Skip to content

Latest commit

 

History

History
159 lines (98 loc) · 3.42 KB

quickstart.md

File metadata and controls

159 lines (98 loc) · 3.42 KB
Raw

Quickstart Guide for SnowDev

Welcome to SnowDev's quickstart guide! In just a few minutes, you'll be up and running with SnowDev, ready to deploy components related to Snowflake using Snowpark.

Installation

Setup Python Environment

Ensure you have Python version 3.10 and above.

pyenv install 3.10.0 
pyenv local 3.10.0

Install SnowDev

You can install SnowDev using pip or add it with poetry.

pip install snowdev

Initialize SnowDev

Start your SnowDev journey by initializing the directory structure for deployment, this will create necessary folders src and udf, sproc, streamlit components

snowdev init

First Deployment with SnowDev

Now that you have SnowDev installed and initialized, it's time to see it in action! Let's walk through the deployment of various components.

Deploying Stored Procedures

  1. Add a New Stored Procedure

    Add a new stored procedure named test_script:

    snowdev new --sproc test_script

  2. Test the Stored Procedure Locally

    Now, test your stored procedure locally to ensure everything works as expected:

    snowdev test --sproc test_script

  3. Deploy the Stored Procedure to Production

    Satisfied with the local tests? Let's deploy it to production:

    snowdev deploy --sproc test_script

Deploying User-Defined Functions (UDFs)

  1. Add a New UDF

    Create a new UDF named predict_sentiment:

    snowdev new --udf predict_sentiment

  2. Test the UDF Locally

    Before deploying, test the UDF locally to ensure it behaves as intended:

    snowdev test --udf predict_sentiment

  3. Deploy the UDF to Production

    Once local testing is complete, deploy the UDF to production:

    snowdev deploy --udf predict_sentiment

Deploying Streamlit Applications

  1. Add a New Streamlit Application

    Create a new Streamlit application named order_chart_app:

    snowdev new --streamlit "order_chart_app"

  2. Deploy the Streamlit Application to Production

    Ready to go live? Deploy the application to production:

    snowdev deploy --streamlit "order_chart_app"

Deploying Tasks

  1. Add a New snowflake Task

    Create a new task application named sample_task:

    snowdev new --task "sample_task"

    Modify the sql in the src/task/sample_task/app.sql

  2. Deploy the task Application to Production

    Ready to go live? Deploy the application to production:

    snowdev deploy --task "sample_task"

  3. Resume the Task

    Resume the task so it starts the schedule:

    snowdev task --name "sample_task" --action resume

Deploying using AI

  1. Initialize by embedding the knowledge

    This embeds the knowledge base into a in memory chroma db vector DB:

    snowdev ai --embed

  2. Deploy a new stored procedure using AI

    Give a clear description of what you need the stored procedure to do:

    snowdev ai --sproc "I want to fetch data from order table and predict the bad orders using snowflake ml"

Congratulations! You've successfully deployed various components using SnowDev. Dive deeper and explore other commands and options mentioned in the main README to make the most out of SnowDev.