Managing extents
In Chapter 1, Introducing Azure Synapse Data Explorer, you learned about the architecture of Data Explorer pools and how it uses database sharding techniques to split tables and physically store them on cluster nodes. Database sharding is broadly used by database management systems to split large tables (or databases) into smaller, more manageable files, offering performance that scales linearly as you add more compute and removing limitations that would arise from managing very large individual files to persist data on disk.
In Data Explorer, each of these data shards is called an extent. When you create a new table and ingest data, the Data Explorer engine creates and distributes your table into a series of extents across the cluster. These extents contain not only your data, but also some associated metadata that indicates the creation time of the extent, some optional tags (which we will discuss shortly), and information that may help the Data Explorer engine...