Following the last post here is a set of scripts to monitor linux generic performance counters as well as MySQL.

The data that will be gathered will be done according to your request. Data from CPU and / or memory and / or disk io and / or MySQL. The data generated will be dispatched to different csv files according to your configuration choices (by hour, day, week, month). This is mainly to ease the process when you will be in the process comparing a specific time frame with what people are saying abnormal activity against what was considered normal activity from historic baseline.

Script installation step:

• make sure that all the prerequisites and requirements are fulfilled (prerequisites document)
  

• extract all the script files to a folder of your choice. make sure to keep the structure of it as specified in the prerequisites document (baseline-installation-scripts.rar)
  

• define personal script parameters, within baseline.sh file between the tags “BASELINE SCRIPT SETUP PARAMETERS ”
  

• create a crontab item according to what you need to baseline and the frequency that fits your requirements.

• finally, see the data coming ...

Then what ?:

This of course provide the baseline gathering process. Another good thing to do will be of course to find a way to quickly perform analysis of the data and find ways to extract, efficiently, meaningful information from it. To start I will direct you to the web. You will find may project to “graph” csv , others to split the data into quartiles and so forth.

Future step :

This post is part of a series that I will be doing on the baseline subject. The next phase will be to add to the baseline scripts the capability to handle Oracle and mongoDB databases specific counters.

Voilà, I hope you found the information here usefull and beneficial, enjoy. Please leave comment and suggestions , I will be more than glad to ear what you think.