Checkpoints for database tree rebalancing

The internal database stores its data in structures called B-trees. While B-trees are a very common way to structure data for rapid access, over time, the process of adding and deleting elements to and from the trees can eventually lead to imbalances in the data structure.

Eventually, the tree can become sufficiently unbalanced that performance is degraded. The Helix Server checkpoint and restore processes (see Backup and recovery concepts) re-create the trees in a balanced manner, and consequently, you might see some improvement in server performance following a backup, a removal of the db.* files, and the re-creation of the db.* files from a checkpoint.

Given the length of time required for the trees to become unbalanced during normal Helix Server use, we expect that the majority of sites will never need to restore the database from a checkpoint (that is, rebalance the trees) to improve performance.

(The changes to the B-trees between Helix Server 2013.2 and 2013.3 require that any upgrade that crosses this release boundary must be performed by taking a checkpoint with the older release and restoring that checkpoint with the newer release. See the topic on upgrading From prior to 2013.3.)