ColumnStore index fragmentation

Copper Contributor

In an existing clustered columnstore index expanding one of the existing cols. - from int to bigint - with values remaining same - would it lead to any fragmentation in the existing data that could cause any perf. issues for queries.

 

The table is mostly a read-only one with data being added every day and is partitioned month-wise.

 

The existing columnstore table is quite big (approx 400+ GB) already.

 

The need to expand the col. from int to bigint is due to the values for the new data rows coming from an upstream application will roll-past the INT value limit.

The INT col. in question is a non-key column

1 Reply