Snowflake Time Travel and Fail-safe

Pia Riachi
3 min readMay 2, 2022
Photo by Jocke Wulcan on Unsplash

Time Travel is an interesting and powerful feature of Snowflake that makes it possible to access data at a specific point in the past.

This feature comes in very handy in case there are accidental updates on data that you would want to revert. Similarly, if you accidentally drop an object (table, view, database, schema, etc…), you can easily UNDROP it and return to the previous state. You also have the ability to clone databases, schemas, or tables at a specific time.

Another interesting use of time travel is to query the data at different points in time for analytical purposes.

Snowflake Time Travel

How does Snowflake time travel work?

Snowflake seamlessly saves the state of the data before performing any operation on it.

What is the retention period for Snowflake time travel?

Depending on your Snowflake subscription. For a Snowflake Standard account, the retention period is 1 day only. However with an enterprise version of Snowflake, the retention period can be configured and can be anywhere between 0 days and 90 days for permanent tables. For transient and temporary tables, the retention period is 1 day (can be dropped to 0 days). It is also possible to…

--

--

Pia Riachi

Engineer @Google | Advertising Solutions Engineering | Business Intelligence | Data Engineering