How can I make large IN clauses more efficient in SQL Server?

Question

My current query runs very slow when accessing a DB with pretty large tables

SELECT * 
FROM table1
WHERE timestamp BETWEEN 635433140000000000 AND 635433150000000000
  AND ID IN ('element1', 'element2', 'element3', ... , 'element 3002');

As you can see the IN clause has several thousand values. This query is executed roughly every second.

Is there another way to write it to improve performance?

Answer 1

将IN的元素添加到索引临时（如果元素更改）或永久表（如果元素是静态的）中，并对其进行内部联接。

Answer 2

This is your query:

SELECT *
FROM table1
WHERE timestamp BETWEEN 635433140000000000 AND 635433150000000000 AND
      ID IN ('element1', 'element2', 'element3', ... , 'element 3002');

The query is fine. Add an index on table1(id, timestamp) .

Answer 3

The best answer depends on how those element ID listings are selected, but it all comes down to one thing: getting them into a table somewhere that you can join against. That will help performance tremendously. But again, the real question here is how best to get those items into a table, and that will depend on information not yet included in the question.

Answer 4

You should check your execution plan, I guess you could have a parameter sniffing problem caused by your between. Check if the actual rows are way off you expected values. And you can rewrite your IN to a EXISTS, which works inside like a INNER JOIN.