16

What good software is available (free or not free) to help me keep track of information relating to hundreds of servers, their relationships to each other (parent/child, category, type), and information on connecting to them, as well as possibly showing a picture or grid of some kind that allows me to report these relationships and key information to my supervisor.

I am trying to avoid the "spreadsheet solution" or "visio solution" because I want to share this information and make changes with other persons in my server team.

In other words, the solution I am looking for is a cross between a spreadsheet solution and a visio solution, providing both graphing and configuration information WITHOUT monitoring, and in a consistent format.

HopelessN00b
  • 53,795
  • 33
  • 135
  • 209
djangofan
  • 4,182
  • 10
  • 46
  • 59
  • 1
    5 votes up and two stars within 28 minutes. Sounds like a popular question. – Stefan Lasiewski Nov 29 '11 at 20:16
  • A wiki is a good soultion for the documentation end of this. For real-time information you'll want a monitoring system that does some kind of autodiscovery of systems (and at least the network-level relationships between them). – voretaq7 Nov 29 '11 at 20:19
  • 1
    You are all too quick to comment. If you actually read my question, I am NOT asking for a monitoring solution. – djangofan Nov 29 '11 at 20:40
  • 2
    If you need something integrating the documentation capabilities of a CMDB with monitoring, there probably is hardly anything as comprehensive and complete as HP's OpenView and IBMs Tivoli suites. – the-wabbit Nov 29 '11 at 20:41
  • What kind of relationsships do you think about exactly? Can you give some non-obvious examples? – Nils Nov 29 '11 at 22:32

6 Answers6

3

OpenNMS is an option. Slightly less pain and suffering than Nagios.


Edited to add - on the "Not Free" side of things there's InterMapper, which I have pretty extensive experience with.
InterMapper can generate reasonably complete network maps (assuming you have SNMP running on your hosts), but I don't consider it a good substitute for properly maintained vision (or graffle) network diagrams.

voretaq7
  • 79,879
  • 17
  • 130
  • 214
  • I always thought OpenNMS was more like SolarWinds Orion - it diagnosed network issues, but didn't do much in the way of server monitoring. – Driftpeasant Nov 29 '11 at 19:55
  • I haven't worked with Orion, but OpenNMS definitely has monitoring capabilities - I believe it can also do some basic inventory management which is what I think @djangofan is ultimiately driving at) – voretaq7 Nov 29 '11 at 20:17
  • OpenNMS can monitor services and collect data. Notifications may be sent out on either service state or data thresholds. But it should be noted that OpenNMS is significantly more resource-intensive than Nagios and will take more time to configure for a "quick & dirty" setup - many things are just not straightforward. – the-wabbit Nov 29 '11 at 20:30
  • @syneticon-dj My experience with "quick-and-dirty" Nagios setups that outlived their usefulness and just spew junk has biased me against it as a tool. I freely admit that I'm biased though :-) – voretaq7 Nov 29 '11 at 20:35
  • Don't shoot the messenger... – the-wabbit Nov 29 '11 at 20:47
  • collateral damage -- I have other legit complaints too though :) – voretaq7 Nov 29 '11 at 20:48
  • My question is not about monitoring. – djangofan Dec 09 '11 at 22:27
2

There is the ever popular NAGIOS with enough plugins/agents/8x10 glossy photographs with a paragraph on the back of each one that you should be able to get the information you're looking for. It'll take work getting it setup, though.

Driftpeasant
  • 3,217
  • 2
  • 22
  • 28
  • This is a great solution for graphing but doesn't provide a way for me to easily store/share information about the servers and/or information about what is on them and/or information on passwords and connectivity info. – djangofan Nov 29 '11 at 19:53
  • The agents that are available can capture information about the servers/installed applications/etc. Connectivity info as well. Password vaulting, however, you're on your own for. – Driftpeasant Nov 29 '11 at 19:55
  • 1
    @Driftpeasant Can you provide more information about these Nagios agents which can capture information about the servers, installed applications and connectivity info? – Stefan Lasiewski Nov 29 '11 at 21:49
  • You forgot the circles and arrows! – Cascabel Nov 30 '11 at 05:26
2

Take look at device42. (http://www.device42.com). It is available as a virtual appliance, so very quick to setup and test drive.

We are working on auto-discovery module right now for automatically populating the database along with few more user requested features.

EDIT: Software now supports network discovery using SNMP and windows and linux device discovery using auto-discovery client and/or APIs.

Raj J
  • 677
  • 4
  • 7
  • A lot of features since I first asked this question... Even though it is $250, it seems like a great product and it has the network tree visualization I wanted. Thanks. – djangofan Apr 04 '12 at 02:31
1

I am considering RackTables but I am going to need time to review it.

djangofan
  • 4,182
  • 10
  • 46
  • 59
  • While this may theoretically answer the question, [it would be preferable](http://meta.stackexchange.com/q/8259) to include some details along with the link. – freiheit Dec 02 '11 at 18:32
0

You could take a look at Open-AudIT (in particular the second version - OAv2). It will inventory devices on your network, but (at the moment) has no inbuilt auto-generated network maps. This feature is most definitley on the list of things to d, though.

Caveat - I'm the developer...

If I can help anyone out, don't hesitate to join the forums or email me.

http://www.open-audit.org

Mark Unwin
  • 255
  • 1
  • 6
  • This is a great answer but after looking at your screenshots it doenst even seem like I can arrange server data in a "tree" heirarchy either. No visializations of any kind??? – djangofan Dec 09 '11 at 22:29
0

There are several options, and most of those options are opinionated driven. I'll mention a few different off the shelf packages that might help and give you some info on what I do in my environment.

Options for monitoring are a diverse bunch, but depend greatly on what you'll be monitoring and what information you want to obtain and be able to analyze. Here are several common utilized monitor suites and what they are known for:

  • Nagios - best all around system monitoring for immediate problems with a strong plug-in framework for all kinds of services
  • BigBrother - works well for basic system monitoring, generally slightly easier and less complex to get up, running, and usable than nagios is
  • cacti - best for monitoring performance over time, generates lots of graphs over time to see what your real time stats are over time
  • munin - an alternative to cacti

I tend to use a combination of Nagios and Cacti to respectively actively monitor and see trends over time.

As for network topology organization, I simply use a basic wiki like Twiki or MediaWiki and provide details based on links to different system topology or configuration on there.

For system configuration I use a program I developed in-house called 'synchronicity', but now-a-days there are several open source projects that are similar, namely:

Andrew Case
  • 3,489
  • 3
  • 23
  • 39