Setting up A Network Monitoring and Inventory System

From Alpine Linux
Revision as of 19:39, 8 September 2010 by Jbilyk (talk | contribs) (one last change to the linux host-detect script so that it's input is subnet(s) to scan and output is resolvable hostnames)


This material is work-in-progress ...

Do not follow instructions here until this notice is removed.
(Last edited by Jbilyk on 8 Sep 2010.)

This guide will set up (list subject to change):

This guide will assume:

  • You have a knowledge of your network setup (at least know which subnets exist)
  • Your network allows ICMP ping packets to every device that you want discovered and automatically added to the system
  • Your firewalls (whether on your networking devices or end-user devices) allow traffic back to the server that is being set up, and if it is a Windows host, allows WMI queries in
  • If you are auditing a Windows domain, you have access to an account that has local administrator privileges on every computer (on Win XP a user with Power User privileges may be enough)

Goal

Basic concept is to combine the above apps/utilities to create an easy to set up and low upkeep monitoring system that also tracks inventory thanks to OpenAudit. Redundant entries in configs will be kept to a minimum (hopefully). New devices should be auto-added to monitoring systems, along with associated services (for example, a Windows server running MS Exchange might have basic services such as CPU load, HD free space, running services and Exchange queues monitored, whereas a router running on Alpine Linux would might need the routing table, throughput, and opennhrp monitored). Also, software licensing should be monitored so that Windows boxes do not exceed purchased licenses.

Progress Log

Set up Alpine server with Lighttpd 1.4.28-r0, PHP 5.3.3-r2 and MySQL 5.1.50-r0 with OpenAudit 09.12.23. Windows box can be audited with results returned to Alpine.
Set up test cacti system and made sure the basics work with the new aport.
SNMPTT receives snmp traps properly and sends them on to Nagios but they aren't parsed in any way yet.
Installed Nagiosql, need to investigate potential usage + OpenAudit integration.
Got nmap results into OpenAudit. Needed a change in admin_nmap_input.php line 48 ("Interesting ports on" -> "Nmap scan report for"), for which a bug report was filed upstream. Still need to get automatic scanning working of an entire subnet with new devices submitted to OpenAudit.

Operational Concept

Nmap -> OpenAudit

  1. Scheduled nmap ping sweep of subnets (one at a time)
  2. Check if hosts have been audited recently
  3. If not, trigger an nmap OS detection
  4. If Windows, trigger a remote audit
  5. If other device type, do ___ to audit

Scripts to check online hosts (will be /bin/sh scripts soon):

host-detect.sh (replaces all batch files below) 

#!/bin/sh
for IPADDRESS in `nmap -sP 10.14.8-39.0-255 | grep "Nmap scan report for" | awk -F ' ' '{print $NF}' | sed s/\(// | sed s/\)//`;do nslookup $IPADDRESS | grep "name =" | awk -F '=' '{print $2}';done

ping-sweep-main.bat 

cd C:\Program Files\WAMP\www\openaudit\scripts\scans
for /F %%i in (subnets.txt) do ping-sweep-backend.bat %%i

ping-sweep-backend.bat 

@echo off
cd C:\Program Files\WAMP\www\openaudit\scripts\scans
del %1.ip.txt
nmap -sP %1 | "C:\program files\utilities\bin\grep" "Nmap scan report for" | "C:\Program Files\Utilities\bin\awk" -F " " "{print $NF}" | "C:\Program Files\Utilities\bin\awk" -F "(" "{print $NF}" | "C:\Program Files\Utilities\bin\awk" -F ")" "{print $1}" >> %1.ip.txt

lookup-main.bat 

cd C:\Program Files\WAMP\www\openaudit\scripts\scans
for /F %%i in (subnets.txt) do lookup-subnet.bat %%i

lookup-subnet.bat 

@echo off
cd C:\Program Files\WAMP\www\openaudit\scripts\scans
del %1.host.txt
for /F %%i in (%1.ip.txt) do @lookup-backend.bat %%i >> %1.host.txt

lookup-backend.bat 

@nslookup %1 | "C:\program files\utilities\bin\grep" "Name:  " | "C:\program files\utilities\bin\awk" -F " " "{print $2}"

subnet.txt (snippet) 

10.14.8.0-255
10.14.9.0-255

OpenAudit -> Nagios (or Cacti, or Smokeping)

  1. Schedule a query of Windows Servers from OpenAudit + installed software
  2. Check that nagiosql has each server in it
  3. If it doesn't, add host monitoring + base services
  4. For each MS software (such as Exchange, SQL, DPM), add services

For bonus points:

  1. Query LLDP to find out where device is and set up dependencies automatically
  2. SNMPTT integration to Nagios (setup based on device info in OpenAudit)
Retrieved from "https://wiki.alpinelinux.org/w/index.php?title=Setting_up_A_Network_Monitoring_and_Inventory_System&oldid=4394"