Unused tables

Your warehouse might not be top of mind when it comes to Spring cleaning, but our latest feature update might make you reconsider your (data) hygiene priorities. The newest addition to Data Insights comes in the form of an easy-to-use report of tables in your warehouse that aren’t being used. You'll see two lists when you click into the tab: completely unused tables and tables with writes but no reads.

Completely unused tables

We’ve all been there. Maybe you created a staging table for merging a data source that’s no longer being used. Maybe you created a table last year to track metrics for a now-finished product rollout. It can often be worth keeping those tables around just in case, but at a certain point, it’s not worth the storage costs any longer. Now you can know exactly which tables to clean up when you’re looking to optimize.

Tables with writes but no reads

Maybe your team leaned into “storage is basically free” a bit too much when setting up your ELT strategy. In any case, new members of the team are now getting confused about whether to reference “salesforce.customer_health_metrics” or “salesforce.customer_health_metrics_updated” when building a new table, and it’s time to clear up the confusion.

To support your ongoing quest for data hygiene, you’ll also be able to drill into what queries are updating tables with writes but no reads. Who knows–you might even be able to save a few bucks after disabling those tables.

Export to CSV

The tables in Data Insights provide a great starting point for looking into your warehouse performance and optimizations. 

When you’re ready to drill in further to understand performance and cost, Data Explorer gives you options including:

• Filters by specific time ranges, users, roles, and more to help you isolate specific incidents in your data
• Group by options can be thought of as “dimensions” and include users, roles, and compute resources to help you optimize cost or performance for a particular aspect of your warehouse
• Aggregation options can be thought of as “metrics” and include almost every performance-related metadata field from your warehouse, including bytes scanned or credits used. These can be used to sort your dimensions listed above.

Our latest addition to data insights is the ability to export these tables to CSVs. This ability makes it easier to:

1. Share your findings with a broader team–for example, imagine you’re pulling a list of the most expensive queries additionally grouped by workload-dependent warehouses. You’ll then be able to share this list with analytics teams using different warehouses to help them understand where they can focus their spend optimization efforts.
2. Expand your scope–in the UI, you’ll be able to see the top 50 rows for cost and performance optimization. Exporting to CSV gives you the ability to look into your top 5000 queries, tables, users, roles, and/or almost any other breakdown you’re interested in.