On Wed, 16 Apr 2003, Stanley Hopcroft wrote:
> Dear Ladies and Gentlemen,
>
> I am writing to suggest that Nag 2.0 (or later) consider reporting
> availability against SLA.
>
> It may be possible to
>
> 1 Define SLAPeriods like the current TimePeriod objects. This attribute
> is a property of host and service objects.
This objects change is the only impact on the monitoring core.
Is the SLATimePeriod
> 2 Define an SLAGuarantee for a host or service specifying the lower
> bound of the available time that the measured availability must exceed
> to meet the guarantee.
>
> (eg SLAGuarantee_999, SLAGuarantee_9999 so that if these are applied to
> a service or host the availability of that service or host must be more
> than 99.9 and 99.99% respectively for the specified SLA period.
>
>
> It _may_ be possible for Nagios to calculate availability after each
> service check [it does already I think] and submit an alert or
> escalation ...
>
>
> 3 Have avail.c discard events that fall outside the SLA period for that
> service or host.
>
> 4 Have the availability CGI return an extra indication of whether or not
> the SLA has been met (if avail >= SLAGuarantee)
>
> Yours sincerely.
>
>
--
-sg
This post was automatically imported from historical nagios-devel mailing list archives
Original poster: sghosh@sghosh.org