8-16      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

04:10:01      dev8-16      0.14      0.00      2.07     15.33      0.00      2.83      0.88      0.01

04:20:01      dev8-16      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

04:30:01      dev8-16      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

04:40:01      dev8-16      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

00:00:01        IFACE   rxerr/s   txerr/s    coll/s  rxdrop/s  txdrop/s  txcarr/s  rxfram/s  rxfifo/s  txfifo/s

04:00:01         eth0      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

04:10:01         eth0      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

04:20:01         eth0      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

04:30:01         eth0      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

04:40:01         eth0      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

AUTHORS

So far it's just me, ryran, aka rsaw, aka Ryan Sawhill Aroha

Please contact me if you have ideas, suggestions, questions, or want to collaborate on this or something similar. For specific bugs and feature-requests, you can post a new issue on the tracker.

LICENSE

Copyright (C) 2012-2013, 2015 Ryan Sawhill Aroha

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License @gnu.org/licenses/gpl.html for more details.

[转帖]rsar - Extract data from plain-text sar files的更多相关文章

  1. [转帖]浅谈程序中的text段、data段和bss段

    作者:百问科技链接:https://zhuanlan.zhihu.com/p/28659560来源:知乎著作权归作者所有.商业转载请联系作者获得授权,非商业转载请注明出处. 一般情况,一个程序本质上都 ...

  2. Insert Plain Text and Images into RichTextBox at Runtime

    Insert Plain Text and Images into RichTextBox at Runtime' https://www.codeproject.com/Articles/4544/ ...

  3. Plain text considered harmful: A cross-domain exploit

    referer:http://balpha.de/2013/02/plain-text-considered-harmful-a-cross-domain-exploit/ Data from aro ...

  4. 如何撤销 PhpStorm/Clion 等 JetBrains 产品的 “Mark as Plain Text” 操作 ?

    当把某个文件“Mark as Plain Text”时,该文件被当做普通文本,就不会有“代码自动完成提示”功能,如下图所示: 但是呢,右键菜单中貌似没有 相应的撤销 操作, 即使是把它删除,再新建一个 ...

  5. Jade之Plain Text

    Plain Text jade提供了3种得到纯文本的方法. Piped Text 添加纯文本的一个最简单的方法就是在文本最前面加|符号即可. jade: p | It must always be o ...

  6. [Regular Expressions] Find Plain Text Patterns

    The simplest use of Regular Expressions is to find a plain text pattern. In this lesson we'll look a ...

  7. [转] bss段、data段、text段

    1.前言 一个程序本质上都是由 BSS 段.DATA段.TEXT段三个组成的. 本文主要分编译时和运行时分别对 对data段 bss段 text段 堆 栈作一简要说明 2. 程序编译时概念说明 程序与 ...

  8. How to return plain text from AWS Lambda & API Gateway

    With limited experience in AWS Lambda & API Gateway, it's struggling to find the correct way to ...

  9. (深入理解计算机系统) bss段，data段、text段、堆(heap)和栈(stack)

    bss段: bss段(bss segment)通常是指用来存放程序中未初始化的全局变量的一块内存区域. bss是英文Block Started by Symbol的简称. bss段属于静态内存分配. ...

  10. 【转】(深入理解计算机系统) bss段，data段、text段、堆(heap)和栈(stack)

    bss段: bss段(bss segment)通常是指用来存放程序中未初始化的全局变量的一块内存区域. bss是英文Block Started by Symbol的简称. bss段属于静态内存分配. ...

随机推荐

  1. 小熊派开发实践丨小熊派+合宙Cat.1接入云服务器

    摘要:使用小熊派开发板,以合宙的AIR724为通信模组(Cat.1),以AT指令方式,通过mqtt协议接入云服务器. 本贴使用小熊派开发板+合宙的Air724(Cat.1模组),接入自己搭建的EMQ服 ...

  2. 华为云原生数据仓库GaussDB(DWS)深度技术解读：融、快、大、稳、易

    摘要:云原生数据仓库GaussDB(DWS)架构师应邀为大家解读数仓深度技术. "云原生"在2020年成为备受瞩目的热词,云原生在确保企业数字化转型中资源快速供给.按需使用的同时, ...

  3. 复杂查询so easy ，GaussDB(for Cassandra)推Lucene引擎全新解决方案

    摘要:复杂查询so easy!GaussDB(for Cassandra)新特性来袭. 本文分享自华为云社区<来了!GaussDB(for Cassandra)新特性亮相>,作者:Gaus ...

  4. 结MySQL 的一些知识点：MySQL 安装

    MySQL 安装 所有平台的 MySQL 下载地址为: MySQL 下载 . 挑选你需要的 MySQL Community Server 版本及对应的平台. **注意:**安装过程我们需要通过开启管理 ...

  5. 实用教程｜手把手带你离线部署Walrus

    Walrus 0.4 已于近日发布,新版本中采用的应用模型可以让运维团队仅需配置1次,即可在多模态的基础设施及环境中运行包括应用服务及周边依赖资源在内的全套应用系统.这极大减少了运维人员的工作量,同时 ...

  6. 差点错过！火山引擎VeDI帮这家企业成功挖掘200余条商机

    更多技术交流.求职机会,欢迎关注字节跳动数据平台微信公众号,回复[1]进入官方交流群 与个体消费市场临时性需求大.决策参与人少等情况不同,企业消费市场往往因为长线需求复杂.商品/服务的价格高.参与决策 ...

  7. PPT 如何将大段WORD制作成PPT

    PPT 如何将大段WORD制作成PPT 标记等级 发送到PPT 图片.表格,无法发送

  8. 【已解决】无法打开.xlsx文件，xlrd.biffh.XLRDError: Excel xlsx file； not supported

    原因是最近xlrd更新到了2.0.1版本,只支持.xls文件.所以x1 = xlrd.open_workbook("data.xlsx")会报错. 可以安装旧版xlrd,在cmd中 ...

  9. 收到一封CTO来信，邀约面试机器学习工程师

    大家好,我是北海 很少登陆 Gmail,前天收验证码登了一下,发现居然收到一封某初创公司CTO的来信. 我在Github上看到了您的资料觉得很有意思,请问您是否考虑我们公司的全职工作机会呢?可供考虑的 ...

  10. HanLP — HMM隐马尔可夫模型 -- 维特比(Viterbi)算法 --示例代码 - Java

    Viterbi 维特比算法解决的是篱笆型的图的最短路径问题,图的节点按列组织,每列的节点数量可以不一样,每一列的节点只能和相邻列的节点相连,不能跨列相连,节点之间有着不同的距离,距离的值就不在 题目背 ...