Some cases not tracked to Langfuse

Self Checks

[X] This is only for bug report, if you would like to ask a question, please head to Discussions.
[X] I have searched for existing issues search for existing issues, including closed ones.
[X] I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
[X] [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:）
[X] Please do not modify this template :) and fill in all the required fields.

Dify version

0.6.15

Cloud or Self Hosted

Cloud, Self Hosted (Docker)

Steps to reproduce

Dify: 0.6.15 Langfuse: v2.66.0

We use Langfuse to track Chatflow. Chat is now combined with Knowledge.

There are cases where the content of the chat is tracked and cases where it is not tracked. Although such a case exists in the Cloud version, In the local version of Dify x Langfuse, the tracking works properly. Is this a bug?

Thank you.

✔️ Expected Behavior

No response

❌ Actual Behavior

No response

Hey @ymiyata910! 👋 I'm here to help you with any bugs, questions, or contributions you have for the project. Let's tackle this together!

There are a few potential reasons why Langfuse tracking might not be consistent in the cloud version of Dify compared to the self-hosted version:

Network and Security Configurations: The cloud version of Dify has different network and security configurations compared to the self-hosted version. The cloud version is hosted on AWS in the U.S. region, which might have different network policies and restrictions that could affect Langfuse tracking ^[1]^[2].
Error Logging: The cloud version of Dify generates logs for all interactions, including errors. Specific error codes or messages in the logs, such as error (level) and error (status_message), could indicate failures in Langfuse tracking. Reviewing these logs might help identify the issue ^[3]^[4].
Chat and Knowledge Combination: The implementation of chat and knowledge combination in Dify involves configuring context settings, recall strategies, and enabling citation and attribution. This configuration could affect how user interactions are processed and logged, potentially impacting tracking ^[5]^[6].
Version Differences: Ensure that both the cloud and self-hosted versions are running the same versions of Dify and Langfuse. Any discrepancies in versions could lead to differences in behavior.

Reviewing the logs for specific error messages and ensuring consistent configurations between the cloud and self-hosted versions might help resolve the issue.

To continue talking to Dosu, mention @dosu.

langgenius / dify