Module: check_mk Branch: master Commit: de0b3818e981b90d3e5dfd109d2126c5eec97bed URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=de0b3818e981b9… Author: Bernd Stroessenreuther <bs(a)mathias-kettner.de> Date: Tue May 6 11:50:20 2014 +0200 Updated bug entries #1067 --- .bugs/1067 | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) diff --git a/.bugs/1067 b/.bugs/1067 new file mode 100644 index 0000000..df7d40f --- /dev/null +++ b/.bugs/1067 @@ -0,0 +1,24 @@ +Title: Avoid multiple notifications for a single problem on a single host +Component: checks +State: open +Date: 2014-05-06 11:42:27 +Targetversion: 1.2.5i1 +Class: feature + +you have e. g. a machine with some oracle instances +per instance there are multiple checks +now if one instance goes down, you get multiple notifications (one per check) +this is normally unwanted + +in classical Nagios setups service dependencies are used to avoid these +but service dependencies have conceptual weaknesses +a notification may or may not be sent, depending on the timing of the both checks + +as an alternative solution Mathias suggests: +Throw counter wrap exceptions in case a dependent service fails +dependent services should get stale in this case +but this means a check needs to know on which other services it depends +for the named oracle checks this should work fine + +so let's try it this way +for other szenarios we might find a similar way