Choosing a network monitoring system and software

Question

Monitoring

Greetings!

For a small company (max 50-70 monitored machines) it is necessary to choose a network monitoring system and software. First of all, we will monitor critically important (24x7) software - logs, performance and availability.

The question is in the choice. There is a great temptation to use Zabbix, but still I want to objectively evaluate all the pros and cons of each system.

Candidates:
Cacti
Nagios
OpenNMS
AggreGate Network Manager
Zenoss
Pandora FMS
PRTG Network Monitor
NetDecision
Zabbix
Ganglia

I will be glad to hear feedback on the real (everyone has already read the advertising brochures) use of these systems. Interested in features such as:
— ease of installation and support;
- extensibility;
- productivity;
— reliability;
— data visualization;
— distributed monitoring;
— escalation of incidents;
— a wide range of alerts;
— price;

Also interested in whether “adult” information systems management software, such as IBM Tivoli or HP OpenView, is suitable for such purposes?

I would be glad to hear your answers and suggestions.

Answer 1

Used nagios to monitor about 300+ machines
1) ease of installation and support - available in packages in many distributions, easy to set up
2) extensibility - many plugins, including for network equipment, you can write your own, it will be enough to write a shell script, which writes data in a specific format.
3) performance - for 300+ machines, the server had enough 256 MB, while OCS Inventory was still working there
4) reliability - like a brick, very reliable.
5) data visualization - present, beauty is an amateur.
6) distributed monitoring - did not quite understand what is meant. If that's what I thought, then there is an NRPE that executes plugins remotely on the host and gives the plugin's response back to the monitoring server. If you mean the use of several monitoring servers in conjunction, then I'm not sure that this is necessary in an enterprise of 50-70 machines.
7) escalation of incidents - I’m not ready to say anything specific, I can’t imagine what the situation is with this in your enterprise, what is the hierarchy of incident management. Based on the size of the company, one technician and a couple of management people.
8) a wide range of alerts - you can divide users into groups, specify which groups to send alerts to, three service states or parameters are used as standard - OK, WARNING, CRITICAL. There is also the UNKNOWN status, which occurs in case of incorrect plugin operation.
9) price - free of charge
Plus, it is possible to execute scripts remotely in case of a critical situation. For example, distort the service.
In addition, you can use the coordinate drawing on the map. You set the coordinates for all the machines and see on the diagram where it is located. True, you have to take care a little with setting the coordinates.
Cacti is more suitable for monitoring via SNMP, therefore, it is more convenient for them to monitor network equipment, I personally used nagios plugins for this.
In general, from my experience I advise nagios.

Answer 2

opsview community edition

Answer 3

purely IMHO
cacti is more suitable for monitoring network equipment - its charting system is very convenient.
zabbix is ​​more like a light version of nagius.
nagios is a very maneuverable monster. but the interface is so not beautiful that it's scary.
opsview is an add-on for nagios. very beautiful.
based on opsview, a distributed monitoring network is easily created. You can create clusters.

Answer 4

Regarding the second question.
HP OpenView did not feel, but Tivoli is better than all of the above for your purposes, but there is only one caveat - the price for them is clearly not for a small company.
From myself I advise Cacti(Nagios) + Munin

Answer 5

Author, I recently asked the same question, what did you come to? What is the result?