Vous êtes sur la page 1sur 13

Observer

Framework for monitoring production environment via functional tests


Tomasz Stelmach 2015

All that is too complex is unnecessary, and


it is simple that is needed
Mikhail Kalashnikov

About this presentation

1. No automated monitoring framework


2. Lack of central dashboard
3. Low coverage
4. Performance monitoring

Problem

Solution

Summary

1. Easy to start with implementation


2. Ready to use plugins
3. Easy to integrate
4. Open source with community
5. Alive tool
6. Cost?

Problem

Solution

Summary

Problem

Solution

Summary

Problem

Solution

Summary

Problem

Solution

Summary

define host{
use windows-server
host_name CXL-Prod
alias CXL prod
address 10.8.0.59
initial_state u
max_check_attempts 6
check_interval 6
check_period 24x7
notification_period 24x7
check_command check_http
hostgroups production-servers
}
define service{
use generic-service
host_name CXL-Prod
check_command
check_java!CxlQuantumCheck.CxlQuantumCheckFromIntegrationTables!threshold=30,config=CxlQuantumCheck_Prod.properties!Prod_CXL_Quantum
service_description Performance: CXL-Quantum interface
check_interval
10
notifications_enabled
1
contact_groups
cxl-admins
}
define service{
use generic-service
host_name CXL-Prod
check_command check_java!CxlCheck.CxlCheckLogin!endpoint=http://webapps.ch.glencore.com!Prod_CXL login check
service_description Functional: Login to application
check_interval
20
notifications_enabled
1
contact_groups
cxl-admins
}

Problem

Solution

Summary

Problem

Solution

Summary

Advantages

Disadvantages

Flexibility and simplicity in implementation

Poor authentication mechanism in Nagios Web


Server (via httpasswd)

Easy integration with other tools

Nagwin can call every command via


shell

Http integration via NRDP

Limited number of standard plugins for windows


platform. Most of plugins are for engineering
and Unix environment (host network
management)

Passive and Active checks

Failures have to be investigated manually

Notifications via email about critic state of


service or host
Limited development needed:

java plugin (for calling Selenium,


SOAPUI or other tool)

php page for presenting results


Highly customizable inheritance, templates,
config files etc.
Big community around Nagios

Problem

Solution

Summary

Observer is a monitoring framework for performance and functional tests


which aims to monitor environment from functional point of view
Observer is not another testing tool so doesnt aim to find defects
Production environment tests have to be simple, repeatable and
nonintrusive
Set of tools based on Nagwin (Nagios distribution) which gives possibility to
monitor applications from functional or performance point of view (sample:
Login to application or How fast to login)
Completely web based for end user
Easy to integrate and customize

Problem

Solution

Summary

Time for Q&A

Problem

Solution

Summary

Vous aimerez peut-être aussi