Chapter 1: Introduction to Session Mode

Google has introduced Session Mode in BigQuery, allowing users to manage their sessions either automatically or manually. This feature enables the labeling of all queries within a session and provides access to session metadata through the INFORMATION_SCHEMA.SESSIONS_BY_* views. With Session Mode, users can utilize temporary tables and variables throughout the session, facilitating the interactive creation of multiple queries. Moreover, multiple sessions can run concurrently, and each session's history is preserved for up to 20 days after it concludes.

Chapter 2: How to Enable Session Mode

To begin using Session Mode, first select the Processing Location where the queries will execute. Once a session is established, all queries will be confined to that location, which cannot be altered thereafter. In the Session Management section, simply check the "Use session mode" option and click Save. It’s that straightforward!

Chapter 3: Practical Applications of Session Mode

Google outlines several key advantages of using Session Mode, which I fully support:

1. Management of Temporary Session Data: You can define variables and temporary tables once and utilize them throughout the session.
2. Query History Tracking: For behavior analysis at specific times during the session, the history of changes can be retrieved easily.
3. Multi-Statement Transactions: Within a session, you can initiate a transaction, make modifications, and observe temporary results before deciding to commit or roll back. This flexibility allows for multiple queries without the need to complete everything in a single statement.

Chapter 4: Conclusion

As both a developer and analyst, I find this new feature highly beneficial. Additionally, you might be interested in other recent updates, such as:

• BigQuery's support for JSON as a data type
• Utilizing WITH RECURSIVE in BigQuery SQL
• Implementing Materialized Views without aggregation in BigQuery
• Leveraging Remote Functions in BigQuery