Hi community
I'm trying to generate reasonable problems from event data read from website.
Input data is formatted like this:
2021-11-18T17:05:01.001+02:00 | Event X
2021-11-18T17:05:30.057+02:00 | Event Y
2021-11-18T19:08:30.999+02:00 | Event Z
When something happens a new row is being made and when situation gets back to normal that row is just being removed.
In normal condition input data has value 'N/A'
Example timeline:
N/A
.
.
2021-11-18T17:05:01.001+02:00 | Event X ----> generate problem with Operational data: Event X
.
.
2021-11-18T17:05:30.057+02:00 | Event Y ----> generate problem with Operational data: Event X and Event Y
.
.
2021-11-18T19:08:30.999+02:00 | Event Z ----> generate problem with Operational data: Event X, Event Y and Event Z
.
.
.
Event X fixed ---> No recovery, no new problem
.
.
Event Z fixed ---> No recovery, no new problem
.
.
Event Y fixed ---> Input data -> 'N/A' and that will recover all three problems.
I have 2 items.
One holds above mentioned event data as text (event.data).
Another dependent item holds timestamp of the latest event as unixtime (event.data.last.ts).
Trigger is being fired when timestamp of the latest event gets bigger (new event has occurred).
Event data is involved also in trigger expression because that is used as Operational data.
Problem expression:
{Event test:event.data.last.ts.change()}>0 and {Event test:event.data.strlen()}>3
Recovery expression:
{Event test:event.data.strlen()}<5
The problem is that I will always get duplicate problems.
That's because trigger is being evaluated twice (first when master item gets data and secondly when dependent item gets data).
Would there be any way to avoid duplicate problems?
I just can't generate problem when input data changes because new problem is not needed when some of event is being recovered.
I'm trying to generate reasonable problems from event data read from website.
Input data is formatted like this:
2021-11-18T17:05:01.001+02:00 | Event X
2021-11-18T17:05:30.057+02:00 | Event Y
2021-11-18T19:08:30.999+02:00 | Event Z
When something happens a new row is being made and when situation gets back to normal that row is just being removed.
In normal condition input data has value 'N/A'
Example timeline:
N/A
.
.
2021-11-18T17:05:01.001+02:00 | Event X ----> generate problem with Operational data: Event X
.
.
2021-11-18T17:05:30.057+02:00 | Event Y ----> generate problem with Operational data: Event X and Event Y
.
.
2021-11-18T19:08:30.999+02:00 | Event Z ----> generate problem with Operational data: Event X, Event Y and Event Z
.
.
.
Event X fixed ---> No recovery, no new problem
.
.
Event Z fixed ---> No recovery, no new problem
.
.
Event Y fixed ---> Input data -> 'N/A' and that will recover all three problems.
I have 2 items.
One holds above mentioned event data as text (event.data).
Another dependent item holds timestamp of the latest event as unixtime (event.data.last.ts).
Trigger is being fired when timestamp of the latest event gets bigger (new event has occurred).
Event data is involved also in trigger expression because that is used as Operational data.
Problem expression:
{Event test:event.data.last.ts.change()}>0 and {Event test:event.data.strlen()}>3
Recovery expression:
{Event test:event.data.strlen()}<5
The problem is that I will always get duplicate problems.
That's because trigger is being evaluated twice (first when master item gets data and secondly when dependent item gets data).
Would there be any way to avoid duplicate problems?
I just can't generate problem when input data changes because new problem is not needed when some of event is being recovered.