I have this simple model on a fresh new project:
class A
{
public int Id { get; set; }
public B B { get; set; }
}
class B
{
public int Id { get; set; }
public string Text { get; set; }
}
class C
{
public int Id { get; set; }
public B B { get; set; }
}
The mappings are:
modelBuilder.Entity<A>().HasOptional(x => x.B).WithOptionalDependent();
modelBuilder.Entity<C>().HasOptional(x => x.B).WithOptionalDependent();
I am running this query:
db.As.Include(a => a.B).Where(a => a.B.Text == "10").FirstOrDefault();
Where db.As
is the DbSet of A's.
It generates the following sql
SELECT TOP (1)
[Extent1].[Id] AS [Id],
[Join8].[Id1] AS [Id1],
[Join8].[Text] AS [Text],
[Join11].[Id2] AS [Id2],
[Join14].[Id3] AS [Id3]
FROM [dbo].[A] AS [Extent1]
INNER JOIN (SELECT [Extent2].[Text] AS [Text], [Extent3].[Id] AS [Id4]
FROM [dbo].[B] AS [Extent2]
LEFT OUTER JOIN [dbo].[A] AS [Extent3] ON ([Extent3].[B_Id] IS NOT NULL) AND ([Extent2].[Id] = [Extent3].[B_Id])
LEFT OUTER JOIN [dbo].[C] AS [Extent4] ON ([Extent4].[B_Id] IS NOT NULL) AND ([Extent2].[Id] = [Extent4].[B_Id]) ) AS [Join2] ON [Extent1].[Id] = [Join2].[Id4]
INNER JOIN (SELECT [Extent5].[Text] AS [Text], [Extent6].[Id] AS [Id5]
FROM [dbo].[B] AS [Extent5]
LEFT OUTER JOIN [dbo].[A] AS [Extent6] ON ([Extent6].[B_Id] IS NOT NULL) AND ([Extent5].[Id] = [Extent6].[B_Id])
LEFT OUTER JOIN [dbo].[C] AS [Extent7] ON ([Extent7].[B_Id] IS NOT NULL) AND ([Extent5].[Id] = [Extent7].[B_Id]) ) AS [Join5] ON [Extent1].[Id] = [Join5].[Id5]
LEFT OUTER JOIN (SELECT [Extent8].[Id] AS [Id1], [Extent8].[Text] AS [Text], [Extent9].[Id] AS [Id6]
FROM [dbo].[B] AS [Extent8]
LEFT OUTER JOIN [dbo].[A] AS [Extent9] ON ([Extent9].[B_Id] IS NOT NULL) AND ([Extent8].[Id] = [Extent9].[B_Id])
LEFT OUTER JOIN [dbo].[C] AS [Extent10] ON ([Extent10].[B_Id] IS NOT NULL) AND ([Extent8].[Id] = [Extent10].[B_Id]) ) AS [Join8] ON [Extent1].[Id] = [Join8].[Id6]
LEFT OUTER JOIN (SELECT [Extent12].[Id] AS [Id2]
FROM [dbo].[B] AS [Extent11]
LEFT OUTER JOIN [dbo].[A] AS [Extent12] ON ([Extent12].[B_Id] IS NOT NULL) AND ([Extent11].[Id] = [Extent12].[B_Id])
LEFT OUTER JOIN [dbo].[C] AS [Extent13] ON ([Extent13].[B_Id] IS NOT NULL) AND ([Extent11].[Id] = [Extent13].[B_Id]) ) AS [Join11] ON [Extent1].[Id] = [Join11].[Id2]
LEFT OUTER JOIN (SELECT [Extent15].[Id] AS [Id7], [Extent16].[Id] AS [Id3]
FROM [dbo].[B] AS [Extent14]
LEFT OUTER JOIN [dbo].[A] AS [Extent15] ON ([Extent15].[B_Id] IS NOT NULL) AND ([Extent14].[Id] = [Extent15].[B_Id])
LEFT OUTER JOIN [dbo].[C] AS [Extent16] ON ([Extent16].[B_Id] IS NOT NULL) AND ([Extent14].[Id] = [Extent16].[B_Id]) ) AS [Join14] ON [Extent1].[Id] = [Join14].[Id7]
WHERE (N'10' = [Join2].[Text]) AND ([Join5].[Text] IS NOT NULL)
which seems complicated to me. It even "joins" with C which I don't reference from A nor B and is not included in the query.
Now, the thing is: we have a large project with some very complex models, we are having some performance issues, when we profiled the queries we found that there were queries that needed to "include" a lot of entities and they generated SQL of more than 3k lines.
Is there something wrong in our mappings or our configuration? We have lazy loading and proxy creation disabled.
You can improve data access performance in Entity Framework Core in several ways. These include enabling eager loading, disabling lazy loading, using streaming instead of buffering, and disabling change tracking.
In LINQ, an inner join is used to serve a result which contains only those elements from the first data source that appears only one time in the second data source. And if an element of the first data source does not have matching elements, then it will not appear in the result data set.
A join clause takes two source sequences as input. The elements in each sequence must either be or contain a property that can be compared to a corresponding property in the other sequence. The join clause compares the specified keys for equality by using the special equals keyword.
The Join operator joins two sequences (collections) based on a key and returns a resulted sequence. GroupJoin.
I have no real answer for you, but with the same specifications I would have writen:
modelBuilder.Entity<A>().HasOptional(x => x.B).WithMany();
modelBuilder.Entity<C>().HasOptional(x => x.B).WithMany();
which gives the same database schema as yours.
Why would I have done that: because I often think in terms of 'what schema do I want ?', and with my practice, for this schema, the code is the one above.
BTW: resulting query for ctx.Set<A>().Include(x => x.B).Where(x => x.B.Text == "10").Select(x => x)
is:
SELECT
[Extent1].[Id] AS [Id],
[Extent3].[Id] AS [Id1],
[Extent3].[Text] AS [Text]
FROM
[dbo].[A] AS [Extent1]
INNER JOIN [dbo].[B] AS [Extent2] ON [Extent1].[B_Id] = [Extent2].[Id]
LEFT OUTER JOIN [dbo].[B] AS [Extent3] ON [Extent1].[B_Id] = [Extent3].[Id]
WHERE
N'10' = [Extent2].[Text]
but still here: why extend3 ?
Because Extend2 allows to filter the A with no B: that is inner join ... extend2
is a where (really clear, no ?)
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With