back to all blogsSee all blog posts

A Deep Dive into MicroProfile 4.0 with Open Liberty

image of author
Emily Jiang on Mar 19, 2021
Post available in languages:

The latest release of MicroProfile 4.0 aligns with Jakarta EE 8. There are many new and exciting features introduced in MicroProfile 4.0. In this post, we’ll take a look at all of the MicroProfile 4.0 updated features and then show you how to use them in Open Liberty.

What’s New In MicroProfile 4.0?

The full list of MicroProfile 4.0 specification includes the updates from Jakarta EE 8 and the MicroProfile specification updates.

Jakarta EE 8 Specification updates:

Open Liberty releases (19.0.0.6 and onwards) are Jakarta EE 8 compatible. Jakarta EE 8 features are functionally identical to their Java EE 8 counterparts with the only difference being the maven coordinates. The alignment with Jakarta EE 8 demonstrates MicroProfile welcomes the new releases from Jakarta EE and tries to keep up with the new releases from Jakarta EE.

MicroProfile Component Specification Updates:

In the next section, we will go through each MicroProfile Specification and take a further look at their new and exciting features.

Getting started

MicroProfile 4.0 implementation is now available in Open Liberty 21.0.0.3 and future releases.

To make use of MicroProfile 4.0 you must add the feature to both pom.xml and server.xml. Add the following Maven dependency to your pom.xml:

<dependency>
    <groupId>org.eclipse.microprofile</groupId>
    <artifactId>microprofile</artifactId>
    <version>4.0.1</version>
    <type>pom</type>
    <scope>provided</scope>
</dependency>

or, if you use Gradle:

dependencies {
    providedCompile 'org.eclipse.microprofile:microprofile:4.0.1'
}

Please note: you need to use the version 4.0.1 as the version 4.0 does not pull in any dependencies (see here for more details).

Also, make sure to configure your Liberty server with the microProfile-4.0 feature in the server.xml:

<server>
  <featureManager>
    <feature>microProfile-4.0</feature>
    <!-- ... -->
  </featureManager>
  <!-- ... -->
</server>

That’s it! Now that we’ve got our development and deployment environments set up, it’s time to dive into each individual MicroProfile Specification to learn more about these highly anticipated updates!

What’s New in MicroProfile Config 2.0

Quite a few nice capabilities were provided by MicroProfile Config 2.0, such as Config Profile, Config Properties, ConfigValue support, and more.

Open Liberty’s MicroProfile Config 2.0 feature is based on SmallRye’s implementation of the specification.

For more information about how to use the awesome new features, check out the MicroProfile Config 2.0 blog post.

What’s New in MicroProfile Fault Tolerance 3.0

MicroProfile Fault Tolerance allows you to easily apply strategies for mitigating failure to your code. It provides annotations that you can add to methods to use bulkhead, circuit breaker, retry, timeout and fallback strategies.

Fault Tolerance 3.0 introduces two main features.

MicroProfile Metrics tags

MicroProfile Fault Tolerance 3.0 overhauls the metrics that are automatically exported, to take advantage of tags. Tags make it easier to use those metrics, to understand when and where your application is failing.

Information which was previously included in the metric name is now included as metric tags. This makes it much easier to query for data from multiple methods and pick out those which may be causing an issue.

Example:

Old metric: application:ft.<name>.timeout.callsTimedOut.total

New metric: base:ft.timeout.calls.total{method="<name>", timedOut="true"}

Clarify the scope of Bulkheads and CircuitBreakers

The specification now requires that circuit breakers and bulkheads are singletons, identified by the bean class and the guarded method. For example, if a @RequestScoped bean has a @CircuitBreaker method, all invocations of that method will share the same circuit breaker state, even though each request has a different instance of the bean.

Get an introduction to MicroProfile Fault Tolerance with the Open Liberty guides Failing fast and recovering from errors and Preventing repeated failed calls to microservices.

Try it now

Enable Fault Tolerance 3.0 and CDI in the server.xml, along with any other features you’re using.

<featureManager>
  <feature>mpFaultTolerance-3.0</feature>
  <feature>cdi-2.0</feature>
  <feature>jaxrs-2.1</feature>
</featureManager>

For more information:

What’s New in MicroProfile Metrics 3.0

MicroProfile Metrics 3.0 introduces new metric values for the existing SimpleTimer and Timer metrics. Additionally a new REST metric is introduced for better monitoring and handling of unmapped exceptions. Manual configuration for re-usability has been removed. A notable change to the MicroProfile Metrics programming model regarding CDI Producers has been made. Lastly a medley of API improvements and refactoring have been added in this release.

SimpleTimer

The SimpleTimer metric now tracks and reports the highest and lowest recorded time duration of the previous complete minute. See SimpleTimer Javadoc for more information.

Timer

The Timer metric now tracks and reports the total elapsed time duration. See Timer Javadoc for more information.

REST metric

A new REST.request.unmappedException.total metric that is backed by a counter metric has been introduced. Similar to the REST.request metric, there will be one unique metric for each REST endpoint identified by a class and method label. The new REST metric will count the amount of times the request ends in an unmapped exception. The REST.request metric corresponding to this REST endpoint will not record any values if an unmapped exception has occured.

CDI Producer

The @Metrics annotation will no longer support the method target (i.e it can not be annotated on a method). Additionally, it will not support usage with CDI Producers.

Try it now

Enable Metrics 3.0 in the server.xml, along with any other features you’re using.

<featureManager>
    <feature>mpMetrics-3.0</feature>
</featureManager>

More information:

New changes in MicroProfile Health 3.0

MicroProfile Health 3.0 enables you to provide your own health check procedures to be invoked by Open Liberty, to verify the health of your microservices.

MicroProfile Health allows services to report their health, and publish overall health status to a defined endpoint. A service reports "UP" if it is available and reports "DOWN" if it is unavailable. In addition to reporting the state of individual services, MicroProfile Health also agregates the overall status as "UP" if all the services are "UP". If one of the services is "DOWN" then the overall status is also "DOWN". A service orchestrator can then use the health statuses to make decisions.

MicroProfile Health checks its own health by performing necessary self-checks and then reports its overall status by implementing the API provided by MicroProfile Health. A self-check can be a check on anything that the service needs, such as a dependency, a successful connection to an endpoint, a system property, a database connection, or the availability of required resources. MicroProfile offers checks for both liveness and readiness.

In the mpHealth-3.0 feature for Open Liberty:

  • The overall default Readiness status was changed to "DOWN", with an empty response until all the deployed application(s) have been started. A new MicroProfile Config property (mp.health.default.readiness.empty.response=UP) is introduced to change the overall default Readiness check status to "UP", during application start up, that do not have any user-defined health checks.

  • The HealthCheckResponseBuilder.state(Boolean UP) method was also renamed to HealthCheckResponseBuilder.status(Boolean UP) for HealthCheckResponse deserialization compatibility, where the JSON health check response string can now be deserialized into an HealthCheckResponse object.

  • The deprecated @Health qualifier was removed. Instead, the @Liveness or @Readiness qualifiers should be used as appropriate in your HealthCheck implementations.

Applications are expected to provide health check procedures by implementing the HealthCheck interface with the @Liveness or @Readiness annotations. These are used by Open Liberty to verify the Liveness or Readiness of the application, respectively. Add the logic of your health check in the call() method, and return an instance of HealthCheckResponse by using the simple up and down API methods:

//**Liveness Check**
@Liveness
@ApplicationScoped
public class AppLiveCheck implements HealthCheck {
...
    @Override
     public HealthCheckResponse call() {
       ...
       HealthCheckResponse.up("my-liveness-check");
       ...
     }
}

//**Readiness Check**
@Readiness
@ApplicationScoped
public class AppReadyCheck implements HealthCheck {
...
    @Override
     public HealthCheckResponse call() {
       ...
       HealthCheckResponse.named("my-app-readiness").status(isMyAppReady()).build();
       ...
     }
}
...

To view the status of each health check, access either http://<hostname>:<port>/health/live or http://<hostname>:<port>/health/ready endpoints.

More information:

New updates in MicroProfile Open Tracing 2.0

MicroProfile OpenTracing 2.0 can be used to profile and monitor applications built using microservice architecture.

MicroProfile OpenTracing 2.0 has upgraded the OpenTracing API to version 0.33.0. This allows the use of tracing backends and their libraries that are built on OpenTracing API 0.33.0.

Try it now

Include the following in the server.xml:

    <feature>mpOpenTracing-2.0</feature>

Next, configure a tracing backend such as Jaeger or Zipkin.
For Jaeger, add the following maven dependencies in the application’s pom.xml.

<dependency>
    <groupId>io.jaegertracing</groupId>
    <artifactId>jaeger-client</artifactId>
    <version>1.2.0</version>
</dependency>
<dependency>
    <groupId>org.slf4j</groupId>
    <artifactId>slf4j-api</artifactId>
    <version>1.7.30</version>
</dependency>
<dependency>
    <groupId>org.slf4j</groupId>
    <artifactId>slf4j-jdk14</artifactId>
    <version>1.7.30</version>
</dependency>

You can find out more about about configuring Jaeger settings using environment variables by looking at jaeger-client-java readme.

We also have an Open Liberty guide on enabling distributed tracing using Jaeger which you can access here.

For the JAEGER_PASSWORD environment variable, the password can be encoded using the securityUtility command.

Depending on Jaeger’s sampling settings JAEGER_SAMPLER_TYPE and JAEGER_SAMPLER_PARAM, Jaeger may not report every span created by the applications.

For Zipkin, take a look at the sample project to see how to implement a tracer for Liberty.

Define your application in the server.xml:

<webApplication location="yourapp.war" contextRoot="/yourapp">
    <!-- enable visibility to third party APIs -->
    <classloader apiTypeVisibility="+third-party" />
</webApplication>

We also have an Open Liberty guide on enabling distributed tracing with Zipkin too which you can access here.

Once you have hit some JAX-RS endpoints of your application, you should be able to find spans in the user interface of your tracing backend.

More information:

What’s New in MicroProfile Open API 2.0

MicroProfile OpenAPI 2.0 builds on top of the OpenAPI v3 specification. The OpenAPI v3 specification defines a standard, language-agnostic, interface for describing REST APIs which allows documentation to be generated from the APIs themselves. The MicroProfile OpenAPI specification provides a unified Java API for the OpenAPI v3 specification which allows Java developers to generate OpenAPI v3 documents from their JAX-RS applications. MicroProfile OpenAPI 2.0 introduces some new annotations that simplify the process of generating OpenAPI documentation for your REST APIs. It also introduces a new MicroProfile Config property prefix that can be used to define the schema for Java classes that you do not have the source code for.

@SchemaProperty

The @SchemaProperty annotation has been added to allow you to include properties in the definition of a schema when using the @Schema annotation. Previously, the @Schema annotation did not allow you to define any properties. Properties were introspected from the implementation class. With the introduction of the @SchemaProperty annotation, a list of properties can now be defined using the properties attribute on the @Schema annotation. For example;

@Schema(
    name = "User",
    description = "Telephone number to contact the user”,
    implementation = User.class,
    properties = {
        @SchemaProperty(
            name = "phone",
            description = "Telephone number to contact the user
         )
    }
)

@RequestBodySchema

The @RequestBodySchema annotation has been added to provide a shorthand mechanism to specify the schema for a request body. Previously, defining the schema for a request body required the use of three different annotations, as follows:

@RequestBody(
    content = {
        @Content(
            schema = @Schema(
                implementation = MyRequestObject.class
            )
        )
    }
)

With the introduction of the @RequestBodySchema annotation, this can be simplified to the following single annotation:

@RequestBodySchema(MyRequestObject.class)

@APIResponseSchema

Similarly, the @APIResponseSchema annotation has been added to provide a shorthand mechanism to specify the schema for a response body. Previously, defining the schema for a response body required the use of three different annotations:

@APIResponse(
    content = {
            @Content(
                schema = @Schema(
                    implementation = MyResponseObject.class
                )
            )
        }
    )

With the introduction of the @RequestBodySchema annotation, this can be simplified to the following single annotation:

@APIResponseSchema(MyResponseObject.class)

The "mp.openapi.schema." property prefix

The mp.openapi.schema.* MicroProfile Config property prefix has been added to allow you to define the schema for Java classes using configuration rather than code. The use of this property is functionally equivalent to the use of the @Schema annotation on a Java class, but may be used in cases where you do not have access to the source code of a class. The remainder of the property key must be the fully-qualified class name and the value must be a valid OpenAPI schema object, specified in the JSON format. For example, in the case where an application needs to represent dates in epoch milliseconds, the following configuration could be used (line escapes and indentation added for readability):

mp.openapi.schema.java.util.Date = { \
   "name": "EpochMillis" \
   "type": "number", \
   "format": "int64", \
   "description": "Milliseconds since January 1, 1970, 00:00:00 GMT" \
 }

Enable OpenAPI 2.0 in the server.xml, along with any other features you’re using.

<featureManager>
    <feature>mpOpenAPI-2.0</feature>
</featureManager>

For more information:

Awesome new features in MicroProfile Rest Client 2.0

MicroProfile REST Client is a type-safe client API enabling rapid development of applications capable of consuming RESTful services. Version 2.0 is the latest update and adds support for HTTP proxy servers, automatically following HTTP redirects, Server Sent Events, and additional configuration options for JSON-B providers and multiple query parameters.

Please check out the following dedicated blog post solely on MicroProfile Rest Client 2.0, which demonstrates the new features and how to use them.

Enhancements in MicroProfile JWT Auth 1.2

MicroProfile JWT 1.2 simplifies the configuration for managing the validation of the JWT by introducing new MicroProfile Config properties. Enhanced signature algorithm support is added in this Open Liberty implementation.

MicroProfile JWT 1.2 adds a number of MicroProfile Config properties to control different aspects of the JWT validation. The specification also adds support for the ES256 signature algorithm, while this Open Liberty beta supports using the RS384, RS512, HS384, HS512, ES256, ES384, and the ES512 signature algorithms.

MicroProfile JWT 1.2 also supports the JWT specified in a cookie.

For more information about the exciting new features, check out the MicroProfile JWT 1.2 blog post.

Summary

As you can see, MicroProfile 4.0 has many new updates. It is time to try out MicroProfile 4.0 using Open Liberty 21.0.0.3 !

As always, let us know if you have any questions with this new feature. Thanks for checking it out!