Application monitoring

[PaulVM]

hello all,
newbie looking for solution
I have read around, but I need some advice to plan

The need is to monitor the use of certain predefined applications during business hours on workstations.
For example, writer/word, calc/excel, thunderbird, firefox, ...
No matter if they are simply open in background or are actively used.
The only interest at the moment is to extrapolate how long are active and able to do the statistics on these data.
My idea is to install Nagios to a VM somewhere on the net.
The stations are not directly accessible from the Nagios server, so I have to use an agent in passive mode (NCPA or NSClient++ I suppose).
I am new to Nagios, but this seems simple (or not too complex ).
Am I right until now or I already had done some wrong evaluation?

If The previous considerations are fine, I now have an info registered in some DB, that report the application status for a specific workstation every N minutes.
MySQL or RRD?
Is there any simple way to do the statistics on these data?
Obviously I am interested to evaluate the workstation's activity and the applications activity during business hours (my be 8-12, 14-18 or simply when the PC is ON).
I am not interested to reinvent the wheel, so, if someone already do something similar ...

Thanks, P.

[slansing]

If The previous considerations are fine, I now have an info registered in some DB, that report the application status for a specific workstation every N minutes.
MySQL or RRD?
Is there any simple way to do the statistics on these data?

I'm a bit puzzled by this, are you already monitoring the metrics you need? You just don't know how to send that back up to nagios?

[mickem]

NSClient++ actually supports statefull scripts /python/lua) so you could do it there I guess.
Another options would be to put the performance data (i.e. time application has been running) inside something like graphite or TSDB and graph it (or possibly query it) from there.
It depends a bit on how you want the data (i.e. aggregated or per system (per user?), trends or absolutes etc etc.,..)

// Michael Medin

[PaulVM]

If The previous considerations are fine, I now have an info registered in some DB, that report the application status for a specific workstation every N minutes.
MySQL or RRD?
Is there any simple way to do the statistics on these data?

I'm a bit puzzled by this, are you already monitoring the metrics you need? You just don't know how to send that back up to nagios?

Sorry for my poor english

Try to clarify:

My need: monitoring applications usage and do statistics (data collected from windows workstations in various offices, under firewalls, ...)

Is Nagios a good solutions for this?

Is there something similar already developed (addon, plugin)?

Any hints on configurations and tools to achieve the result is appreciated.

Thanks, P.

[slansing]

Based on what you are checking, and what protocol you are using "and what plugin," you can pull the above information, including performance data "graphing." What are you specifically looking to check? If we have a bit more detail we can point you to a solution. There are a number of ways you can check the "vitals" of your Windows machines, be it NCPA, NRDS, SNMP, WMI or other agent/agent-less solutions

[PaulVM]

Based on what you are checking, and what protocol you are using "and what plugin," you can pull the above information, including performance data "graphing." What are you specifically looking to check? If we have a bit more detail we can point you to a solution. There are a number of ways you can check the "vitals" of your Windows machines, be it NCPA, NRDS, SNMP, WMI or other agent/agent-less solutions

Starting the development clarify something

I installed Nagios (VM in datacenter). I can monitor agentless server/services.
I want monitor the windows workstations inside the LAN (under firewall).
I installed NSClient++.
Now I need some config example for passive check using NCPA .
Found this tutorial (http://nsclient.org/nscp/wiki/doc/usage/nagios/nsca), but it is incomplete.
I now have a client that insert messagges in the server queue.

Where can I find detailed working example (windows NSClient ++ and Linux server config)?
I need monitor processes state (application active or not).

Next I need to extract statistics usage for the business hours.

Thanks, P.

[abrist]

I want monitor the windows workstations inside the LAN (under firewall).

You can do so by opening ports, using a pn, passive reporting, etc.

Now I need some config example for passive check using NCPA .
Found this tutorial (http://nsclient.org/nscp/wiki/doc/usage/nagios/nsca), but it is incomplete.

NCPA or NSCA?

Where can I find detailed working example (windows NSClient ++ and Linux server config)?

Config for what check and what agent?

I need monitor processes state (application active or not).

This is possible with most agents.

Next I need to extract statistics usage for the business hours.

For statistics other than states, you will most likely need to set up a performance data graphing addon. I suggest pnp4nagios.

[PaulVM]

I want monitor the windows workstations inside the LAN (under firewall).

You can do so by opening ports, using a pn, passive reporting, etc.

Ok. installed server on CentOS 6.4 VM
Using NSClient++ on windows clients for passive reporting

Now I need some config example for passive check using NCPA .
Found this tutorial (http://nsclient.org/nscp/wiki/doc/usage/nagios/nsca), but it is incomplete.

NCPA or NSCA?

Sorry. NSCA. If there is a better/simpler way to do passive check I can change

Where can I find detailed working example (windows NSClient ++ and Linux server config)?

Config for what check and what agent?

I have NSClient++ that reach the server and insert messages in the queue.
Tried to modify the examples, but with very limited results
I need the right syntax to define services in the server part and the check in the agent.
Currently I have agent and server that are speaking but not understanding each other ...

I need monitor processes state (application active or not).

This is possible with most agents.

Searching for syntax example in nsclient.ini (and correspondig syntax service definition in the server).

Next I need to extract statistics usage for the business hours.

For statistics other than states, you will most likely need to set up a performance data graphing addon. I suggest pnp4nagios.

Start reading docs. Thanks

Regards, P.

[slansing]

Can you restructure that last reply so that we know what you need help with?

And what do you mean by a passive check you can change? NRDP would probably be a good solution here as it will use standard plugins which you can modify:

http://exchange.nagios.org/directory/Ad ... PA/details

[PaulVM]

Can you restructure that last reply so that we know what you need help with?

And what do you mean by a passive check you can change? NRDP would probably be a good solution here as it will use standard plugins which you can modify:

http://exchange.nagios.org/directory/Ad ... PA/details

Sorry. Tried to quote.

Currently, I have the server side that seems working. It receives the NSCA notifications and queue it. But I haven't a valid host+services config because I modified the standard template, but with por results. I am searching for a NSCA passive check template/example config.

In the client side I currently have NSClient++.
Started with basic configuration to test it is working (reported below). I can see the warning in the server log so I can suppose communication is working. Now I am searching for some config example for both client and server to use as template insted that start from scratch. My primary interest is to monitor activity of a pool of application (if Explorer, Firefox, Thunderbind, Word, Excel, ... are currently running or not).

Is there any example available?

Your link point to NCPA. Do you mean this? http://exchange.nagios.org/directory/Ad ... or/details

Thanks, P.

My client basic config (NSClient++):
--------------------------------------
[/modules]
CheckSystem=enabled
CheckDisk=enabled
CheckExternalScripts=enabled
CheckHelpers=enabled
Scheduler=enabled
NSCAClient=enabled
[/settings/scheduler/schedules/default]
interval=5m
[/settings/scheduler/schedules]
cpu=alias_cpu
mem=alias_mem
disk=alias_disk
service=alias_service
[/settings/NSCA/client]
hostname=win_${host}
[/settings/NSCA/client/targets/default]
address=1.2.3.4
encryption=1
password=nscpass
--------------------------------------