B-tree (regular) and inverted indexes never require explicit
optimization by the database administrator, as they are automatically
kept up-to-date (optimized) at every table modification
However, this is not possible for Metamorph indexes due to their fundamentally different nature. Instead, table changes are logged for incorporation into the index at the next optimization (index update), and Texis must linearly search the changed data until then. Thus, the more a table has been modified since a Metamorph index's last optimization, the more its search performance potentially degrades. When the index is re-optimized, those changes are indexed and merged into the Metamorph index, restoring its performance. A Metamorph index may be optimized in one of several ways, as follows.