Delay WARN/CRITICAL notifications - no delay for OK

[dianeghir]

HI

I have a couple services that are a little flaky so we delay notifications to allow them to stabilize. However, when they come back I'd like the OK notification to be sent right away (not wait). Is there a way to do this?

thanks

diane

[slansing]

Are you currently using the 'Delay next notification' function? If so, what happens when you delay it, and it is still in a problem state? Are you saying that you click that until the object returns to an OK, and then you want an immediate notification? It should send you a recovery notification regardless, unless you are suppressing your notifications by other means.

[dianeghir]

When I get the first warning/critical I want it to try 5 more times before sending (to avoid false-positives).

When it recovers, I want to receive the OK message immediately.

The problem is that the OK notification also seems to be retrying 5 times before sending the notification.

In my Notification message I have
***** Nagios Service Alert *****
Attempts before Notification: %currentattempt%
Report for %host%::%service%
Received from address: %hostaddress%
Report Status Level: %servicestate%
Date: %datetime%

These are the 2 messages I got:

***** Nagios Service Alert *****
Attempts before Notification: 5
Report for hostname.net::Promo Ping
Received from address: 192.168.1.15
Report Status Level: CRITICAL
Date: 2014-04-08 00:03:52

***** Nagios Service Alert *****
Attempts before Notification: 5
Report for hostname.net::Promo Ping
Received from address: 192.168.1.15
Report Status Level: OK
Date: 2014-04-08 00:08:53

Here is my service defn:
define service {
host_name hostname.net
service_description Promo Ping
use a4_host_services
display_name Promo Ping
check_command check_a4_ping_version!promo!30112!!!!!!
is_volatile 0
initial_state o
max_check_attempts 5
check_interval 5
retry_interval 1
active_checks_enabled 1
passive_checks_enabled 0
check_period 24x7
obsess_over_service 0
check_freshness 0
event_handler_enabled 0
flap_detection_enabled 0
process_perf_data 1
retain_status_information 0
retain_nonstatus_information 0
notification_interval 60
first_notification_delay 5
notification_period 24x7
notification_options w,c,r,
notifications_enabled 1
contact_groups Admin Group
register 1
}

[lmiltchev]

Can't you enable flap detection to avoid false positives?