rigagent / vagrant-ansible-nagios

Vagrant Ansible Nagios

This project allows to set up environment on a Vagrant box VM with a monitoring system, a web server and a simple application. System monitors certain application parameter with alerting enabled when parameter exceeds certain threshold.

Stack of technologies:

Following technologies were choosen for task implementation:

1. Ubuntu 14.04 - operation system
2. Apache2 - web server
3. Nagios3 - monitoring system
4. Bottle Py - Python framework as a simple application
5. SSL - cryptographic protocol

Preconditions:

To get started you should check that following preconditions on your local PC are met

Hardware:

"Virtualization" option is enabled in BIOS setup

Software:

Next tools should be installed:

1. Virtual Box - tool for virtualization
2. Vagrant - tool for virtualization
3. Ansible - configuration management tool for provisioning
4. Locust - load testing tool

If you don't have these tools and you have Debian or RHEL 6.x you can simple install all needed preconditions running:

sudo ./install_env_preconditions.sh

Workflow:

1. Run "install_env_preconditions.sh" script (optional)

2. Run command:

   vagrant up

3. After testing environment deployment open next links:

   • HTML page - this is a simple HTML page for Apache server testing
   • Nagios - Nagios monitoring system via SSL
   • Application - Simple Application which returns system load average per one minute. Here is implemented a reverse proxy from web server.

4. Credentials for Nagios:

   • login: nagiosadmin
   • pass: nagiosadmin

5. New plugin services added to Nagios:

   • APPLICATION - simple Application monitoring
   • LINUX STATS CPU - CPU monitoring
   • LINUX STATS DISK - Disk monitoring
   • LINUX STATS PROCESS - Processes monitoring

6. Stress testing uses Locust framework to simulate the load of clients and requests. Application Nagios plugin measures system load average per one minute and checks next ranges:

   • OK - from 0.00 to 0.05
   • WARNING - from 0.05 to 0.1
   • CRITICAL - from 0.1 and above

7. Run stress test using Locust wrapper with 100 numbers of clients and 5 rates per second in which clients are spawned by default:

   ./stress_testing.sh

8. Also you can specify number of concurrent clients and rates per second:

   ./stress_testing.sh [CLIENTS] [HATCH-RATE]

9. Please keep in mind that Nagios plagin updates every few minutes so there may be delays