Catalog compaction

Catalog maintains a history of the evolution of its objects.  The history may grow over time, which leads to the occupation of a significant amount of storage space to keep the update log, the occupation of a significant amount of memory since every node in the cluster maintains its on-heap cache of version snapshots, and increased time of node startup since every node should restore version history and populate its on-heap cache. To address these issues, catalog compaction was introduced. The problem is, at the moment the compaction is not triggered at all, because truncation of the catalog's history may affect other components.

The goal of the epic collect together tickets on implement the triggers of the compaction