Host/Service check scheduling

[Denise Sandell]

On Mon, 28 Apr 2003, Ethan Galstad wrote:

> Host checks are performed when:
> 
> 1) A service changes state

Right, but this is intended to be a soft state, correct?

> 2) A service is first checked and is found to be OK (if aggressive 
> checking is on)
> 3) A service is checked and remains in a non-OK state (if aggressive 
> checking is on)
> 
> They may also occur if an upstream or downstream host changes state.  
> In this case, host checks might be propagated up and down the 
> parenting tree to see if the states of other hosts have changed.
> 
> Based on your description, you need to choose another method of 
> checking your host.  If 5 consecutive host checks are failing and the 
> host is really up, something needs to be reconfigured.  IMHO, a 50 
> second failure for a host check should indicate that there's some 
> kind of problem.

The case where the host check failed for 5 consecutive attempts
was correct, it _should_ have failed because of a soft service state
change (and did).  The inference made to the host staying up was
upon completely disabling the host check_command for that particular
host.

The desired function I want is not to receive host notifications
for things down under 5 minutes -- which to me means I need to increase
the normal_check_interval for my services on said host, thereby decreasing
the possibility of a service changing state within that period of time.

> 
> 
> On 28 Apr 2003 at 13:45, Denise Sandell wrote:
> 
> > 
> > Howdy,
> > 
> > I have a question/concern about the way host and service checks are
> > handled.
> > 
> > Host checks are only supposed to be initiated when a service check
> > fails.  So at what point does the host check initiate, when a service
> > returns a "soft" non-OK state, or a "hard" non-OK state?  I've
> > poured over the documentation and can't find anything that explicitly
> > states what the function is.
> > 
> > As far as I can tell operationally speaking, a host check is initiated
> > immediately when a service check fails (ie a soft fail state).  Yet,
> > this is not clear from the event logs.  I never see a failed/soft
> > service check before a host check is initiated noted in the logs.
> > As a result, my service checks fail once, a host check is initated,
> > it counts through max attempts defined for a host, and then issues
> > a notification.
> > 
> > This is rather annoying since I would like to receive notifications 
> > for things down >5 minutes.  My max_attempts for my host template is
> > currently set at 5 (with an interval_length of 60), so once Nagios
> > initiates host checks, it does so every 10 seconds 5 times in a row.
> > As a result, I get notifications for things down for 50 seconds.  
> > 
> > The only way i've managed to get around this is to completely disable
> > check_commands for hosts.  This leads to some general retardedness for
> > availability of hosts/services.  If my service check for ping fails,
> > the host goes down, but it is never marked as such by the system.
> > 
> > So, the questions i have are these:
> > 
> > . is this the intended function for network availability?
> > . is there an undocumented way to apply check intervals to host checks?
> >   and if not, is this being considered as an option for future development?
> > 
> > -- 
> > - Denise Sandell                                 network operations -
> > - dsandell at voyager.net              voyager.net, a CoreComm company -
> > 
> > 
> > 
> > -------------------------------------------------------
> > This sf.net email is sponsored by:ThinkGeek
> > Welcome to geek heaven.
> > http://thinkgeek.com/sf
> > _______________________________________________
> > 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
> > 
> 
> 
> 
> Ethan Galstad,
> Nagios Developer
> ---
> Email: nagios at nagios.org
> Website: http://www.nagios.org
> 
> 
> 
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> 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
> 

-- 
- Denise Sandell                                 network operations -
- dsandell at voyager.net              voyager.net, a CoreComm company -



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
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