https://github.com/bbonkr/sample.ef.batch

Delete rows batch using Entity Framework Core
https://github.com/bbonkr/sample.ef.batch

dotnet dotnet6 entityframeworkcore

Last synced: 25 days ago
JSON representation

Delete rows batch using Entity Framework Core

• Host: GitHub
• URL: https://github.com/bbonkr/sample.ef.batch
• Owner: bbonkr
• Created: 2022-01-07T14:36:20.000Z (almost 3 years ago)
• Default Branch: main
• Last Pushed: 2022-01-08T04:15:10.000Z (almost 3 years ago)
• Last Synced: 2024-04-09T21:56:09.156Z (7 months ago)
• Topics: dotnet, dotnet6, entityframeworkcore
• Language: C#
• Homepage:
• Size: 15.6 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
A repository I wrote out of curiosity about the benefits of using raw queries when using EntityFrameworkCore.

## About

### Environment

* Visual Studio 2022 (v17.0.4)

* Localdb (Please see appsettings.json)

* .NET 6

* EntityFrameworkCore v6.0.1

Target is 1,000 rows. 

> if you need dramatic result, Change target rows to up to 100,000. 🤔

### Generate migration code

If entity changes, you should generate migration code.

```powershell

PS> cd src/Sample.Data

PS> dotnet ef migrations add "Initialize database" --context AppDbContext --startup-project ../Sample.App --project ../Sample.Data.SqlServer --json

```

## Jobs

Each job is isolated to measure execution time.

### Add Some rows (#1-1, #1-2)

I think, insert batch feature is good to go.

```sql

-- Declaring variables 

DECLARE @inserted0 TABLE ([Id] bigint, [_Position] [int]);

MERGE [UserToken] USING (

VALUES (@p0, @p1, @p2, @p3, 0),

(@p4, @p5, @p6, @p7, 1),

(@p8, @p9, @p10, @p11, 2),

-- <>

(@p160, @p161, @p162, @p163, 40),

(@p164, @p165, @p166, @p167, 41)) AS i ([CreatedAt], [ExpiresAt], [Purpose], [Token], _Position) ON 1=0

WHEN NOT MATCHED THEN

INSERT ([CreatedAt], [ExpiresAt], [Purpose], [Token])

VALUES (i.[CreatedAt], i.[ExpiresAt], i.[Purpose], i.[Token])

OUTPUT INSERTED.[Id], i._Position

INTO @inserted0;

```

### Remove each rows #2

It runs DELETE statement to every row it needs.

This job takes 507 milliseconds.

```csharp

var deleteCandidate = Context.UserTokens.Where(x => x.ExpiresAt <= DateTime.UtcNow);

Context.UserTokens.RemoveRange(deleteCandidate);

await Context.SaveChangesAsync();

```

```sql

-- Declaring variables 

DELETE FROM [UserToken]

WHERE [Id] = @p0;

SELECT @@ROWCOUNT;

DELETE FROM [UserToken]

WHERE [Id] = @p1;

SELECT @@ROWCOUNT;

DELETE FROM [UserToken]

WHERE [Id] = @p2;

SELECT @@ROWCOUNT;

-- <>

```

### Remove rows #3

This job verify raw sql execution and transaction supporting.

It does not delete rows, because Its transaction will be rollback.

### Remove rows #4

Delete rows using raw SQL.

It may depend on DBMS.

I think, to use ANSI SQL for dependency lower when writes sql statement.

This job takes 17 milliseconds.

## References

* [EntityFrameworkCore: Bulk updates](https://docs.microsoft.com/en-us/ef/core/performance/efficient-updating#bulk-updates)