NC_Net -passive checks (test version)

[Anthony Montibello]

The documentation on passive checks was written prior to implementing
the passive checks.  What you are requesting is for the most part what
I am planning on implementing.  the other port thing was a earlier
thought about having a management application in windows that can
modify the configuration (this is probably not going to happen).

The current plan that should be implemented next week some time, is to
have check_nt amended with a new option that will push the passive
check command to NC_Net.

The passive checks can be configured at startup.  the startup.cfg file
is where all the startup information about flags, ports and intervals
goes.  the passive.cfg file is read each iteration that passive check
should run,  currently prior to starting NC_net you can add a bunch of
checks to the end of the passive.cfg file. Passive.cfg is parsed every
iteration for what passive checks to do.  So if the passive.cfg is
manually changed the new changes will occur at the next passive check.
 I use the file "Sample Passive.cfg" and save as to modify my checks
while NC_Net is running.

For Changing the checks via Check_nt, I need to add 1) a few active
checks that change current configuration (this is already in NC_Net
but not in check_nt) 2) a active check that changes startup.cfg (not
in either check_nt nor NC_Net) 3) a switch to add to a passive check
that will add it to the passive.cfg ( i am thinking of using the [-s
serviceDescrip]  each passive check MUST have a service description so
the command should fail if it is not present.  Also the result from
this command should be OK if the check was added and tested with an OK
result.  2) Warning - if the check was added and tested with a non OK
result.  3)Critical - if the check was not added - a check will not be
added if it testing it returned an ERROR 4) Unknown - if the check
fails for some other error.  check_nt needs to catch the -s switch
prior to its main service check routine because - check_nt sends very
little information to NC_Net for a service check and we basically need
it to send over the entire service check options. ( -d,-w,-c,-l,-v,-s)

I also will need to add an active check command to delete a passive
check - this will be just the service description with another
switch.(any suggestions)

A point worth mentioning about the changing the configuration. 
several flags exist in the Setup.cfg that will disable the ability to
make changes to NC_Net.  setting lock_active_config to true prevents
changes from an active check.  While lock_passive_config set to true
prevents a configuration change in  passive.cfg from running.  Also
the Mode variable at the end of the Startup.cfg is not an actual
variable, it is actually a mask com mad that overwrites the explicit
settings of these variables that were set earlier in the file. (this
was done intentionally, the mode is the primary interface for changing
all true/false variables at the same time, so I use ti to change the
configuration during testing.   One of the goals of having the
Startup.cfg file was to enumerate all the configuration variables that
a user has access to change since this list will not be available from
a command sent to NC_Net.  The Testrun also overwrites other variables
that may have been set. testrun disables the actual sending of the
passive checks via NSCA as well as increases the frequency to check
each minute.

I hope this post more appropiatly lays out the intentions of the
updating passive checks remotly via Check_nt.
Tony


On Fri, 4 Feb 2005 11:10:25 -0800, Chris Hale <chris.hale at fiducial.com> wrote:
> I've read the documentation on passive checks... and I like where you are
> going, but wanted to give some feedback on the remote controllability of the
> passive checks.
> 
> By the documentation... It looks like you are planning on having any remote
> configuration of the passive commands be done on a separate port from the
> active checks.
> 
> Before reading the docs... I always thought it would be nice, if check_nt
> could send just another optional variable along with an active check to set
> a timer, having nc_net then send back passive results to the configure NSCA
> server
> 
> IE:
> 
> 1) NC_Net is installed and configured with information about the NSCA to
> report too.  (But unlike right now... it does not have a local config of
> passive checks)
> 
> 2) Nagios is setup to receive passive only results for a particular test
> 
> 3) This service is also set with freshness checking and the command for the
> check is the appropriate check_nt active check command including the extra
> --passive (X seconds) variable.
> 
> So the real time scenarios are:
> 
> A)
> 
> A server reboots... therefore the service starts with no passive checks
> configured.  The Nagios server comes to the freshness check limit and send
> the active check (which gets the first test result after the reboot) and
> also set the passive check that will be used from there on.
> 
> B)
> 
> Someone in management decides that they want the alerts for drive space on
> 500 servers changed from 85% to 86%.  You change the check command for this
> one service check (with the 500 hosts :) ) and then force an active check on
> these hosts.  The nc_net receives the check_command with the new warning
> level and changes the passive check that is on timer.
> 
>


-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl
_______________________________________________
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