[Nagiosplug-help] Running local checks on windows systems?

[gmartin]

\\Greg



On Fri, Jan 8, 2010 at 1:10 PM, Dan Stromberg <dstromberglists at gmail.com>wrote:

> There appears to be multiple ways of doing the local checks on Windows
> systems, and while there is commentary on the strengths and weaknesses
> of them in google, I've found the contrasts somewhat vague.
>
> So please allow me to describe my understanding of the strengths and
> weaknesses here, and hopefully People In the Know(tm) will jump in and
> correct me where I'm off.
>
> Here goes:
>
>
> nsclient:
> * Both a windows service and a protocol,
> * I've not looked at this one that much, as it appears to be superseded
> by nsclient++.
> * Connected to using Nagios plugin check_nt.
>
>
> nsclient++
> * A windows service that can serve up either or both of the nsclient and
> NRPE protocols.
> * Has an assortment of DLL's implementing the various checks, whether
> you connect using the nsclient protocol (check_nt) or the NRPE protocol
> (check_nrpe).
> * Allows you to continue using check_nt, which is nice if you already
> have some nagios checks based on that.
> * Allows you to start using check_nrpe for some things, but those things
> appear to require a dll?  As such, it seems to paint you into a corner;
> if you get a bunch of NRPE checks going based on nsclient++, it appears
> that a future move to NRPE_NT is made more complicated.
> * Granted, you could probably write a dll that just accepts arguments
> and treats them as a bash/powershell command.  Or is there something
> that already does this in nsclient++?
>
>  I'm running batch-file based checks using nsclient++ & check_nrpe witout
using a dll.  nsclient++ does this just fine.  The config take a little
investigation, but once you understand it, its easy.


> nrpe_nt
> * A windows service that serves up the NRPE protocol, just like on *ix.
> * Allows you to use the same sort of plugin architecture for local
> checks that is used on *ix - enabling one to use the many plugins for
> Windows at
>
> http://exchange.nagios.org/directory/Plugins/Uncategorized/Operating-Systems/Windows-NRPE
> * Is this the reason why people are saying that NRPE is the way forward,
> not NSClient?
>
>
> sshd
> * Available as a cygwin sshd or via freesshd.  Speaks the ssh protocol(s)
> * Allows you to use the same sort of plugin architecture for local
> checks that is used on *ix - again enabling one to use the many plugins
> for Windows at
>
> http://exchange.nagios.org/directory/Plugins/Uncategorized/Operating-Systems/Windows-NRPE
> * Perhaps a little slower than NRPE, especially if you use NRPE without SSL
> * Much better authenticated than NRPE_NT, because NRPE_NT just uses
> IP-based authentication, while ssh is RSA public key with mutual auth
> * Probably harder to lock down to a list of specific commands than NRPE_NT
>
>
> SNMP
> * Another possibility I've not looked into much, at least not in a
> Nagios context.
> * Decent authentication in version 3, not so good in versions 1 and 2c.
> * Allows access to a LOT of stuff
>
>
We use SNMP for hardware monitoring of WIndows servers in addition to
nsclient++


>
> I'll add that one could probably run nsclient++ with the nsclient
> protocol enabled and nsclient++'s NRPE protocol support disabled, and
> also run NRPE_NT - on the same machine.  It seems like this would allow
> compatibility with existing check definitions, and give a nicer path
> forward - because where new plugins are needed, they could just be
> scripts in your favorite scripting language; they wouldn't need to be
> dll's in Windows-specific C or something.
>
>
> OK folks - where am I off?  Please tear it to shreds.
>
> Thanks!
>
> \\Greg <file://greg/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.monitoring-plugins.org/archive/help/attachments/20100108/e1509d7c/attachment.html>