Greetings Spring Community!

I am pleased to announce the second milestone of both Spring Session Data Geode for Apache Geode and Spring Session Data GemFire for Pivotal GemFire.

The 2nd milestone release brings many welcome improvements, including:

• Upgrades to Spring Framework 5.0.0.RELEASE.

• Upgrades to Spring Data Kay-RELEASE.

• Upgrades to Spring Session 2.0.0.M4.

• Upgrades to Spring Boot 2.0.0.M4.

• Adds support for Pivotal GemFire and Apache Geode PDX Serialization.

• Introduces a new Serialization framework and adapter for Pivotal GemFire/Apache Geode’s Data Serialization and PDX frameworks.

Both artifacts can be downloaded from Spring’s libs-milestone Repository using Maven …

  <dependency>
    <groupId>org.springframework.session</groupId>
    <artifactId>spring-session-data-geode</artifactId>
    <version>2.0.0.M2</version>
  </dependency>

Or, using Gradle…​

compile 'org.springframework.session:spring-session-data-geode:2.0.0.M2

To use Spring Session with Pivotal GemFire, just switch the artifact from spring-session-data-geode to spring-session-data-gemfire. The version stays the same.

From Data Serialization to PDX

Many users have been requesting that Spring Session for Pivotal GemFire and Apache Geode support PDX Serialization. This makes sense given that PDX is highly "portable", as explained further in the Reference Guide.

One of the key benefits of PDX is that you do not need Spring Session for Pivotal GemFire or Apache Geode on the classpath of the servers in the cluster to use Spring Session in Spring Boot, GemFire/Geode cache client applications using the client/server topology. In fact, you do not even need to put your application domain object classes stored in an (HTTP) Session, and persisted to GemFire/Geode servers in the cluster, on the servers' classpath, either.

Spring Session Data GemFire/Geode now uses PDX by default. However, it is very easy to switch between Data Serialization and PDX just by switching the bean reference in the @EnableGemFireHttpSession annotation.

OOTB, Spring Session Data GemFire/Geode register 2 beans on the user’s behalf implementing each GemFire/Geode Serialization strategy.

To explicitly declare that you want your Spring Boot, Spring Session enabled application to use PDX Serialization, set the new sessionSerializerBeanName attribute in the @EnableGemFireHttpSession annotation to GemFireHttpSessionConfiguration.SESSION_PDX_SERIALIZER_BEAN_NAME…​

@SpringBootApplication
@EnableGemFireHttpSession(sessionSerializerBeanName =
    GemFireHttpSessionConfiguration.SESSION_PDX_SERIALIZER_BEAN_NAME
class Application {
  ...
}

To use Data Serialization, and leverage Deltas, set sessionSerializerBeanName to GemFireHttpSessionConfiguration.SESSION_DATA_SERIALIZER_BEAN_NAME…​

@SpringBootApplication
@EnableGemFireHttpSession(sessionSerializerBeanName =
    GemFireHttpSessionConfiguration.SESSION_DATA_SERIALIZER_BEAN_NAME
class Application {
  ...
}

More details on using PDX with Spring Session and Pivotal GemFire or Apache Geode can be found in the Reference Guide.

Serialization Framework

Spring Session Data GemFire/Geode introduces a new SessionSerializer interface allowing users to customize how a Session gets serialized to either Pivotal GemFire or Apache Geode.

interface SessionSerializer<T, IN, OUT> {

  void serializer(T session, OUT out);

  T deserialize(IN in);

  boolean canSerialize(Class<?> type);
}

Users can provide an custom implementation of this interface, register it as a bean in the Spring context, set the sessionSerializerBeanName attribute to the name of the custom SessionSerializer bean definition, and Spring Session Data GemFire/Geode will adapt and register this bean as a PdxSerializer in Pivotal GemFire or Apache Geode.

For instance…​

@SpringBootApplication
@EnableGemFireHttpSession(sessionSerializerBeanName = "customSerializer")
class Application {

  @Bean
  SessionSerializer<Session, PdxReader, PdxWriter> customSerializer() {
    // return an instance of a custom `SessionSerializer` implementation
  }
}

This is just the tip of the iceberg. More details can be found in the Reference Guide on the new Serialization framework.

Feedback

As always, any feedback on this release is highly appreciated. Users can reach out on StackOverflow, file GitHub Issues, or share your comments below. You can also reach me on Twitter, @john_blum.

SpringOne Platform 2017

Also make sure to checkout the lineup this year at SpringOne Platform 2017. This is going to be a year to remember with so much great content…​ Spring Framework 5 with early support of JDK 9, Reactive and Functional programming, integration with JUnit 5 and Kotlin. So much cool stuff to learn!