With DBT Cloud deprecating support for jobs running on versions earlier than 1.0, all DBT Cloud projects need to be upgraded from 0.x DBT versions by July 1st, 2022.
DBT has published a guideline to assist with upgrades to 1.0. Here are some additional tips to help with your migration:
- On your local dev environment, if you are running into issues with a direct upgrade from 0.x to 1.0, first uninstall all 0.x components which include dbt-core and the dbt-adapters and then install 1.0.0.
- Run a DBT model in order to ensure that your project compiles correctly without errors on 1.0 before pushing changes to the main branch.
- Add the new dbt_packages directory to the .gitignore file. If you chose to remove the previous dbt_modules directory from .gitignore, delete it before pushing changes.
- Nice to do: There are new default directories for seed files, tests and analyses. You may choose to rename your current directories to match the new default names for forwards compatibility.
- If you find that you need to upgrade from python 3.6, as it is no longer supported on DBT Core v1.0, please note that only versions 3.7, 3.8, and 3.9 are supported. You will get errors if DBT runs on python 3.10, such as ImportError: cannot import name xxxx from xxxxx