This is with Shinken 2.4.1 running on RedHat 6.3
We are facing an issue about business_impact.
For example
Hosts: hosta, hostb, hostc
Business Impact for hosts : 2 (default)
Service for each host: serviceA
Business Impact for serviceA : 3
business rule is defined like
Code:
define service{
display_name Test BP
service_description test_bp
host_name host_global
max_check_attempts 1
check_interval 1
business_impact 4
check_command bp_rule!3of: hosta,serviceA|hostb,serviceA|hostc,serviceA
servicegroups test_group
contact_groups test_prod
use test-service
}
Immediately after shinken starts, business_impact for host remains 3, but when this business_rule become active (checked ?), business_impact for these hosts will be 4.
This was noticed under the situation where all 3 hosts were shutdown, and the "priority" in Thruk went from the default, to a level 4, which paged engineers .. even though
the business_impact for the servers was at a lower level. If the servers are up and running, but the individual services being checked go offline, the engineers did not get paged,
and the "priority" remained at the default level.
I am not sure if it is the behavior of shinken, but different than what I want.
I am am trying to find out if this is working as designed? And, if there is a way to not inherit the business_impact from the service definition.
The thought is that the host might be at a lower business_impact if "Xof: " changes to something like "1of:" .. so it does not seem to make sense for all the hosts to be at the
same business_impact as the service (eg, if you have 20 web servers, the http service definition might be higher than the definition for individual hosts, and you might not want
to get notifications at an elevated business_impact level for each host).
Or is this an odd behavior in Thruk? I would think that Shinken is sending "host down" alerts to Thruk, but I do not know why that would elevate the "priority".
Any ideas?