Bedrock Zero-Code Instrumentation Example
=========================================
This is an example of how to instrument Bedrock calls with zero code changes,
using `opentelemetry-instrument`.
When examples are run, it exports traces and logs to an OTLP
compatible endpoint. Traces include details such as the model used and the
duration of the chat request. Logs capture the chat request and the generated
response, providing a comprehensive view of the performance and behavior of
your OpenAI requests.
Note: `.env <.env>`_ file configures additional environment variables:
- `OTEL_LOGS_EXPORTER=otlp` to specify exporter type.
Available examples
------------------
- `converse.py` uses `bedrock-runtime` `Converse API <https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_Converse.html>_`.
- `converse_stream.py` uses `bedrock-runtime` `ConverseStream API <https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_ConverseStream.html>_`.
- `invoke_model.py` uses `bedrock-runtime` `InvokeModel API <https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_InvokeModel.html>_`.
- `invoke_model_stream.py` uses `bedrock-runtime` `InvokeModelWithResponseStrea API <https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_InvokeModelWithResponseStream.html>_`.
Setup
-----
Minimally, update the `.env <.env>`_ file with your "AWS_SECRET_ACCESS_KEY",
"AWS_SECRET_ACCESS_KEY", "AWS_DEFAULT_REGION" and if you are using temporary
credentials "AWS_SESSION_TOKEN". An
OTLP compatible endpoint should be listening for traces and logs on
http://localhost:4317. If not, update "OTEL_EXPORTER_OTLP_ENDPOINT" as well.
Next, set up a virtual environment like this:
::
python3 -m venv .venv
source .venv/bin/activate
pip install "python-dotenv[cli]"
pip install -r requirements.txt
Run
---
Run the example like this:
::
dotenv run -- opentelemetry-instrument python converse.py
You should see a poem generated by Bedrock while traces exported to your
configured observability tool.
95873604ab