Streamlit pyarrow error on connecting to snowflake

Summary

I have installed the pre requisites using the snowflake link
https://quickstarts.snowflake.com/guide/frosty_llm_chatbot_on_streamlit_snowflake/index.html?index=..%2F..index#1

And when I tried to connect to snowflake through streamlit using the following code, I got pyarrow error:

Steps to reproduce

Code snippet:

import streamlit as st

conn = st.experimental_connection("snowpark")
df = conn.query("select current_warehouse()")
st.write(df)

If applicable, please provide the steps we should take to reproduce the error or specified behavior.
https://quickstarts.snowflake.com/guide/frosty_llm_chatbot_on_streamlit_snowflake/index.html?index=..%2F..index#1
Expected behavior:

I should see the following on browser
CURRENT_WAREHOUSE()
COMPUTE_WH

Actual behavior:

Error Message:
Failed to import ArrowResult. No Apache Arrow result set format can be used. ImportError: DLL load failed while importing arrow_iterator: The specified procedure could not be found.

Debug info

  • Streamlit version: Streamlit, version 1.25.0
  • Python version: Python 3.10.12
  • Using Conda? Yes
  • OS version: Edition Windows 10 Pro for Workstations
    Version 21H2
  • Browser version: Chrome Version 116.0.5845.97 (Official Build) (64-bit)
  • Python packages
    pyarrow 10.0.1
    snowflake-connector-python 3.1.0
    snowflake-snowpark-python 1.6.1
    streamlit 1.25.0

Hey @reema_dangwal,

Thanks for sharing this question!

What happens if you uninstall snowflake-connector-python? You should only need the Snowpark, Streamlit, and OpenAI packages for that quickstart, so I’m wondering if the Snowflake connector might be contributing to that error…

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.