From 2ee231dee24a90b36041f898fb544af07f66ee48 Mon Sep 17 00:00:00 2001 From: Juergen Hoeller Date: Tue, 17 Nov 2020 14:43:17 +0100 Subject: [PATCH] Document that @Transactional does not propagate to new threads Closes gh-25439 --- .../transaction/annotation/Transactional.java | 19 +++-- .../RuleBasedTransactionAttribute.java | 2 +- src/docs/asciidoc/data-access.adoc | 73 +++++++++++-------- 3 files changed, 57 insertions(+), 37 deletions(-) diff --git a/spring-tx/src/main/java/org/springframework/transaction/annotation/Transactional.java b/spring-tx/src/main/java/org/springframework/transaction/annotation/Transactional.java index 4af6be7675..caad238aae 100644 --- a/spring-tx/src/main/java/org/springframework/transaction/annotation/Transactional.java +++ b/spring-tx/src/main/java/org/springframework/transaction/annotation/Transactional.java @@ -38,16 +38,25 @@ import org.springframework.transaction.TransactionDefinition; * {@link org.springframework.transaction.interceptor.RuleBasedTransactionAttribute} * class, and in fact {@link AnnotationTransactionAttributeSource} will directly * convert the data to the latter class, so that Spring's transaction support code - * does not have to know about annotations. If no rules are relevant to the exception, - * it will be treated like - * {@link org.springframework.transaction.interceptor.DefaultTransactionAttribute} - * (rolling back on {@link RuntimeException} and {@link Error} but not on checked - * exceptions). + * does not have to know about annotations. If no custom rollback rules apply, + * the transaction will roll back on {@link RuntimeException} and {@link Error} + * but not on checked exceptions. * *

For specific information about the semantics of this annotation's attributes, * consult the {@link org.springframework.transaction.TransactionDefinition} and * {@link org.springframework.transaction.interceptor.TransactionAttribute} javadocs. * + *

This annotation commonly works with thread-bound transactions managed by + * {@link org.springframework.transaction.PlatformTransactionManager}, exposing a + * transaction to all data access operations within the current execution thread. + * Note: This does NOT propagate to newly started threads within the method. + * + *

Alternatively, this annotation may demarcate a reactive transaction managed + * by {@link org.springframework.transaction.ReactiveTransactionManager} which + * uses the Reactor context instead of thread-local attributes. As a consequence, + * all participating data access operations need to execute within the same + * Reactor context in the same reactive pipeline. + * * @author Colin Sampaleanu * @author Juergen Hoeller * @author Sam Brannen diff --git a/spring-tx/src/main/java/org/springframework/transaction/interceptor/RuleBasedTransactionAttribute.java b/spring-tx/src/main/java/org/springframework/transaction/interceptor/RuleBasedTransactionAttribute.java index ecb231abec..0451f0518e 100644 --- a/spring-tx/src/main/java/org/springframework/transaction/interceptor/RuleBasedTransactionAttribute.java +++ b/spring-tx/src/main/java/org/springframework/transaction/interceptor/RuleBasedTransactionAttribute.java @@ -28,7 +28,7 @@ import org.springframework.lang.Nullable; /** * TransactionAttribute implementation that works out whether a given exception * should cause transaction rollback by applying a number of rollback rules, - * both positive and negative. If no rules are relevant to the exception, it + * both positive and negative. If no custom rollback rules apply, this attribute * behaves like DefaultTransactionAttribute (rolling back on runtime exceptions). * *

{@link TransactionAttributeEditor} creates objects of this class. diff --git a/src/docs/asciidoc/data-access.adoc b/src/docs/asciidoc/data-access.adoc index a2ca75459c..0fa295f6b6 100644 --- a/src/docs/asciidoc/data-access.adoc +++ b/src/docs/asciidoc/data-access.adoc @@ -566,7 +566,7 @@ abstractions mentioned earlier. [[transaction-declarative]] -=== Declarative transaction management +=== Declarative Transaction Management NOTE: Most Spring Framework users choose declarative transaction management. This option has the least impact on application code and, hence, is most consistent with the ideals of a @@ -637,7 +637,7 @@ around method invocations. NOTE: Spring AOP is covered in <>. -Spring Frameworks's `TransactionInterceptor` provides transaction management for +Spring Framework's `TransactionInterceptor` provides transaction management for imperative and reactive programming models. The interceptor detects the desired flavor of transaction management by inspecting the method return type. Methods returning a reactive type such as `Publisher` or Kotlin `Flow` (or a subtype of those) qualify for reactive @@ -648,6 +648,18 @@ Transaction management flavors impact which transaction manager is required. Imp transactions require a `PlatformTransactionManager`, while reactive transactions use `ReactiveTransactionManager` implementations. +[NOTE] +==== +`@Transactional` commonly works with thread-bound transactions managed by +`PlatformTransactionManager`, exposing a transaction to all data access operations within +the current execution thread. Note: This does _not_ propagate to newly started threads +within the method. + +A reactive transaction managed by `ReactiveTransactionManager` uses the Reactor context +instead of thread-local attributes. As a consequence, all participating data access +operations need to execute within the same Reactor context in the same reactive pipeline. +==== + The following image shows a conceptual view of calling a method on a transactional proxy: image::images/tx.png[] @@ -2844,30 +2856,29 @@ specific to each technology. Spring provides a convenient translation from technology-specific exceptions, such as `SQLException` to its own exception class hierarchy, which has `DataAccessException` as -the root exception. These exceptions wrap the original exception so that there is never any -risk that you might lose any information about what might have gone wrong. +the root exception. These exceptions wrap the original exception so that there is never +any risk that you might lose any information about what might have gone wrong. In addition to JDBC exceptions, Spring can also wrap JPA- and Hibernate-specific exceptions, -converting them to a set of focused runtime exceptions. -This lets you handle most non-recoverable persistence exceptions -in only the appropriate layers, without having annoying boilerplate -catch-and-throw blocks and exception declarations in your DAOs. (You can still trap -and handle exceptions anywhere you need to though.) As mentioned above, JDBC -exceptions (including database-specific dialects) are also converted to the same +converting them to a set of focused runtime exceptions. This lets you handle most +non-recoverable persistence exceptions in only the appropriate layers, without having +annoying boilerplate catch-and-throw blocks and exception declarations in your DAOs. +(You can still trap and handle exceptions anywhere you need to though.) As mentioned above, +JDBC exceptions (including database-specific dialects) are also converted to the same hierarchy, meaning that you can perform some operations with JDBC within a consistent programming model. -The preceding discussion holds true for the various template classes in Spring's support for various ORM -frameworks. If you use the interceptor-based classes, the application must care -about handling `HibernateExceptions` and `PersistenceExceptions` itself, preferably by -delegating to the `convertHibernateAccessException(..)` or -`convertJpaAccessException()` methods, respectively, of `SessionFactoryUtils`. These methods convert the exceptions +The preceding discussion holds true for the various template classes in Spring's support +for various ORM frameworks. If you use the interceptor-based classes, the application must +care about handling `HibernateExceptions` and `PersistenceExceptions` itself, preferably by +delegating to the `convertHibernateAccessException(..)` or `convertJpaAccessException(..)` +methods, respectively, of `SessionFactoryUtils`. These methods convert the exceptions to exceptions that are compatible with the exceptions in the `org.springframework.dao` -exception hierarchy. As `PersistenceExceptions` are unchecked, they can get -thrown, too (sacrificing generic DAO abstraction in terms of exceptions, though). +exception hierarchy. As `PersistenceExceptions` are unchecked, they can get thrown, too +(sacrificing generic DAO abstraction in terms of exceptions, though). -The following image shows the exception hierarchy that Spring provides. (Note that the -class hierarchy detailed in the image shows only a subset of the entire +The following image shows the exception hierarchy that Spring provides. +(Note that the class hierarchy detailed in the image shows only a subset of the entire `DataAccessException` hierarchy.) image::images/DataAccessException.png[] @@ -4232,14 +4243,14 @@ interface that wraps a single `Connection` that is not closed after each use. This is not multi-threading capable. If any client code calls `close` on the assumption of a pooled connection (as when using -persistence tools), you should set the `suppressClose` property to `true`. This setting returns a -close-suppressing proxy that wraps the physical connection. Note that you can no longer -cast this to a native Oracle `Connection` or a similar object. +persistence tools), you should set the `suppressClose` property to `true`. This setting +returns a close-suppressing proxy that wraps the physical connection. Note that you can +no longer cast this to a native Oracle `Connection` or a similar object. -`SingleConnectionDataSource` is primarily a test class. For example, it enables easy testing of code outside an -application server, in conjunction with a simple JNDI environment. In contrast to -`DriverManagerDataSource`, it reuses the same connection all the time, avoiding -excessive creation of physical connections. +`SingleConnectionDataSource` is primarily a test class. It typically enables easy testing +of code outside an application server, in conjunction with a simple JNDI environment. +In contrast to `DriverManagerDataSource`, it reuses the same connection all the time, +avoiding excessive creation of physical connections. @@ -8810,8 +8821,8 @@ can do so by using the following `applicationContext.xml`: ---- This application context uses XStream, but we could have used any of the other marshaller -instances described later in this chapter. Note that, by default, XStream does not require any further -configuration, so the bean definition is rather simple. Also note that the +instances described later in this chapter. Note that, by default, XStream does not require +any further configuration, so the bean definition is rather simple. Also note that the `XStreamMarshaller` implements both `Marshaller` and `Unmarshaller`, so we can refer to the `xstreamMarshaller` bean in both the `marshaller` and `unmarshaller` property of the application. @@ -8829,8 +8840,8 @@ This sample application produces the following `settings.xml` file: [[oxm-schema-based-config]] === XML Configuration Namespace -You can configure marshallers more concisely by using tags from the OXM namespace. To -make these tags available, you must first reference the appropriate schema in the +You can configure marshallers more concisely by using tags from the OXM namespace. +To make these tags available, you must first reference the appropriate schema in the preamble of the XML configuration file. The following example shows how to do so: [source,xml,indent=0] @@ -9073,7 +9084,7 @@ vulnerabilities do not get invoked. NOTE: Note that XStream is an XML serialization library, not a data binding library. Therefore, it has limited namespace support. As a result, it is rather unsuitable for usage -within Web services. +within Web Services.