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

Disabling anomaly detection for set of services across multiple application

suresh230591
Advisor

Hi,

Is there a way to disable the anomaly detection for set of services across multiple application? If we disable it from global setting it will disable for entire environment.

Any chance of using tag or services IDs or Service name through configuration API.


Thanks

Suresh




4 REPLIES 4

You can only start maintenance window for them.

https://www.dynatrace.com/news/blog/manage-planned-downtimes-maintenance-windows/

You can disable alerting or even whole problem generation (not only alerts will not be sent via integrations, but problems wont be fired at all). But It's nor recommended to putting some things out of scope because of potential correlation between issues on services you want monitor and you don't. If you have to many alerts from some portions of services I suggest tweaking anomaly detection settings for them. Remember that after week dynatrace should learn failure rate level and other important metrics for those services and stop alerting if this is normal.

Sebastian


Keep maintenance window for life long time does not look ideal solution. In our case app team do not bother about transaction monitoring as this is PaaS service and vendor is doing that so all the host are monitored in Infra only mode.

We were under the impression that Infra mode also stop detection of services on PG. However, we could see services are detected and purepath data seen.


Thanks

Suresh


Infra mode indeed is disabling services but you may see opaque services. This is something different. When monitored entity is reaching process working on host in infra mode, dynatrace may create opaque service. It's not working on this process (infra only), it's virtual representation of reaching this process by another monitored in full stack. If you see something like this you shouldn't bother about it. In general there you will see single element purepaths with only information about time, status code and error message (if it was thrown on response).

Sebastian


You are right, this is opaque service, another service from a full stack monitored host is calling this process. So this is only detected as long as we are receiving requests to underlying process.

In order to disable the anomaly then we may have to turn this off from GUI for individual services.


Thanks

Suresh