We have trigger dependency to configure exact other trigger in relation.

We have feature for tagging triggers with dynamicaly added name and/or value to tags,

Add posibility to define dependent trigger key using compose with own macros also not only in host-item params.

find(/{{ITEM.VALUE1}.regsub(".,.,(.),.", "\1")}/globalLogAlert,,"regexp","New")=1 

Simplified use case:

We have definned many common hosts and all they with metrics from templates. We have one other cental server which logs prolems with information of related common host.

If common host identified problem and populate it to service, we don’t need to alert problem from central server log. But if central server log find problem for common host which looks OK without problems (or is for just new host-item-trigger) , then we need that alarm to see.

Workflow

If by test process of main trigger expression, in first internal step fires Problem (after receive new item.value) ,

1. then in next step is checked status of dependency.trigger(lookup for dependent.trigger.key)
2. and than … trigger tags are composed for event.tags (if dependent.trigger is not in problem) and event is created with problem (and …) .

but if the second step looks like

1. then in next step firstly evaluate expression to compose key of depentent trigger, update this key for (next created) event.dependent.trigger.key and then that event.dependent.trigger.key is located for check dependency trigger status …
2. …

How we should search dependencies?

In that steps (---^) I tell to use small preprocess for compose dependent key before it used to search dependencies.

Where that dynamic dependency would be defined?

… some preprocessing steps added before lookup

Do you want to suppress triggers in case of dependency or make it OK?

Good idea for customize decision as needed in other scenarios. But for me it could be just default OK status.