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

Service anomaly detection - unusual drop of responce time and throughput (any ideas welcomed!)

imoskovko1
Inactive

Hello!

There was sudden service interruption to external unmonitored host that we can not instrument. Exchange server mailbox emails parsing got stuck by this service.

How to set up anomaly warnings on such behaviour? I am not sure that i see answer at Settings --> Anomaly detection ---> Services

So far i have created custom alert on this service static threshold, but looking the way to catch anomalies on service responce time and throughput.

There are some screenshots:

Regards,

Igor


3 REPLIES 3

aaltuhov
Newcomer

Hi Igor,
not sure whether that's the best option, but..:
In the SaaS Dynatrace we have this option in the Anomaly Detection for Applications.
So you would have to create a separate application in Dynatrace to include you service's requests and then set up the rules for the traffic spikes/drops of anomaly detection settings of your new application or in global Anomaly Detection settings.
dt-app-anomaly-traffic.png
Hope that helps,
Alex


Hi Alex,


I have a similar issue, key requests coming from Deep monitoring on a jetty container where the throughput suddenly disappears.

I want to monitor these in drop in throughput.


How can I create a custom application and monitor the jetty key requests from there?


Thanks,

Patrick


imoskovko1
Inactive

Thank you Aleex.

I need to think. Honestly do not wand to introduce new top level application, cause it is only small service that every minute check Exchange mailbox for new incident related mails. But if it gets stuck SD people missing incidents logged via emails.

Are there any other ideas?

So far i have created custom alert on this service parameters with static threshold. But actualy looking for way to catch anomalies.

Regards, Igor


Featured Posts