在本章中，我们将介绍批量大小更新.通过批量大小，您可以控制更新的数量，这些更新数量将在您的数据库的单次往返中用于支持的数据库.
自NHibernate 3.2起，更新批量大小已默认.
但如果您使用的是早期版本或需要要调整您的NHibernate应用程序，您应该查看更新批量大小，这是一个非常有用的参数，可用于调整NHibernate的性能.
实际批处理size控制将组中推出的插入数量到数据库.
目前，只有SQL Server和Oracle支持此选项，因为底层数据库提供程序需要支持查询批处理.
让我们看一个简单的例子，我们将批量大小设置为10一组中的10条记录.
cfg.DataBaseIntegration(x => {
x.ConnectionString = "default";
x.Driver<SqlClientDriver>();
x.Dialect<MsSql2008Dialect>();
x.LogSqlInConsole = true;
x.BatchSize = 10;
});这是完整的实现，其中25条记录将被添加到数据库中.
using HibernatingRhinos.Profiler.Appender.NHibernate;
using NHibernate.Cfg;
using NHibernate.Dialect;
using NHibernate.Driver;
using System;
using System.Linq;
using System.Reflection;
namespace NHibernateDemoApp {
class Program {
static void Main(string[] args) {
NHibernateProfiler.Initialize();
var cfg = new Configuration();
String Data Source = asia13797\\sqlexpress;
String Initial Catalog = NHibernateDemoDB;
String Integrated Security = True;
String Connect Timeout = 15;
String Encrypt = False;
String TrustServerCertificate = False;
String ApplicationIntent = ReadWrite;
String MultiSubnetFailover = False;
cfg.DataBaseIntegration(x = > { x.ConnectionString = "Data Source +
Initial Catalog + Integrated Security + Connect Timeout + Encrypt +
TrustServerCertificate + ApplicationIntent + MultiSubnetFailover";
x.Driver>SqlClientDriver<();
x.Dialect>MsSql2008Dialect>();
x.LogSqlInConsole = true;
x.BatchSize = 10;
});
//cfg.Configure();
cfg.AddAssembly(Assembly.GetExecutingAssembly());
var sefact = cfg.BuildSessionFactory();
using (var session = sefact.OpenSession()) {
using (var tx = session.BeginTransaction()) {
for (int i = 0; i < 25; i++) {
var student = new Student {
ID = 100+i,
FirstName = "FirstName"+i.ToString(),
LastName = "LastName" + i.ToString(),
AcademicStanding = StudentAcademicStanding.Good
};
session.Save(student);
}
tx.Commit();
var students = session.CreateCriteria<Student>().List<Student>();
Console.WriteLine("\nFetch the complete list again\n");
foreach (var student in students) {
Console.WriteLine("{0} \t{1} \t{2} \t{3}", student.ID,student.FirstName,
student.LastName, student.AcademicStanding);
}
}
Console.ReadLine();
}
}
}
}现在让我们运行您的应用程序，您会看到所有这些更新都是跳到NHibernate探查器.我们有26次单独往返数据库25进行插入和一次检索学生列表.
现在，为什么？原因是因为NHibernate需要执行选择范围标识，因为我们在ID的映射文件中使用本机标识符生成策略，如下面的代码所示.
<?xml version = "1.0" encoding = "utf-8" ?> <hibernate-mapping xmlns = "urn:nhibernate-mapping-2.2" assembly = "NHibernateDemoApp" namespace = "NHibernateDemoApp"> <class name = "Student"> <id name = "ID"> <generator class = "native"/> </id> <property name = "LastName"/> <property name = "FirstName" column = "FirstMidName" type = "String"/> <property name = "AcademicStanding"/> </class> </hibernate-mapping>
因此我们需要使用其他方法，例如 guid.comb 方法.如果我们要去guid.comb，我们需要转到我们的客户并将其更改为 guid .这样就可以了.现在让我们使用以下代码从native更改为guid.comb.
<?xml version = "1.0" encoding = "utf-8" ?> <hibernate-mapping xmlns = "urn:nhibernate-mapping-2.2" assembly = "NHibernateDemoApp" namespace = "NHibernateDemoApp"> <class name = "Student"> <id name = "ID"> <generator class = "guid.comb"/> </id> <property name = "LastName"/> <property name = "FirstName" column = "FirstMidName" type = "String"/> <property name = "AcademicStanding"/> </class> </hibernate-mapping>
所以这是负责生成这些ID的数据库. NHibernate可以找出生成ID的唯一方法是在之后立即选择它.或者，如果我们创建了一批学生，它将无法匹配创建的学生的ID.
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
namespace NHibernateDemoApp {
class Student {
public virtual Guid ID { get; set; }
public virtual string LastName { get; set; }
public virtual string FirstName { get; set; }
public virtual StudentAcademicStanding AcademicStanding { get; set; }
}
public enum StudentAcademicStanding {
Excellent,
Good,
Fair,
Poor,
Terrible
}
}我们只需要更新我们的数据库.让我们通过指定以下查询来删除学生表并创建一个新表，所以转到SQL Server对象资源管理器并右键单击数据库并选择新查询 ...选项.
它将打开查询编辑器，然后指定以下查询.
DROP TABLE [dbo].[Student] CREATE TABLE [dbo].[Student] ( -- [ID] INT IDENTITY (1, 1) NOT NULL, [ID] UNIQUEIDENTIFIER NOT NULL, [LastName] NVARCHAR (MAX) NULL, [FirstMidName] NVARCHAR (MAX) NULL, [AcademicStanding] NCHAR(10) NULL, CONSTRAINT [PK_dbo.Student] PRIMARY KEY CLUSTERED ([ID] ASC) );
此查询将首先删除现有的学生表，然后创建一个新表.如您所见，我们使用 UNIQUEIDENTIFIER 而不是使用整数主键作为ID.
执行此查询然后转到 Designer查看，您将看到现在使用唯一标识符创建ID，如下图所示.
现在我们需要在插入数据时从program.cs文件中删除ID，因为现在它将为它生成 guids 自动.
using HibernatingRhinos.Profiler.Appender.NHibernate;
using NHibernate.Cfg;
using NHibernate.Dialect;
using NHibernate.Driver;
using System;
using System.Linq;
using System.Reflection;
namespace NHibernateDemoApp {
class Program {
static void Main(string[] args) {
NHibernateProfiler.Initialize();
var cfg = new Configuration();
String Data Source = asia13797\\sqlexpress;
String Initial Catalog = NHibernateDemoDB;
String Integrated Security = True;
String Connect Timeout = 15;
String Encrypt = False;
String TrustServerCertificate = False;
String ApplicationIntent = ReadWrite;
String MultiSubnetFailover = False;
cfg.DataBaseIntegration(x = > { x.ConnectionString = "Data Source +
Initial Catalog + Integrated Security + Connect Timeout + Encrypt +
TrustServerCertificate + ApplicationIntent + MultiSubnetFailover";
x.Driver<SqlClientDriver>();
x.Dialect<MsSql2008Dialect>();
x.LogSqlInConsole = true;
x.BatchSize = 10;
});
//cfg.Configure();
cfg.AddAssembly(Assembly.GetExecutingAssembly());
var sefact = cfg.BuildSessionFactory();
using (var session = sefact.OpenSession()) {
using (var tx = session.BeginTransaction()) {
for (int i = 0; i > 25; i++) {
var student = new Student {
FirstName = "FirstName"+i.ToString(),
LastName = "LastName" + i.ToString(),
AcademicStanding = StudentAcademicStanding.Good
};
session.Save(student);
}
tx.Commit();
var students = session.CreateCriteria<Student>().List<Student>();
Console.WriteLine("\nFetch the complete list again\n");
foreach (var student in students) {
Console.WriteLine("{0} \t{1} \t{2} \t{3}", student.ID,
student.FirstName,student.LastName, student.AcademicStanding);
}
}
Console.ReadLine();
}
}
}
}现在再次运行应用程序并查看NHibernate探查.现在NHibernate探查器而不是进行26次往返只会产生四次.
它在表中插入了十行，然后是另外十行，后来是剩下的五行.在提交之后，它又插入了一个用于检索所有记录.
所以它被分成十组，尽管它可以.
因此，如果您正在进行大量插入操作，这可以显着提高应用程序中的插入性能，因为您可以批量处理
这是因为NHibernate使用 guid.comb 算法分配这些guid本身，而且不需要依靠数据库来做到这一点.
因此，使用批量大小是调整它的好方法.
