Finding and exporting your Snowflake data

What it holds

Snowflake stores relational warehouse data and semi-structured values. Current tables may be accompanied by historical versions in Time Travel, while Fail-safe serves a separate recovery purpose.

Where Snowflake shows the size

1. Open Snowsight Admin > Cost management > Consumption and select Storage.
2. Run SELECT table_catalog, ROUND(SUM(active_bytes)/POW(1024,3),2) AS active_gb, ROUND(SUM(time_travel_bytes)/POW(1024,3),2) AS time_travel_gb, ROUND(SUM(failsafe_bytes)/POW(1024,3),2) AS failsafe_gb FROM SNOWFLAKE.ACCOUNT_USAGE.TABLE_STORAGE_METRICS WHERE deleted = FALSE GROUP BY table_catalog ORDER BY active_gb DESC;
3. Inspect individual tables under Catalog > Database Explorer. DATABASE_STORAGE_USAGE_HISTORY provides daily database totals, including Time Travel in average_database_bytes.

Plan and admin role. ACCOUNTADMIN can inspect usage by default. Delegate with GRANT DATABASE ROLE SNOWFLAKE.USAGE_VIEWER TO ROLE your_role; cost pages also require the APP_USAGE_VIEWER application role for delegated access. Unloads need table, stage and warehouse permissions.

Exporting the full history

1. Create an internal stage with CREATE STAGE export_stage; or configure an external stage using a storage integration.
2. Run COPY INTO @export_stage/orders/ FROM mydb.public.orders FILE_FORMAT = (TYPE = PARQUET) HEADER = TRUE MAX_FILE_SIZE = 268435456;
3. For internal output, download through SnowSQL using GET @export_stage/orders/ file:///data/export/. External-stage output is already in cloud storage.

Snowflake storage is compressed. Account-usage database totals need not match billed account storage.

COPY INTO supports CSV, JSON and Parquet; JSON unloading requires VARIANT columns. CSV defaults to gzip and Parquet to Snappy; MAX_FILE_SIZE defaults to 16 MB and supports up to 5 GB.

Unloads consume warehouse credits and capture the queried state, not every historical version. Query AT or BEFORE for retained Time Travel states; users cannot query Fail-safe data.

You don't need to run an export to talk to us, and there's no need to clean it first — removing personal data is part of our work.