Extracting anomaly detection setting for each service

[pshinde2]

Hi , We have made threshold changes for response time and failure rate alerting for various services since the time of deployment. Now we wanted to override those custom settings with global default values . Problem here is that the services which are using custom/fixed values did not take global defined values and we did not kept the track of those when we made threshold changes to various services individually . Is their any way or any API call available to pull the details which will show if service is using global threshold values or custom values. This details will us to make required changes to all those services to use global values instead of fixed values.

[skrystosik]

Unfortunately anomaly detection API is only for global settings, not for those related to particular entities.

Sebastian

[Julius_Loman]

It's not in the Config API yet. If you know your service entity IDs, it's possible to emulate requests from UI and call the same rest API requests the UI is calling. It's nasty, unsupported, may change between Dynatrace updates, but it's the only way to automate it until the anomaly detection is part of config API.

[wolfgang_beer · Accepted solution]

A simple workaround would be to define a custom event on a selected service or service method with a threshold and use its customMetricEvent API endpoint to update that config.

Custom event configs are fully configurable on instance level as well as on group level and available through the API.

Best greetings,

Wolfgang

[pshinde2]

Hi Wolfgang ,

Thanks , Currently our concern is to figure out which all services using custom/fixed values as threshold . We don't have exact record of service which are using those custom thresholds. It would be great if Dynatrace provide some solution around it . Atleast list of services using Global values and list of services not using global anomaly settings.

[diana]

@Wolfgang B. Any idea when the anomaly detection API endpoints will be available at the service level?

[tibebe_m_digafe]

Hello,

This is along the line of "enabling DevSecOps teams" to  handle their own monitoring settings instead of handing off that responsibility to a very few folks administering Dynatrace.   This is not just for anomaly detection but to all that can/are scoped to MZ (in our case, MZ identifies the "tenant" in a single DT environment. We can't do environment per tenant due to shared AWS and k8s resources across tenants).

Has there been any plans to make this at entity/service level (access based on MZ) instead of "globally". Obviously, entities/services are NOT global.

Thanks. Tibebe