Invalid instrumentation key

[tyler-suard-parker]

Describe the bug

I am following this tutorial: https://learn.microsoft.com/en-us/azure/machine-learning/prompt-flow/how-to-trace-local-sdk?view=azureml-api-2&tabs=python

I logged in using azure cli and then entered the command to set up my trace destination:
pf config set trace.destination=azureml://subscriptions/<your_subscription_id>/resourcegroups/<your_resourcegroup_name>/providers/Microsoft.MachineLearningServices/workspaces//<your_azureml_workspace_name>

This created a CosmosDB. I promptflow does not do this every time I run it on different machines.

I tried to send traces using the trace-autogen-groupchat.ipynb, and I got this error:

ERROR:opencensus.ext.azure.common.transport:Non-retryable server side error 400: {"itemsReceived":2,"itemsAccepted":0,"appId":null,"errors":[{"index":0,"statusCode":400,"message":"Invalid instrumentation key"},{"index":1,"statusCode":400,"message":"Invalid instrumentation key"}]}.
ERROR:opencensus.ext.azure.common.transport:Non-retryable server side error 400: {"itemsReceived":1,"itemsAccepted":0,"appId":null,"errors":[{"index":0,"statusCode":400,"message":"Invalid instrumentation key"}]}.

How To Reproduce the bug
Steps to reproduce the behavior, how frequent can you experience the bug:

1. Download the trace-autogen-groupchat.ipynb notebook from this repo
2. Follow the instructions for setting the trace destination https://learn.microsoft.com/en-us/azure/machine-learning/prompt-flow/how-to-trace-local-sdk?view=azureml-api-2&tabs=python
3. Try to run the code in the autogen section, where agents generate messages and send them using tracing.

Expected behavior
I expected the information to be sent without any error.

Using Windows, Python 11, and the latest version of PromptFlow.

[zhengfeiwang]

Hi @tyler-suard-parker , thank you for reaching out. I think there are several questions in this issue:

• pf config set trace.destination does create a Cosmos DB, but one workspace will only have one Cosmos resource, so I understand when you run same command on different machines, only the first time you will see the Cosmos setup process (and wait for that).
• For the error, it's wired to me because we did use opencensus before, but in Migration OpenCencus to opentelemetry #2175 , which is merged 3 months ago, we already migrate to Open Telemetry. Could you please share some screenshots for your error? and provide your prompt flow version with pf -v? so that maybe we can re-produce your error and better investigate on that.

[github-actions]

Hi, we're sending this friendly reminder because we haven't heard back from you in 30 days. We need more information about this issue to help address it. Please be sure to give us your input. If we don't hear back from you within 7 days of this comment, the issue will be automatically closed. Thank you!