Azure / Azure/azure-sdk-for-python

Agent Monitoring Continuous Evaluation for Custom Agents Failing with error "No trace data found"

Aperta
#47,812 4 commenti 1 reazione 1 assegnatario Assegnata a @posaninagendra Vedi su GitHub
AI Projects Client customer-reported needs-team-attention question Service Attention
Lingua principale
Python
Stelle
5.6k
Fork
3.4k
Merge medio
2g 2h
PR unite (30g)
213

Descrizione

- **Package Name**: Agent Monitoring Continuous Evaluation
- **Package Version**:
- **Operating System**:
- **Python Version**:

**Describe the bug**
We have setup continuous monitoring via the UI for our custom agent, however evaluation run fails with this error even though traces exist:

**There was an error with this evaluation run.No trace data found where customDimensions.gen_ai.input.messages or customDimensions.gen_ai.output.messages contain data. Please check the agent running status for agent 'ifd-mono' during time range: from 2026-07-02T05:50:08+00:00 to 2026-07-02T06:50:08+00:00**

Image

Image

Image

**To Reproduce**
Set up monitoring for your custom agents
1. Onboard your custom agent to Foundry using the instructions in [Register and manage custom agents](https://learn.microsoft.com/en-us/azure/foundry/control-plane/register-custom-agent).
2. Instrument your agent to comply with the [semantic conventions for generative AI solutions in the OpenTelemetry standard](https://opentelemetry.io/docs/specs/semconv/gen-ai/).
3. Configure your agent to send telemetry to the same Application Insights instance as your Foundry project to enable continuous evaluation features.
4. In Foundry Control Plane, go to the Asset page and select your agent.
5. Select the Monitor tab to view your metrics and charts.
6. Set up continuous evaluations using the methods outlined in [Configure settings](https://learn.microsoft.com/en-us/azure/foundry/observability/how-to/how-to-monitor-agents-dashboard?tabs=python#configure-settings).

**Note: We only implemented the configure settings via the UI and didn't create the python implementation as we understood it achieves the same thing?**

**Expected behavior**
A clear and concise description of what you expected to happen.
Evalution result should appear in the Evaluation page successfully

**Screenshots**
If applicable, add screenshots to help explain your problem.

**Additional context**
See user_log attached

[user_logs (12).txt](https://github.com/user-attachments/files/29587816/user_logs.12.txt)

Guida per i contributori

Apri la guida per i contributori

Valutazione

Questa issue non è ancora stata valutata.

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.