SQL Server Enterprise Edition is supported for online index rebuilds. So how does an online index rebuild work and how does it affect our queries?
Since it is online, SQL Server keeps the existing index available to the user, which means that it does not modify the existing index and its associated statistics. Then it is assumed that the approach is similar to SQL Server switch partition-minimizing data offline time. It should be Mr. Index and statistics, this time the new index and statistics are definitely allocated object_id, internal must be already visible, but for users outside access is not visible. After the index rebuild is complete and the new statistics are established, the Sch-x lock on the application table points to the new index for the existing index. The new indexes and statistics are visible after completion. It then re-compiles all queries that refer to this table.
So what about index reorganization? The work of index reorganization is to fill the end data of the index data as far as possible into the front empty page, it does not update the statistics as the index rebuilds, and will not cause the query to recompile.
As to whether this is related to the mode of isolation. Isolation mode is only related to data page access, how can it be related to statistical data?
Reference:
Would SQL Server use ' incomplete ' or ' dirty ' statistics during online index rebuild?
SQL Server->> Online index rebuilding (online indexing rebuild)