Hi,
We want to get alerted on anomalous behaviour in our 4XX errors. In the particular, the ones affecting out front end and middleware.
Constraints:
- We do not want to define 4XX errors as server-side as this will affect the failure rates and cause poor error detection to the mix of business/expected errors inherent to 4XX responses.
- We want to minimize the number of CSMs and Metric Events, including the volume of data ingested by a Metric Event
- Any solution has to tell us which application, service & HTTP response code is driving the issue (ie dimensions to split)
- Preferably we can drill down the specific request/endpoint
- Preferably, when an increase in 4XX errors occurs, a solution would be able to tell us what request/endpoint is driving the issue.
- Strong preference for anomaly detection vs static threshold
Any suggestions for the recommended solution to the above problem statement in Dynatrace Classic, we do not have Grail/DPS.
Thanks,
Hari
