azure 托管实例 sql server 数据库上的基本查询非常慢

Question

On an Azure managed sql server instance database, I do

Select top 1000 * from _CSP_Detail.

The data table is empty because I deleted all data from the table yesterday (Delete * from _CSP_Detail).
It used to have a million rows, and it took a while to delete. There are a hand full of indexes / indices defined on the table.

After doing the delete, today, I am finding that

Select top 1000 * from _CSP_Detail

does pull nothing, as expected, but takes over a minute to run.

Puzzled. Any ideas?

Answer 1

It seems your table is a heap (no clustered index). Space is not reclaimed when you delete rows from a heap so, even though the table is empty, the allocated pages must be scanned by the SELECT query to find rows.

Explicitly TRUNCATE the table or create/drop a clustered index to reclaim space and improve performance against the empty table.

Note that tables should generally have a clustered index according to the documentation .