使用INNER JOIN作为过滤器的性能和优点/缺点

Question

I have been working on MS SQL Server and has been following standard SQL Query to do INNER JOIN and filter the resultset based on the criteria.

Recently, I was working with a large table where it has somehow cause the whole query to take ridiculous long time to do a query.

Below is just a mock query of the situation.

Select * FROM Customer a
INNER JOIN ProcessTable PT WITH(NOLOCK) on PT.custID = a.ID
INNER JOIN AuditTrailTable AT WITH(NOLOCK) on AT.custID = a.ID
WHERE AT.DATESTAMP > 'Some date time'

My colleague has come up with a new query as below

Select * FROM Customer a
INNER JOIN ProcessTable PT WITH(NOLOCK) on PT.custID = a.ID
INNER JOIN AuditTrailTable AT WITH(NOLOCK) on AT.DATESTAMP > 'Some date time' AND AT.custID = a.ID

in which I observed he has put the filter clause as part of the INNER JOIN.

Despite it returns the same result, I would like to know the performance gain of this approach and pros and cons that I have not aware of please. Will it be a good approach of using this?

Answer 1

The two should be exactly the same. SQL Server's compiler will look at the queries and realize they are describing the same result set. It will then optimize the query plan based on the data.

Whether the condition goes in the on clause or the where clause does not matter for an inner join . (It does matter for an outer join .)