Start your Coalesce Implementation right - with these tips! 🤖

Aug '24 • General

Coalesce is one of the newer Data Warehouse Automation tools in the market, so maybe you find yourself in a position where you're freshly getting familiar with the tool.

Consider my following advices to make the start as smooth as possible:

↗️ Out-of-the-box before custom nodes:

Coalesce is fully in tune with all Snowflake functionality, through the use of templates that can be downloaded from the Marketplace. Such templates or 'User Defined Nodes' can also be modified to make the code even more specific for a particular requirement. Of course, before experimenting with the latest Snowflake functionality based UDN's, it is worth spending time getting to know how Coalesce works with its out of the box offerings.

🚔Spend enough time defining standards upfront:

Although any configuration can be changed inside Coalesce during the development workflow, I recommend that a couple of things should be defined before you start rolling out Coalesce enterprise wide. Topics like git strategy, Snowflake security strategy, storage locations, and naming standards are best to be defined, before all workspaces and branches are created. This helps keeping the impact of changes low.

🔀Do EVERYTHING to keep column-level lineage intact:

Coalesce has amazing lineage capabilities, especially on a column level. But lineage can only be populated when developers don't break the system. The system might be broken when using subselects, CTE's, or if column references are not created by the autocomplete functionality.

If you avoid these, you will be able to have full column level lineage, from Information Marts to source objects!

Do you have any other recommendations to new Coalesce users? Let me know!

2 comments