Rewriting nsclient? (was NSClient 1.0.8.0 Installation Problem)

Michael.Wirtgen at miwi-dv.com Michael.Wirtgen at miwi-dv.com
Fri May 9 14:27:40 CEST 2003


Tim, List

>> I run NSClient on servers provided my clients where I purchase my own
development tools.
Miracle solved, thanks :-)

>>Monolithic vs. Module
backwards compabillity is a valid point, haven't thought of this yet, just
looked at it like "could be a plugin as well" for the checks I've seen.
However, there may be some stuff, that can't be a plugin, as "permanent"
monitoring is required. What I don't like is the idea of adding even more
complexity into NSClient...

>>For the future going to a NRPE type approach
I think, we should look how to avoid any double work. I don't see any reason
in implementing nrpe functionality for windows twice....
Immediately three scenarios come to my mind:

a)  npre on ux, nsclient with nrpe on windows, nrpe_nt (that's what I call
my code currently...) on windows
    That's what I'd consider the worst case, as it would waste development
time better spent on writing windows plugin or sitting at the beach ;-)

c)  npre on ux, nsclient with nrpe on windows, nrpe_nt discarded
    Not to much time wasted, and I'd offer a hand on your effort (haven't
seen Delphi since 3.0 however...)
 
b)  npre on ux, nsclient w/o nrpe on windows, nrpe_nt on windows
    In that case, I'd suggest to focus nsclient on real "agent" stuff (Read:
things that can't technically be plugins, as they need to register callbacks
for events etc.). 
    Also nsclient submitting passive checks back into nagios would be nice
for this (that's because the kind of checks I have in mind, would mostly be
submitted passive by nature...)


>>To permit this type of power, it needs the same safeguards as NRPE 2.0 is
implementing.
...[snip]....
>>...the service itself is best written in a compiled language and interact
properly with the SCM.
I couldn't agree more...

Just some thoughts, any comments welcome
cheers
MiWi


-----Original Message-----
From: Tim Shouldice [mailto:tim at tsmgsoftware.com]
Sent: Friday, May 09, 2003 4:15 AM
To: Michael.Wirtgen at miwi-dv.com
Subject: Re: [Nagios-users] Re: Rewriting nsclient? (was NSClient 1.0.8.0
Installation Problem)


Michel:

Many good points and thoughts but I have to start off by saying I run
NSClient on servers provided my clients where I purchase my own development
tools.

Monolithic vs. Module

I prefer modular, however as the current NSClient is somewhat monolithic and
it needs backwards compatibility I would leave its current functionality
pretty much 'as is'.

For the future going to a NRPE type approach, the NSClient would be given
the functionality of runing a plugin with a parameter on the Windows and
returning the results back to check_nt / check_nrpe. To permit this type of
power, it needs the same safeguards as NRPE 2.0 is implementing.

The plugins could then be written in the language of choice much as the Unix
ones are now. However, as I've expressed before (and you agree with), the
service itself is best written in a compiled language and interact properly
with the SCM.

Tim


-------------------------------------------------------
Enterprise Linux Forum Conference & Expo, June 4-6, 2003, Santa Clara
The only event dedicated to issues related to Linux enterprise solutions
www.enterpriselinuxforum.com

_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. 
::: Messages without supporting info will risk being sent to /dev/null





More information about the Users mailing list