Serialization

The flow of messages between (micro)services and storage of events requires preparation of the messages through serialization. Axon uses the XStreamSerializer by default, which uses XStream to serialize into and deserialize from XML. XStream is reasonably fast, and the result of serialization is human-readable. This makes it quite useful for logging and debugging purposes.

The XStreamSerializer allows further customization if that’s required. You can, for example, define aliases for specific packages, classes, or even fields. In addition to being an excellent way to shorten potentially long names, you can also use aliases when class definitions of the serialized objects change. For more information about aliases, visit the XStream website.

Additionally, Axon provides the JacksonSerializer. This Serializer implementation uses Jackson to serialize objects into and deserialize from JSON. It produces a more compact serialized form, while requiring those classes to stick to Jackson’s conventions (or configuration). The compact format makes it ideal for events, commands, and queries, as it minimizes the storage space and package size.

You may also implement your own serializer simply by creating a class that implements Serializer and setting it within Axon’s configuration for the desired infrastructure components.

Serializer Implementations

Serializers come in several flavors in Axon Framework and are used for various things. Currently, you can choose between the XStreamSerializer and JacksonSerializer to serialize messages (commands/queries/events), tokens, snapshots, deadlines and sagas in an Axon application.

As there are several objects to be serialized, it is typically desired to chose which serializer handles which object. To that end, the Configuration API allows you to define default, message and event serializers, which lead to the following object-serialization break down:

  1. The Event Serializer is in charge of (de)serializing event message payload and metadata. Events are typically stored in an event store for a long period of time. This is the main driver for choosing the event serializer implementation.

  2. The Message Serializer is in charge of (de)serializing the command and query message payload and metadata (used in a distributed application setup). Messages are shared between nodes and typically need to be interoperable and/or compact. Take this into account when choosing the message serializer implementation.

  3. The default Serializer is in charge of (de)serializing the remainder, being the messages (except the payload and metadata), tokens, snapshots, deadlines and sagas. These objects are generally not shared between different applications, and most of these classes aren’t expected to have some of the getters and setters that are, for example, typically required by Jackson based serializers. For example, a QueryMessage consists of a payload and ResponseType, which will respectively be (de)serialized using the message and the default serializer, the query request and response payload will be (de)serialized using the`message`serializer. A flexible, general-purpose serializer like XStream is ideal for this purpose.

By default, all three Serializer flavors are set to use the XStreamSerializer, which internally uses XStream to serialize objects to an XML format. XML is verbose, but XStream has the major benefit of being able to serialize virtually anything.

XStream and JDK 17

Although XStream can "serialize virtually anything," more recent versions of the JDK impede its flexibility. This predicament comes down to XStream’s reflective approach to finding out how to de-/serialize any object, which has become problematic with Java’s intent to secure its internals. Hence, if you’re using JDK 17, the chances are that objects (for example, your sagas) intended for serialization require additional configuration.

On some occasions configuring XStream’s security settings is sufficient. Other times you will have to introduce custom Converters to de-/serialize specific types. If you prefer not to deal with specific XStream settings, it might be better to use the JacksonSerializer throughout your Axon application.

XML’s verbosity is typically fine when storing tokens, sagas, or snapshots, but for messages (and specifically events) XML might cost too much due to its serialized size. Thus for optimization reasons you can configure different serializers for your messages. Another valid reason for customizing serializers is to achieve interoperability between different (Axon) applications, where the receiving end potentially enforces a specific serialized format.

There is an implicit ordering between the configurable serializer. If no event Serializer is configured, the event de-/serialization will be performed by the message Serializer. In turn, if no message Serializer is configured, the default Serializer will take that role.

See the following example on how to configure each serializer specifically, were we use the XStreamSerializer as the default and the JacksonSerializer for all our messages:

Axon Configuration API

public class SerializerConfiguration {

    public void serializerConfiguration(Configurer configurer) {
        // By default, we want the XStreamSerializer
        XStream xStream = new XStream();
        // Set the secure types on the xStream instance
        XStreamSerializer defaultSerializer = XStreamSerializer.builder()
                                                               .xStream(xStream)
                                                               .build();

        // But for all our messages we'd prefer the JacksonSerializer due to JSON's smaller format
        JacksonSerializer messageSerializer = JacksonSerializer.defaultSerializer();

        configurer.configureSerializer(configuration -> defaultSerializer)
                  .configureMessageSerializer(configuration -> messageSerializer)
                  .configureEventSerializer(configuration -> messageSerializer);
    }
}
Spring Boot - configuration class
@Configuration
public class SerializerConfiguration {

   // By default, we want the XStreamSerializer
   @Bean
   public Serializer defaultSerializer() {
      // Set the secure types on the xStream instance
      XStream xStream = new XStream();
      return XStreamSerializer.builder()
                              .xStream(xStream)
                              .build();
   }

   // But for all our messages we'd prefer the JacksonSerializer due to JSON's smaller format
   @Bean
   @Qualifier("messageSerializer")
   public Serializer messageSerializer() {
      return JacksonSerializer.defaultSerializer();
   }
}
Spring Boot - properties file
# Possible values for these keys are `default`, `xstream`, `java`, and `jackson`.
axon.serializer.general
axon.serializer.events
axon.serializer.messages
Spring Boot - YAML file
# Possible values for these keys are `default`, `xstream`, `java`, and `jackson`.
axon:
    serializer:
        general:
        events:
        messages:

Serializer tuning

Several things might be considered when the serialization process proofs to not be up to par with the expectations.

XStreamSerializer

XStream is extremely configurable and extensible. If you just use a plain XStreamSerializer, there are some quick wins ready to pick up. XStream allows you to configure aliases for package names and event class names. Aliases are typically much shorter (especially if you have long package names), making the serialized form of an event smaller. And since we’re talking XML, each character removed from XML is twice the profit (one for the start tag, and one for the end tag).

A more advanced topic in XStream is creating custom converters. The default reflection based converters are simple, but do not generate the most compact XML. Always look carefully at the generated XML and see if all the information there is really needed to reconstruct the original instance.

Avoid the use of upcasters when possible. XStream allows aliases to be used for fields, when they have changed name. Imagine revision 0 of an event, that used a field called "clientId". The business prefers the term "customer", so revision 1 was created with a field called "customerId". This can be configured completely in XStream, using field aliases. You need to configure two aliases, in the following order: alias "customerId" to "clientId" and then alias "customerId" to "customerId". This will tell XStream that if it encounters a field called "customerId", it will call the corresponding XML element "customerId" (the second alias overrides the first). If XStream encounters an XML element called "clientId", it is a known alias and will be resolved to field name "customerId". Check out the XStream documentation for more information.

For ultimate performance, you’re probably better off without reflection based mechanisms altogether. In that case, it is probably wisest to create a custom serialization mechanism. The DataInputStream and DataOutputStream allow you to easily write the contents of the events to an output stream. The ByteArrayOutputStream and ByteArrayInputStream allow writing to and reading from byte arrays.

Preventing duplicate serialization

Especially in distributed systems, event messages need to be serialized on multiple occasions. Axon’s components are aware of this and have support for SerializationAware messages. If a SerializationAware message is detected, its methods are used to serialize an object, instead of simply passing the payload to a serializer. This allows for performance optimizations.

When you serialize messages yourself, and want to benefit from the SerializationAware optimization, use the MessageSerializer class to serialize the payload and metadata of messages. All optimization logic is implemented in that class. See the JavaDoc of the MessageSerializer for more details.

Different serializer for events

When using event sourcing, serialized events can stick around for a long time. Therefore, consider the format to which they are serialized, carefully. Consider configuring a separate serializer for events, carefully optimized for the way they are stored. The JSON format generated by Jackson is generally more suitable for the long term than XStream’s XML format.

Lenient Deserialization

"Being lenient" from the Serializer’s perspective means the Serializer can ignore unknown properties. If it thus was handling a format to deserialize, it would not fail when it is incapable of finding a field / setter / constructor parameter for a given field in the serialized format.

Enabling lenient serialization can be especially helpful to accommodate different message versions. This situation would occur naturally when using an event store, as the format of the events would change overtime. This might also happen between commands and queries if several distinct versions of an application are run concurrently. A scenario when you would hit this is when going for a rolling upgrade pattern to deploying a new service.

To accommodate more closely with the desire to ignore unknown fields, both the XStreamSerializer and JacksonSerializer can be enabled as such. How to achieve this is shown in the following snippet:

XStreamSerializer
public class SerializerConfiguration {

    public Serializer buildSerializer() {
        return XStreamSerializer.builder()
                                .lenientDeserialization()
                                .build();
    }
}
JacksonSerializer
public class SerializerConfiguration {

    public Serializer buildSerializer() {
        return JacksonSerializer.builder()
                                .lenientDeserialization()
                                .build();
    }
}

Generic types

Sometimes the objects serialized by Axon will contain lists or collections of data. In other words, objects containing generics. For XStream, this poses no problem, as it will automatically add the type information to the serialized format. Jackson does not do this out of the box, however.

The recommended approach by Jackson is to use the @JsonTypeInfo to define type information. Doing so will attach the object type to your, for example, serialized list/collection. This approach ensures that the other end can reconstruct the concrete type for you.

Additionally, you can configure the ObjectMapper to add default typing information directly. However, note that this may not be a secure approach, as is described on Jackson’s Polymorphic Deserialization page.

If you prefer to enable default typing regardless, the JacksonSerializer’s builder provides a method to allow this for you. With JacksonSerializer.Builder#defaultTyping, you will automatically enable the addition of types to the serialized format for lists and collections. Consider the following sample on how to enable default typing for the JacksonSerializer:

public class SerializerConfiguration {
    // ...
    public Serializer buildSerializer() {
          return JacksonSerializer.builder()
                                  .defaultTyping()
                                  .build();
    }
}

ContentTypeConverters

An upcaster works on a given content type (for example, dom4j Document). To provide extra flexibility between upcasters, content types between chained upcasters may vary. Axon will try to convert between the content types automatically by using a ContentTypeConverter. It will search for the shortest path from type x to type y, perform the conversion and pass the converted value into the requested upcaster. For performance reasons, conversion will only be performed if the canUpcast method on the receiving upcaster yields true.

The ContentTypeConverter may depend on the type of serializer used. Attempting to convert a byte[] to a dom4j Document will not make any sense unless a Serializer was used that writes an event as XML. Axon Framework will only use the generic content type converters (such as the one converting a String to byte[] or a byte[] to InputStream) and the converters configured on the Serializer that will be used to deserialize the message. That means if you use a JSON based serializer, you would be able to convert to and from JSON-specific formats.

ContentTypeConverters and Upcasting

To achieve the best performance, ensure that all upcasters in the same chain (where one’s output is another chain’s input) work on the same content type.

If Axon does not provide the content type conversion that you need, you can always write one yourself by implementing the ContentTypeConverter interface.

The XStreamSerializer supports dom4j as well as XOM as XML document representations. The JacksonSerializer supports Jackson’s JsonNode and ObjectNode.