Showing results for 
Show  only  | Search instead for 
Did you mean: 

Is it possible to define incident rules only for certain BT splittings?


Assuming I have a BT which calculates response times and creates splittings based on method argument values (as an example). Now, I would like to define incident rules on the response times but with different violation thresholds depending on the splitting. Is this possible without having to create separate BT's each one filtering a particular method argument value?


Hi Enrico,

In short: it is currently not possible to create incident rules based on a BT split. There is an RFE: Feel free to upvote :-).

As a possible workaround, you could create a separate BT for each "split" and create incident rules on those.

Since it is on response time in your case, you could add the BT to the monitoring dashlet and define individual thresholds for each split. However then the baseline engine will take over and wil require significant measurements. It is also not possible to define which aggregation (max/min/avg/...) you want to use and which evaluation timeframe.

Hope this helps.


Thanks, I suspected this was not currently possible...

Anyway, I upvoted the RFE - that would be a highly useful addition!

Alternatively (and perhaps easier to implement) it would also be useful to be able to include matching splittings into account in the incident conditions (in order to limit the incident only to certain splittings). But this would still have the drawback of having the same threshold values applied to all splittings...

I think from this perspective the core design decision of defining threshold values as an attribute of the measure definition seems a bit unfortunate (although I'm sure there are valid reasons for it). IMHO threshold values should have been part of the incident definition from the start (at least I think that's where one would naturally expect them to be defined).

kind regards,