09 May 2025 01:03 PM
Hello Team,
Currently our applications are monitored by Dynatrace and the problem tickets are being raised in Dynatrace . We are trying to develop an automated Agentic AI SRE assistant for resolving application problems . The assistant will action programmatically what an SRE will do manually That is Login to Dynatrace UI and based on analysis of the error details available in Dynatrace UI, perform some actions to resolve the issue.
Example – Dynatrace screenshot below showing OOM(Out Of Memory)
Our ask is the feasibility of programmatically extracting these error details(OOM ) that we see in Dynatrace UI through API calls to Dynatrace
09 May 2025 02:33 PM - edited 09 May 2025 02:37 PM
Hi @SaiVaarmah
You can utilize /api/v2/problems and use problemSelector with text("Out-of-memory kills") that should work
From UI
From API
HTH
Yos
09 May 2025 08:39 PM
Are you using any incident management platforms like ServiceNow? If you are, you can send these Problem card alerts right into you IM platform, from there it/Dynatrace can kick off your automation, be it Dynatrace Workflows, Ansible etc... then ServiceNow can post comments in the Problem card ID by way of the API as well as manually close out these problems all via the API integration with the data you want to include as to what actions were performed etc.
This way its all recorded and traceable, from Event start, to notice, to automation, to closure.