Support for Nagios 1.0, 1.1, 1.2 and 2.0b3

Realtime display of Nagios status

Uses colouring/highlighting for troubled items

Low-bandwidth fullscreen display works on any tty; if vi works, nsc probably does too

Displays all essential Nagios status info on a single screen

Clean & simple display & usage

Adjustable display layout

nsc runs on Unix and look-alikes - I test on Linux, Solaris & FreeBSD

Read-access to Nagios status.log

Perl 5.x

Curses libraries installed (ncurses may be required)

Perl `Curses´ module, available from CPAN

 nsc.pl [options] [statuslog]

If a statuslog isn't specified on the commandline, nsc will use the one from your ~/.nsc.conf.

--term=

When using this option, the terminal type is changed:

   nsc --term=vt220

This is the same thing as "env TERM=vt220 nsc".

-T

Disables the terminal type replacement logic (described later in this document).

-h, --hosts=

If this option is given, nsc will filter out all hosts and services for hosts not matched by the --hosts list - this will only show items from the hosts calvin & hobbes:

   nsc --hosts=calvin,hobbes

Wildcards are available using either "*" or "%":

   nsc --hosts=*.myowndomain.com
   nsc -h firewall,%proxy%,router%

   ./nsc.pl nag1-status.log

This won't give you the live display from Nagios, of course. The example status.log's are in the /usr/share/doc/nsc* directory, if you installed from the rpm package.

If you see no colours in nsc, it is very likely that your $TERM variable indicates a monochrome terminaltype (like vt100). See below for how to get around this.

The reasons for this are historic. Before personal computers, we had dumb terminals and most often, they had a green or amber monochrome display. This is what Unix terminal handling is geared towards, although today most of us have 16.7 million available colours at the desktop. So often we end up with a terminal type that won't do colours, usually due to settings on either the host OS or in the terminal application.

Getting colours working is usually just a matter of setting $TERM right - you may just need to start nsc a bit differently, for example:

   nsc --term=xterm-color

   term.linux.vt220=xterm-color

.. then when you run nsc on Linux with TERM=vt220, nsc will replace vt220 with xterm-color instead. The operating system name is retrieved from `uname -a` - nsc uses the first word and lowercases it, before looking for a replacement terminal type.

nsc knows about a number of terminal mappings - add others as needed to the .nsc.conf file.

Linux (Mandrake 9.1-10.2, RH9)

Within putty/screen, "TERM=xterm-color" works well.

Solaris 9

"TERM=xterm-color" works, at least for putty/screen. CDE will likely work too with this setting.

FreeBSD 4.7

"TERM=vt340-color" works as advertised (putty/screen).

``works'' as advertised" means that nsc will produce colour and that it looks right, specifically regarding producing an even-coloured background.

The above list is much shorter than what nsc actually will work with. So if you find new platform/terminal combos that just works, including colours, please send me info (nsc@gothix.biz) about it, so that I can incorporate the mappings in future nsc versions.

The "!normal" colour/attribute refers to the 'normal' colour/attribute and is just a shorthand to avoid specifying 'white on magenta' seven times.

   nsc my.host.net:/usr/local/nagios/var/status.log

This will use the remote filename from the commandline together with your existing settings from .nsc.conf. You can also set the remote filename in .nsc.conf (keyname 'nslog'). Another approach is to install various aliases in the shell:

   alias nsc1='nsc my.host.net:/usr/local/nagios/var/status.log'
   alias nsc2='nsc nagios.host2.com:/var/opt/nagios/var/status.log'

You will get the best results with the remote Nagios feature, if you ensure that the involved hosts have their clocks synchronized (using NTP or similar).

The remote copy interval may be changed in .nsc.conf - the default is 60 seconds and it may overridden with the keyword 'remote.copy_interval'.

The default remote.copy_cmd setup requires that:

 - You have rsync(1) installed
 - You have ssh installed (OpenSSH or similar)
 - You can login to the Nagios host using ssh without specifying a password

If these conditions are met, then you are ready to go - just run nsc:

   nsc remote.host.org:/path/to/nagios/status.log

If you want to access the remote status.log by other means than rsync/ssh, edit ~/.nsc.conf and specify a different remote.copy_cmd. This command should work for a pure ssh-/scp-setup with no rsync:

  remote.copy_cmd=scp -q -p -C %H:%F %D >%L 2>&1

%-tokens available are:

   %H => Hostname
   %F => Remote filename
   %D => Local destination filename
   %L => Logfile for output from the copy command.

In case of difficulties, try out the copy command in the shell - to test the above scp setup, use something like:

   scp -q -p -C my.host.net:/usr/local/nagios/var/status.log /tmp

You should get no output, no errors and /tmp/status.log should now be identical to the file on the remote host. The return value should also be zero, check it with 'echo $?' after executing the copy command. If the scp command works and you have rsync installed, then the default remote copy setup with rsync should work as well.

If problems arise, check the logfile from the copy operation:

   ls -lt /tmp/status* | head -10

Looking at the logfile needs to be done whilst nsc is running as it removes the logfile, when cleaning up on exit.

The above examples are based on ssh access to the Nagios host, but you can almost as easily use http and wget(1) to retrieve a remote Nagios status.log - or any other script or program that can retrieve a plain file from another host.

   usercmd_V=$VISUAL /usr/local/nagios/etc/services.cfg; /etc/init.d/nagios reload

Any of the above may be reversed with 'r' :-)

roll_wait

'roll_wait' is the number of seconds to wait since last keypress until rolling starts.

roll

'roll' is the time in seconds between shown pages - if this is set to -1, then rolling will be disabled.

Various other (Nagios) files are also used - you may edit ~/.nsc.conf as needed to reflect proper placement of these files.

EDITOR

The text editor (program) to be used for editing files, such as hosts.cfg. Defaults to `vi´.

VISUAL

VISUAL is used in preference to `EDITOR' if set (visual editor).

PAGER

For viewing files, such as status.log, defaults to `more´.

HOME

Where nsc expects to find your .nsc.conf.

If you include nsc with a Linux distribution or create a package for it for another platform, you should let me know. Obviously, I'm interested in how nsc is doing and where it goes, so please do drop me a mail at input@gothix.biz with a couple of words on what you are doing - thanks.

Versions prior to v0.80 (april, 2005) were distributed under the GNU General Public License and those versions may be used as per the GPL license.