Previously, when writing a query against our data shares involving Usage Groups, the full Usage Group definition would be present in the query leading to increased query complexity and damaging readability.
with select start_day as day, sum_query_cost from ( select start_day as day, -- full usage group definition included case when warehouse_name in ('DBT_S', 'DBT_M') or user_name in ('SVC_DBT') then 'Analytics Engineering' when warehouse_name in ('Warehouse 1', 'Warehouse 2') or user_role in ('MARKETING_ANALYST') or database_name in ('MARKETING') then 'Marketing Team' when ... when ... else 'Uncategorized' end as usage_group, sum(sum_query_cost) as sum_query_cost from select_datashare.transformed.daily_workloads_usage_groups where start_day >= '2024-09-01' and start_day <= '2024-09-14' group by all ) where usage_group in ('Marketing Team') order by day desc nulls last
We've since created new views, which abstract away the Usage Group definition leading to simpler queries.
select start_day as day, sum(sum_query_cost) as sum_query_cost from select_datashare.transformed.daily_workloads_usage_groups where usage_group in ('Marketing Team') and start_day >= '2024-09-01' and start_day <= '2024-09-14' group by all order by day desc nulls last
Get up and running with SELECT in 15 minutes.
Gain visibility into Snowflake usage, optimize performance and automate savings with the click of a button.