Open Source Monitoring Tool

[Hunki]

Does anyone can suggest a tool through which I can monitor Unix processes real time through a website.I know of Big Brother but its installation and documentation is not streamlined, but I wud be looking at something similar to big brother and easy to configure.

Hunki

[RAC_1]

There is nagios and webmin. I do not if webmin is full fledged monitoring tool or not. Serach for nagios on google.

[Arunvijai_4]

Hi Hunki,

You have to take a look Internet express suite for HP-UX,

http://h20293.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=HPUXIEXP1111

-Arun

[Roberto Arias]

Hello Hunki:


I suggest you the system nagios, is very good and GNU. you can download of www.nagios.org. The manual in spanish is in http://nagios.linuxbaja.org/?q=node/15

Best regards.

Roberto Arias

[Ralph Grothe]

Hi Hunki,

there are several Open Source monitoring tools available,
of which some were already mentioned.
It depends on your requirements, prerequisites,
likes or dislikes.
But afaik Webmin isn't a monitoring tool but rather a webfrontend to do your sysadmin tasks via a browser.
If you have a large site to monitor with up to some hundreds/thousends of hosts (where a host can be anything from pretty dumb network device to enterprise server),
and if you maybe need to monitor very individual quantities/states where you want to provide your own custom plug-ins,
then I would go for Nagios.
It has a pretty good documentation with interspersed example configurations, and a very broad community with active mailing lists of Nagios experts who are very willing to help.
The only difficult thing maybe is getting a working starting configuration because it usually requires six config files to be edited which mutually depend on each other.
But the tarball includes many sample files that you only need to customize to get started.
And the nagios daemon does a so called pre-flight check of your configuration and would abort if it detects any malconfiguration with telling of the violating file and line No.
The most important part of the Nagios functionality however are the Plug-ins.
So you also need to install the plug-in bundle to have anything useful.
The plug-ins supply already check commands for the most common services.
If you have some scripting or programming experience it is pretty easy to extend the Nagios functionality be writing your own plug-ins.
This is the starting link for your Nagios endaevours
http://www.nagios.org/

A similar project is Zabbix

http://www.zabbix.org/

Then there is a more lightweight monitoring tool based on Perl which you can obtain from the Linux Kernel site, and which is simply called mon.
The configuration is a bit simpler than that of Nagios, and some Perl knowledge wouldn't hurt.

http://www.kernel.org/software/mon/

Then there are those tools that are more suited for collecting performance data
(although Nagios can do this to a certain degree as well).
Of these especially in multi node clusters the Ganglia tool is widespread.

http://ganglia.sourceforge.net/

Another one is Orca

http://www.orcaware.com/orca/

Orca is based on the round robin database and graphing tool RRDTool

http://oss.oetiker.ch/rrdtool/

Also based on RRDTool is the Multi Router Traffic Grapher MRTG

http://oss.oetiker.ch/mrtg/

Another performance collection and charting tool entirely written in Perl and also based on RRDTool is Munin

http://munin.projects.linpro.no/

I use the latter on our Solaris and Linux boxes because it is on those platforms right out of the configure script immediately usable.
The reason being the easy access to performance data on those platforms,
with the procfs on Linux and kstat on Solaris.
Sadly HP-UX hides this data and requires you to buy e.g. Glance to access the data.
Nevertheless you can use plug-ins based on common user space Unix commands such as
uptime, ps, vmstat, iostat, sar, nfsstat etc.

[Lars Ebeling]

BigBrother is not fully free for commercial use. However there is a new product called Hobbit created by Henrik Stoerner who developed bbgen for BigBrother.

Have a look at
http://hobbitmon.sourceforge.net/

Regards

Lars

[Rick Garland]

As with any monitoring tool, configuration time and effort need to be part of the decision making process.

Supportability is also very important issue to consider.

As mentioned, there are several packages to consider and in making your decision, factor in the complexity of your environment with the time/effort required to configure the various packages for your environment.