Specify "no data" as a different type of alert for routing | Archive

Antony_Hawkins · 2014-06-09T00:00:00Z

It would be useful in some cases if "No Data" alerts from a datapoint could be routed via a different Alert Rule and to a different Escalation Chain than that specified for threshold-breaching alerts. Examples: 1. A custom datasource recording values from an application database may be required to alert the database / application manager, or application developer, if alerting thresholds are breached. However, if "no data" is being returned, this may indicate a problem with the datasource query rather than with the application, and it may be that a different person or team needs alerting in order to modify the datasource, reconfigure connections and credentials from the LogicMonitor collector, etc. 2. An MSP may want to pass alerts on various metrics to their customers, while "no data" may indicate a connectivity or setup problem within LogicMonitor, when the MSP themselves may need to be alerted, without troubling their customers. Currently this is only possible if a datapoint is set to alert on "no data" only, i.e. is not used for other, threshold-based, alerts. This could be changed via the provision, in Alert Rule configuration, of a "no data" filter, such that a rule could be created to only catch "no data" alerts from e.g.: "Group X / Host Y / Datasource Z / No Data" ...leaving alerts from datapoints breaching thresholds to pass to another rule. The filter would allow users to segregate "no data" and "threshold breaching" alerts, or include both alert types in a rule.

MoreDiscussions

Specify "no data" as a different type of alert for routing | Archive | LogicMonitor Community