Skip to content

Understanding Data Partitioning in Tracardi

Data partitioning is a database management technique where a large dataset is divided into discrete segments, or partitions, which can be managed and accessed independently. This strategy is particularly useful for improving performance and managing data at scale.

Tracardi, leverages data partitioning to handle its indices efficiently. Indices, which are essentially the databases for storing various types of data, can grow to be very large in volume. When such large indices are divided into smaller, time-based indices, it improves data manageability significantly. This approach facilitates:

  • Improved Performance: Smaller indices are faster to read and write to, which enhances the overall system performance.
  • Easier Data Management: It's simpler to manage and archive data based on time-based criteria. For example, older data that is less frequently accessed can be moved to cheaper, slower storage.
  • Efficient Deletion: Purging old data becomes more straightforward and less resource-intensive when it's contained within specific time-based partitions.

In Tracardi, this partitioning method is applied to:

  • Events: User actions and interactions that are recorded by the system.
  • Profiles: Information about users, such as demographic data and preferences.
  • Sessions: Data pertaining to a user's specific interaction session with a platform.
  • Logs: System logs that record various events and transactions for monitoring and debugging purposes.