Break-2964 Open Source Monitoring- So Many Options!

Break-2964

Open Source Monitoring- So Many Options!

Nick Rowlett – Sparta Area School District

Kevin Capwell – School District of Onalaska

So you want to monitor your network…• Know about an event before the phone rings• Keep tabs on critical processes• Graph important metrics• Visualize trends• Keep you in control of your environment

What’s that? No money?

Commercial• Supported• Well-developed interface• Get started quickly• Costs Money – possibly A LOT

Open Source (or otherwise Free)• Community-driven• Time-tested• Outdated programming• Ongoing time-sink

What to know

• Most open source monitoring systems run on Linux• Some are available as a pre-configured VM • Each one looks GREAT on the product website• You will probably spend way too much time configuring it• Most of the open systems are a gateway to a paid system

What features are important?

• What OS does it run on? • Does it come as a preconfigured VM?• Dashboard• Agentless host monitoring (and for which host OS’s)• Ease of configuration• Pretty interface / graphs / charts• LDAP or AD integrated login• Device auto-discovery• Service monitoring, hardware monitoring

The Contenders

Open Source Kinda Free

Nagios• Defacto standard?

• Difficult setup (until you get used to it)

• Text-based config files

• Rock-solid back end

• No dashboard

• Variety of front-ends available

• Can monitor almost anything

• http://exchange.nagios.org

• Performance graphing not native but can integrate

Icinga• Nagios fork

• https://wiki.icinga.org

• Slick AJAX interface

• Still need to know text-based config

• Comes as a VM

• Graphing & mapping are not native functions

• Integrated mobile interface

• Custom, shareable dashboards

OpenNMS• Automated discovery

• Event management w/ traps

• Mobile client available

• Runs on Linux/OSX

• SNMP traps, Syslog

• Integrated performance graphing

• RSS integration

• Interface is in Java (AJAX dashboard)

Zenoss• Installs on Linux

• Integrated graphing & mapping

• Agentless monitoring

• Configuration workflow is unintuitive (until you get used to it)

• Community-submitted configuration plugins (ZenPacks)

• Configurable dashboard

• Hardware monitoring via ZenPacks

• Resource hog

Zabbix• Monitors everything including web

services

• Performance graphing

• Hardware monitoring

• Event management

• No agentless Windows monitoring via WMI

• Autodiscovery of devices

• Custom ‘screens’ and ‘slideshows’ for easy visual monitoring

Cacti• Arguably the prettiest graphs of any

system

• Individual graphs can be used in other web apps or dashboards

• Excellent performance monitor

• Not a great hardware / system monitor

• Runs on Linux or Windows w/ PHP

Opsview• Built on Nagios 4 Core

• Application, network, virtualiziation monitoring

• Mobile app

• Dashboard, Autodiscovery, Reports, SNMP traps all require Commercial subscription

Spiceworks• All-in-one helpdesk, inventory,

purchasing, monitoring & management

• Crazy easy to setup

• Runs on Windows

• Not very granular

• Can graph performance data, but not too detailed

• Uses targeted IT-specific ads

• Great for quick/easy

• Not great for large/granular

Hyperic• VMware product

• Optimized for vSphere

• Application & Systems monitoring

• Autodiscovery

• Nagios integration

• Agent-based

• Reports, dashboards, remote agent install, etc. require Enteprise subscription

PRTG• Freeware edition limited to 10 sensors

• Smartphone apps available

• Beautiful graphs

• Easy setup

• autodiscovery

• Dashboards & maps

• Priced per ‘sensor’ – can get expensive

• http://www.paessler.com/tools for free SNMP, MIB, NetFlow, WMI, etc. testing tools

How to get started

• Download this presentation from http://www.brainstormk20.com/• Clear your calendar• Lock your door, phone on DND• Caffeine• Use Pre-configured VMs to demo• Use a VM snapshot or template of CentOS 6

Thanks for coming!