When I wrote about "not-so-well-defined alarming layer" that was as well about not-so-well-defined set of dependencies between triggers.
Just try to have look on for example OS Linux template from my repo (URL is below) and look on lis of templates on that template. You will fins=d many green annotations in most of the triggers in that template.
If more important trigger A is already active that and another trigger B (B depends on A) just changed state it causes not action not generated.
This is very effective mechanism provided by zabbix which allows in proper design reduce number of actions logged in action table.
Just try to have look on for example OS Linux template from my repo (URL is below) and look on lis of templates on that template. You will fins=d many green annotations in most of the triggers in that template.
If more important trigger A is already active that and another trigger B (B depends on A) just changed state it causes not action not generated.
This is very effective mechanism provided by zabbix which allows in proper design reduce number of actions logged in action table.
Comment