[Pkg-nagios-devel] Bug#528262: Bug#528262: nagios-plugins-standard: Error in command_line in nt.cfg causes errors
sorens at h3solution.com
Tue May 12 15:47:19 UTC 2009
> For me "check_nt!'MEMUSE -w 80 -c 90'" and "check_nt!'USEDDISKSPACE -l c -w 90
> -c 95'" works without problems.
> Actual I don't see any need for changes. Using different ports is an issue,
> but I'm not willing to break existing systems with changing it (default port
> is 1248).
> With kind regards, Jan.
I'm new to Nagios and spent the last few days setting it up and learning
the ins and outs of the system. I did a fresh install of nagios on my
Debian server and a new install of the latest version of NSClient++ from
their website (version 0.3.5.2). From the conversations that has
occurred in this bug report, it appears that the command_line that ships
with this package worked correctly with older versions of NSClient++.
However, in my case, all of the changes I mentioned were required for
NSClient++ to work correctly.
It seems to me that the default config files should work on a virgin
installation of the package. Of course, advanced admins always have the
right to tweak them to meet specific needs.
Let me step through each of them with a little more explanation.
I didn't realize until reading the responses to this bug report that the
default port for NSClient++ has changed. That probably explains why it
must be explicitly stated in the command_line. Not stating it causes
all checks to fail. Making this change to the config file will cause
problems for those who are using older version of NSClient++ and those
who have manually changed their port numbers, but it seems to me that
deference should me paid to working correctly with current versions.
Having two arguments ($ARG2$) is required if using the sample
windows.cfg file contained in nagios3-common. For example, it contains
the following section:
service_description CPU Load
check_command check_nt!CPULOAD!-l 5,80,90
which is split into two arguments. Not listing $ARG2$ causes this
command to fail because -l 5,80,90 is not passed on. This check_command
could be modified to group everything into one argument and it would
probably work that way (I didn't test it, so your mileage may vary).
However, all the documentation on both the Nagios website and included
in the various Debian Nagios packages lists the examples this way.
Including $ARG2$ causes no damage and matches this documentation, so it
seems like an obvious choice.
The ' marks around the arguments also causes problems for me on some of
the commands. It occurred on all the commands that contained two -.
service_description Memory Usage
check_command check_nt!MEMUSE!-w 80 -c 90
service_description C:\ Drive Space
check_command check_nt!USEDDISKSPACE!-l c -w 80 -c 90
check_command check_nt!SERVICESTATE!-d SHOWALL -l W3SVC
check_command check_nt!PROCSTATE!-d SHOWALL -l Explorer.exe
Removing the ' marks from the arguments solved the problem. This may be
another case of the system working correctly with older versions of
NSClient++ but not the current version.
All of these recommendations follow the documentation found on
nagios.org and contained in the Debian Nagios packages. Specifically,
on http://nagios.sourceforge.net/docs/3_0/monitoring-windows.html near
the bottom of the page, it recommends the following syntax:
command_line $USER1$/check_nt -H $HOSTADDRESS$ -p 12489 -s PASSWORD -v $ARG1$ $ARG2$
The $USER1$ and -s PASSWORD are optional configurations that should be
made by admins who understand what they are doing, but I felt the other
changes I have recommended are necessary for the system to function
correctly in a virgin installation.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Pkg-nagios-devel