작성자: 주인장 디지문
(http://www.digimoon.net/)
Python Nmon Analyzer의 설치 및 사용 방법입니다.
RHEL 6 & 7 기준으로 설명 드립니다
1. Python Nmon Analyzer 실행에 필요한 사전 환경 구성
1.1. gcc-c++ 설치
# yum install gcc-c++
1.2. 파이썬 도구 설치 - yum 또는 소스 직접 설치 방식 중 택일합니다. 소스 설치 방법은 외부인터넷 연결이 되어야 하므로 yum 권장합니다.
1.2.1. yum으로 설치
① EPEL 레포지토리 구성
( https://fedoraproject.org/
RHEL6
# rpm –Uvh https://dl.fedoraproject.org/
RHEL7
# rpm -Uvh https://dl.fedoraproject.org/
② 설치
# yum install python-pip python-matplotlib
python-pip는 EPEL, python-matplotlib은 base repo에서 받아 설치됩니다.
외부 인터넷 안 되는 시스템에서는 첨부한 rpm을 직접 설치하세요.
http://www.digimoon.net/rpms/python-pip-7.1.0-1.el6.noarch.
http://www.digimoon.net/rpms/python-pip-7.1.0-1.el7.noarch.
1.2.2. 소스 설치
pip 소스 다운로드 및 설치
https://pip.pypa.io/en/latest/
https://bootstrap.pypa.io/get-
# python ./get-pip.py
pip 도구로 matplotlib 설치
# pip install matplotlib
2. Python Nmon Analyzer 다운로드 및 사용
GitHub에서 다운로드
https://github.com/madmaze/
pyNmonAnalyzer-master.zip 다운로드
다운로드하여 압축 풀고 확장자 py인 파일들의 퍼미션을 700으로 부여
# chmod 700 *.py
간단 사용법
# ./pyNmonAnalyzer.py -b -o testReport-localhost_160426_0000 -i localhost_160426_0000.nmon
아래는 README.md 내용입니다.
| pyNmonAnalyzer ======== A tool for parsing and reshuffeling nmon's output into "normal" csv format. Nmon puts out a long file with a system-info header at the beginning, followed by a continuous stream of time stamped readings. This format makes it difficult for analysis by standard Spreadsheet viewers without a fair amount of preprocessing. The pyNmonAnalyzer aims to make this simpler, faster and more effective. In one sweep the pyNmonAnalyzer creates CSV files and two HTML-based reports, one with static graphs and one with interactive graphs powered by [dygraphs](http://dygraphs.com). This project is currently a work-in-progress and therefore will hopefully improve in functionality and usability. If you have questions, find a bug or have feature requests, please do file an issues [here](https://github.com/madmaze/pyNmonAnalyzer/issues) **Example Reports:** --------------- - [Static Report](http://matthiaslee.com/scratch/pyNmonAnalyzer/data/report.html) - [Interactive Report](http://matthiaslee.com/scratch/pyNmonAnalyzer/interactiveReport.html) Goals: ----- - make nmon log file analysis easier and faster - create HTML based reports with embedded graphs - create CSV files for more indepth data analysis - interactive graphs for hands on analysis, perhaps using dygraph.js Getting started: ----- - [Python Nmon Analyzer: moving away from excel macros](http://matthiaslee.com/?q=node/38) Installation: ----- pyNmonAnalyzer is now available through pip and easy_install. If you have pip: ```$> sudo pip install pyNmonAnalyzer``` If you'd like to mess with the source, please feel free to fork this github repo and contribute back changes. Dependencies: ----- This tool depends on the python numpy package and the matplotlib package. * If you are on a Debian/Ubuntu based system: `sudo apt-get install python-numpy python-matplotlib` * If you are on a RHEL/Fedora/Centos system: `sudo yum install python-numpy python-matplotlib` Usage: ----- ``` usage: pyNmonAnalyzer [-h] [-x] [-d] [--force] [-i INPUT_FILE] [-o OUTDIR] [-c] [-b] [-t REPORTTYPE] [-r CONFFNAME] [--dygraphLocation DYGRAPHLOC] [--defaultConfig] [-l LOGLEVEL] nmonParser converts NMON monitor files into time-sorted CSV/Spreadsheets for easier analysis, without the use of the MS Excel Macro. Also included is an option to build an HTML report with graphs, which is configured through report.config. optional arguments: -h, --help show this help message and exit -x, --overwrite overwrite existing results (Default: False) -d, --debug debug? (Default: False) --force force using of config (Default: False) -i INPUT_FILE, --inputfile INPUT_FILE Input NMON file -o OUTDIR, --output OUTDIR Output dir for CSV (Default: ./report/) -c, --csv CSV output? (Default: False) -b, --buildReport report output? (Default: False) -t REPORTTYPE, --reportType REPORTTYPE Should we be generating a "static" or "interactive" report (Default: interactive) -r CONFFNAME, --reportConfig CONFFNAME Report config file, if none exists: we will write the default config file out (Default: ./report.config) --dygraphLocation DYGRAPHLOC Specify local or remote location of dygraphs library. This only applies to the interactive report. (Default: http://dygraphs.com/dygraph-dev.js) --defaultConfig Write out a default config file -l LOGLEVEL, --log LOGLEVEL Logging verbosity, use DEBUG for more output and showing graphs (Default: INFO) ``` Example Usage: ------------- First generate a report config, most likely the default is all you need. This creates ./report.config ```$> pyNmonAnalyzer --defaultConfig``` Build HTML report with *interactive* graphs for test.nmon and store results to testReport ```$> pyNmonAnalyzer -b -o testReport -i test.nmon``` Build HTML report with static graphs for test.nmon and store results to testReport ```$> pyNmonAnalyzer -b -t static -o testReport -i test.nmon``` Compile CSV formatted tables for data in test.nmon and store results to testOut ```$> pyNmonAnalyzer -c -o testOut -i test.nmon``` Configuration: ------------- To control which items get graphed(CPU, MEM etc) you need to configure the report.config file. This is especially important for AIX NMON systems. To get a sense of what the config file should look like, run `pyNmonAnalyzer --defaultConfig` this will generate "report.config" in your local directory. It contians two examples, one for Linux and one for AIX systems. Adjust them according to your device names, for Linux you'll want to set DISKBUSY to your sda1 or sdb1 or what ever. You should be able to use **any** nmon performance stats, so DISKBUSY, DISKREAD, CPU1, CPU2 etc. Troubleshooting: --------------- - **It crashes or my graphs don't show anything!** Have you looked at your current report.config? Is it customized to your device names? - **My interactive report will not display! What did I do wrong?** Since the interactive report uses javascript to load CSV files, your browser needs to be allowed to read local files(if you are viewing locally). Firefox has been th e most reliable for me, chrome currently does not allow JS to access local files. **Any other problems, file an issue or send me an email.** Configuration: ------------- To control which items get graphed(CPU, MEM etc) you need to configure the report.config file. This is especially important for AIX NMON systems. To get a sense of what the config file should look like, run `pyNmonAnalyzer --defaultConfig` this will generate "report.config" in your local directory. It contians two examples, one for Linux and one for AIX systems. Adjust them according to your device names, for Linux you'll want to set DISKBUSY to your sda1 or sdb1 or what ever. You should be able to use **any** nmon performance stats, so DISKBUSY, DISKREAD, CPU1, CPU2 etc. Troubleshooting: --------------- - **It crashes or my graphs don't show anything!** Have you looked at your current report.config? Is it customized to your device names? - **My interactive report will not display! What did I do wrong?** Since the interactive report uses javascript to load CSV files, your browser needs to be allowed to read local files(if you are viewing locally). Firefox has been th e most reliable for me, chrome currently does not allow JS to access local files. **Any other problems, file an issue or send me an email.** License: ------- ``` Copyright (c) 2012-2015 Matthias Lee, matthias.a.lee[]gmail.com Last edited: Feburary 25th 2015 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 for more details. You should have received a copy of the GNU General Public License along with this program. If not, see <http://www.gnu.org/licenses/>. |
