Basically, I want to monitor a failure queue(s) and if the count increases a threshold, I want to create an Alert.
1) The queues can be more than one.
2) The Alert: I want to send the data to Prometheus from where I'll read and display in Alerta.
I'm confused as there seem many ways and I don't know what the correct or best way. I followed this https://community.hortonworks.com/questions/172092/monitoring-nifi-back-pressure-threshold-and-trigg... for queue monitoring which works for a single queue but still, the Prometheus part is left.
Secondly, I read there are ReportingTasks in Nifi which can do something similar and report the matrices. e.g SiteToSiteStatusReportingTask
I'm not sure which approach to choose ReportingTask, create custom ReportingTask or the normal NIfi Workflow to monitor as in the article.
And Secondly how to send data to Prometheus. Some PostProcessor or again a reporting task or custom processor etc??
How would you tackle this and whats a good approach?