OSGi Declarative Services – Configuration Binding

As explained in OSGi Components – Simply Simple – Part II when using Declarative Services (DS) you get support for configurations stored in a Configuration Admin service for free. At runtime especially when checking the status of the system – or looking for misconfigurations, there is always the question how to find out if a component is using a certain configuration.

The ServiceComponentRuntime Service

Since R6 of the OSGi specification, the Declarative Service Specification defines the ServiceComponentRuntime service. This service can be used at runtime to get the status of the components and services managed by DS. The top level DTO available from that service is the ComponentDescriptionDTO. This is basically a representation of the XML provided by a bundle and read by DS. So the first step is to find the ComponentDescriptionDTO for the component you’re interested in.

Once you have found such a description you can again use the ServiceComponentRuntime service to get the ComponentConfigurationDTO(s) for the description. This DTO has some useful information like whether the component is satisfied and if not it provides more info about the reason. For example, a reference might be unsatisfied. You can now iterate over the references and find the one(s) which is not satisfied.

Now for the configurations, the first check is to see whether the PID of the configuration you’re interested in is listed in the configurationPid field of the ComponentDescriptionDTO. However this does not provide you the information whether such a configuration exists and is used by your component.

Therefore you have to check the properties from the ComponentConfigurationDTO and check whether the properties contain a service.pid property and whether the value for this property contains the configuration PID you’re interested in. Note that the type of this property can be a String, an array of Strings or a collection of Strings.

This requires a little bit of coding, iterating over the various DTOs and searching for the info you’re interested in.

The Declarative Services WebConsole Plugin

Another way of figuring this out without writing code is to use the famous Apache Felix Web Console. The web console allows you to manage and monitor a running OSGi application through a web interface. The Apache Felix project has a plugin for the web console. The Apache Maven coordinates for the current release are org.apache.felix/org.apache.felix.webconsole.plugins.ds/2.0.6.

Once this plugin is installed it gives you an UI for the previously mentioned ServiceComponentRuntime service and you can browse through the DTOs and find the ones you’re interested in.

Bound OSGi Configurations

An OSGi configuration supports the concept of configuration binding. By checking the bundle location field of a configuration object, it is possible to figure our whether a managed service is using this configuration.

However, don’t use this information to figure out whether a configuration is used by a DS component. Configuration binding is considered legacy and more important a DS implementation is not required to bind a configuration to the bundle of a bundle containing the component. While older implementations of the Apache Felix SCR did the configuration binding, newer versions don’t.

Therefore don’t even look there if you want to figure out whether your configuration is used by a DS component.

New OSGi Configuration Features in R7

The specification groups of the OSGi Alliance are currently in the process of finalizing the work for the upcoming OSGi R7 release. The major part of the specification work is done. This post explains some new functionality around OSGi configurations. Let’s start with a very brief introduction.

Configuration Admin

One of the most used but on the other hand barely noticed services from the OSGi compendium specification is the Configuration Admin. This service allows to create, update and delete configurations. It is up to the implementation where these configurations are stored. A configuration has a unique persistent identifier (PID) and a dictionary of properties.

Usually configurations are tied to an implementation of an OSGi service but configurations can be used for any purpose like database connections, current temperature or the set of available nodes in a cloud setup. While the Configuration Admin service has an API to find configurations or create them, it supports a more inversion of control like behavior by supporting a callback mechanism. The callback (known as the ManagedService interface) gets invoked for existing/created configurations of a certain PID and also if that configuration is deleted.

While this callback exists, its very often not used directly. The most common and easiest way to develop OSGi components and services is to use Declarative Services. Declarative Services (DS) provides build-in support for configurations. Simply by implementing an activation method the component can get it’s configuration. The implementor of that component does not have to worry whether such a configuration exists, gets deleted or is modified. DS takes care of all of this and invokes the right actions and the component.

In addition to single configurations, Configuration Admin provides support for factory configurations: multiple configurations of the same type, like for example different logger configurations for different categories or different database connection configurations. These factory configurations have a factory PID which is the same for all configurations of that type and a configuration PID to distinguish those configurations.

This should explain big picture, if you’re interested in more details, I refer you to the OSGi specifications. Especially the chapters about Configuration Admin and Declarative Services provide very valuable information. Let’s start talking about the new things.

Configurator Specification

R7 adds chapter 150, Configurator Specification. It basically consists of two parts. The first part describes a common textual representation of a set of configurations. Before this specification each and every tool was using its own format for provisioning configurations. For example, the famous Apache Felix File Install uses a properties like format. Other tools use slightly different formats. One problem is that you can’t simply switch from one tool to another and the other major problem is that some of the formats do not allow to specify the real type of a property value. For example, the value for the service ranking property must be of type Integer or your special implementation is expecting (for whatever reason) a value to be of type Byte. However some tools are simply always using a Long to represent numbers.

Therefore a common definition eliminates these problems and allows interchangeability of configurations between various tools. The format is JSON based and uses the PIDs of a configuration as the keys. The value is the configuration object with the properties:

  "my.component.pid": { 
    "port:Integer" : 300, 
    "array:int[]" : [2, 3, 4], 
    "collection:Collection<Integer>" : [2,3,4], 
    "complex": { 
      "a" : 1, 
      "b" : "two"

As you can see in the example, it is possible to specify the runtime type of a configuration property by separating the property name from the type using a colon. For example, the “port” property value is of type Integer, the “array” property value is an array of ints and the “collection” property value is of type Collection<Integer>. You can specify all allowed types for a configuration and the configurator implementation uses converting rules as defined by the Converter Specification – another of the new specifications of R7.

In addition, a configuration property can hold structured JSON as a string value. In the example above “complex” contains at runtime a string value of the specified JSON.

Factory configurations can be specified by using the following syntax: FACTORY_PID~NAME. With the R7 Configuration Admin it is possible to use a meaningful name to address factory components, the name (see below). The tilde separates the factory PID from the name:

  "my.factory.component~foo" : {
  "my.factory.component~bar" : {

OSGi Bundle Configurations

The second part of the configurator specification describes a new extender based mechanism that picks up configurations from within a bundle and applies them. A bundle can contain one or more JSON files with configurations and once the bundle is started the configurations will be put into Configuration Admin by the Configurator. The configurator manages the state handling and ordering in a deterministic way. For example, if two bundles contain a configuration for the same PID, a ranking mechanism is used to specify which configuration is put into Configuration Admin, regardless of their installation or start order.

With this new feature, provisioning of bundles becomes part of the OSGi specifications. The specification of the Configurator has driven the update of the Configuration Admin specification. So let’s talk about the most important new features.

Improved Factory Configuration Handling

With the upcoming R7 specification handling factory configurations has been greatly improved. Before, when you create a factory configuration, the PID part is randomly generated which makes identifying a particular factory configuration later on much harder. In addition, as the PID is auto generated, it has no meaning. With R7 it is now possible to specify the PID of a factory configuration, removing those problems in the future.

New methods on the Configuration Admin allow to create and retrieve factory configurations based on the factory PID and a name. These methods behave the same as the already existing methods for plain configurations. The PID for those factory components is generated by appending the name to the factory PID separated by a tilde. The Configurator uses this syntax to specify factory configurations.

Improved Configuration Plugin Handling

When a configuration is delivered to a managed service, the configuration is passed through registered configuration plugin services. Such a service can manipulate the configuration. One common use case is to handle place holders in the configuration properties and replace them with real values when delivered. For example  a property of a database connection configuration could just contain the value “${database.url}” which is replaced with the actual url when this configuration is passed to the component processing the configuration. Or if you have sensitive configuration data, you can store it encrypted in the configuration and just decrypt it in a configuration plugin before it is passed to the managed service.

While this mechanism sounds useful, it is only useful if you register a managed service. However, when you’re using Declarative Services for your components, the plugins are not called at all. Starting with R7 this gap is closed and the DS implementation uses a new functionality of the Configuration Admin service and calls the plugins before it is passing the configuration to components. This ensures, regardless how you get your configuration, plugins will be called making those use cases mentioned above possible.


The standard format for OSGi configurations is a great step forward for tooling and the Configurator implementation allows to deploy configurations through bundles in a standardized and well specified way. The update of the Configuration Admin resolves some long outstanding issues and allows for new use cases around configuration management. Of course, the final R7 specification will continue all the details.

Http Whiteboard – Simply Simple – Part III

This is the next part in the series about the OSGi http whiteboard. In my last post I explained how to create an own application servlet context and how to register web components with this context.

Bundle Resources

Doing as outlined in that post, resource handling (the get resource methods on the servlet context) might not work as you expect as you can’t get resources from the bundle where you web component is defined in. If you want this, you need first of all need to register your ServletContextHelper with the service scope Bundle – this ensures that every using bundle gets its own instance of the ServletContextHelper. And with that a servlet from bundle A gets only resources from bundle A, while the servlet from bundle B gets only the resources from that bundle – although they are registered within the same servlet context.

Now using DS this looks a little bit complicated, here is the code:

import java.net.URL;
import java.util.Set;

import org.osgi.service.component.ComponentContext;
import org.osgi.service.component.annotations.Activate;
import org.osgi.service.component.annotations.Component;
import org.osgi.service.component.annotations.ServiceScope;
import org.osgi.service.http.context.ServletContextHelper;
import org.osgi.service.http.whiteboard.HttpWhiteboardConstants;

        service = ServletContextHelper.class,
        scope = ServiceScope.BUNDLE,
        property = {
                HttpWhiteboardConstants.HTTP_WHITEBOARD_CONTEXT_NAME + "=" + AppServletContext.NAME,
                HttpWhiteboardConstants.HTTP_WHITEBOARD_CONTEXT_PATH + "=/guessinggame"
public class AppServletContext extends ServletContextHelper {

    public static final String NAME = "game";

    private ServletContextHelper delegatee;

    private void activate(final ComponentContext ctx) {
        delegatee = new ServletContextHelper(ctx.getUsingBundle()) {};

    public URL getResource(String name) {
        return delegatee.getResource(name);

    public String getMimeType(String name) {
        return delegatee.getMimeType(name);

    public Set<String> getResourcePaths(String path) {
        return delegatee.getResourcePaths(path);

    public String getRealPath(String path) {
        return delegatee.getRealPath(path);

As you can see in the above example, the ServletContextHelper is registered with the Bundle scope. In order to use the provided functionality of the abstract class ServletContextHelper in the activate method a delegatee is created with the bundle using the ServletContextHelper. The different methods then delegate to the delegatee.

This example looks a little bit complicated, but it is due to the fact that ServletContextHelper only provides to set the bundle through the constructor. And with Declarative Services this is currently a little bit more work to do. But for the upcoming OSGi R7 release, constructor injection is planned, then this should look much cleaner.

Http Whiteboard – Simply Simple – Part II

Developing web components with OSGi can be very simple if you’re using the OSGi Http Whiteboard Specification (see OSGi Compendium Chapter 140). For example the Apache Felix Http Implementation supports this standard.

Building your own Application

With traditional web application development, you usually package your application into its own web application archive (war file) and deploy this into an application server (servlet engine). Each application has its own servlet context and is mounted at some path. You can either mount an application at the root or any other path, e.g. at /my-app.

Using OSGI and the OSGi Http Whiteboard Specification, when you develop a web application you usually but the web components into a bundle and deploy this bundle. Depending on your needs you might also split up the web components into different bundles. But how do you create the equivalent of a servlet context known from web archive deployments?

You can create an own servlet context for your application through the Http Whiteboard. But there is no special deployment format, you simply deploy your app through bundles. And in contrast to traditional webapp deployment, this is leveraging all the well-known OSGi features. And of course source a context is represented through a service.

Let’s start with creating a simple application. The first thing you do is you create your own servlet context. This is done by registering a ServletContextHelper service (again for the examples I’m using Declarative Services and the corresponding annotations):

        service = ServletContextHelper.class,
        property = {
                HttpWhiteboardConstants.HTTP_WHITEBOARD_CONTEXT_NAME + "=" + AppServletContext.NAME,
                HttpWhiteboardConstants.HTTP_WHITEBOARD_CONTEXT_PATH + "=/guessinggame"
public class AppServletContext extends ServletContextHelper {}

As you can see from the example, ServletContextHelper is not an interface but an abstract class and your implementation gets all the default behaviour by simply extending this class. An abstract class has been chosen over the interface in order to be able to add methods to the service without breaking existing implementations. More important are the two required properties: a unique name and a context paths. The path is equivalent to the context path of a normal web application. The name is used to reference this servlet context. If you want to associate a servlet, resource, servlet filter, or listener to this context, you can do this by adding an additional property to select this context:

@Component(service = Servlet.class,
           property = { HttpWhiteboardConstants.HTTP_WHITEBOARD_SERVLET_PATTERN + "=/game",
                        HttpWhiteboardConstants.HTTP_WHITEBOARD_CONTEXT_SELECT + "=(" + HttpWhiteboardConstants.HTTP_WHITEBOARD_CONTEXT_NAME +"=" + AppServletContext.NAME + ")"})
public class GameServlet extends HttpServlet {...}

The context select property defines a filter expression to select the servlet context. While you can use any filter expression there, it is pretty common to simply filter against the name of the servlet context. If the context select property is missing, your web component is added to the default context.

The path of the servlet context is prepended before the path of the web component, so in the example above the servlet is mounted at “guessinggame/game”.

Building OSGi Bundles with Apache Maven

It seems that still today some developers think that creating an OSGi bundle is complicated – in contrast to simply creating a jar. Well, an OSGi bundle is a jar. The factor which distinguishes the two is that an OSGi bundle has additional manifest entries. Adding those using the right tooling is really simple.

The more difficult part is creating good modules: clearly separating public API from the implementation and correctly version the API based on the changes you made. But this is a general problem not related to OSGi at all and applies to any Java coding. There are some simple guidelines:

  • Leverage packages – put the API in a different package than your implementation. It’s also good still to use a package name for your implementation which makes clear that this is the implementation, e.g. by using impl as one part of the package name.
  • Start private and only make public if necessary. Once something is public, you never can make it private without breaking your clients. Starting private makes things easier.
  • Use semantic versioning. Whenever you change your public API, make sure to correctly increase the package version depending on the type of change you made.

Again, these are general guidelines and have nothing to do with OSGi. But OSGi uses headers in the manifest of your jar to know which packages from inside the jar are public API and the version of those packages. This is the export package header. Packages not listed there are private and not accessible to other modules. The other important header is the import package header, specifying which other packages this module is using. This usually comes with a version range, like a bundle is using package foo.bar version 1.1 or higher up to but not including 2.0.

Therefore, the only required thing to make a bundle out of your jar is more or less calculating these export and import packages headers and adding them to the manifest. But don’t fear, the tooling does this automatically for you. Let’s build a bundle/jar.

Using Apache Maven to Create a OSGi Bundle

There is always the year old debate on which build tool is the best one, I will not get into this discussion. I simply use what I know best and what I’m forced to use anyway: Apache Maven.

By default, Maven is not adding the additional manifest information to the jar, but there are different plugins for Maven available. For this tutorial, I’ll use the newer bnd Maven Plugin from the bnd project.

Add the following two plugin configurations to your pom. It’s usually a good idea to put this into your parent pom:

             -exportcontents: ${packages;VERSIONED}

The first adds the bnd maven plugin which creates the manifest data and the second instructs the default jar plugin to use the manifest created by the bnd plugin. And that’s it. If you now build your jar project using Maven you’ll have additional manifest entries in there.

Standard OSGi Bundle Headers

The first entries are standard headers which make your jar a bundle and contain some metadata:

  • Bundle-ManifestVersion: This header is required and marks the jar as a bundle, the value is always 2 (newer OSGi specifications might add more features in which case the number would be increased).
  • Bundle-Name: A human readable name for your bundle. This defaults to your project name from the pom.
  • Bundle-SymbolicName: In combination with the bundle version (see below) the symbolic name provides a key to uniquely identify a bundle. This name should be based on the reverse domain name convention and defaults to the artifact id of your project. I’ll talk a little bit more about this soon.
  • Bundle-Version: The version of your bundle, this defaults to the version of your project.

There is no magic in the above entries and having them in your jar does not hurt at all. As mentioned above, the symbolic name should be based on the reverse domain name convention. Usually in the Maven world you follow this rule for your group id, like org.mycompany.myproject and then use a simple name for the artifact id like scheduler-api. However a good practice is to use a full qualified name for the artifact id (which usually means you prefix the artifact id with the group id), e.g. org.mycompany.myproject.scheduler-api. Why is this a good idea? It has nothing to do with OSGi and simply makes sure that the final name of the artifact is fully qualified. If you don’t add the group id to the artifact id, the jar filename would be scheduler-api-1.2.3.jar – which is not very descriptive and gives you no clue what this thing is (and it might clash with other projects using the same artifact id).

The bundle version as any version in software development should increase with continued development/releases. For OSGi this information together with the symbolic name provides a unique key for a bundle and allows OSGi to verify if this exact bundle (name and version) is already installed in the OSGi framework. On a technical level, the bundle version information is more a marketing version.

Public API

More important than versioning your bundle (which still is important), it’s more important to correctly version your public API which might be used by others. For this add the following dependency to your pom:


The above dependency adds some annotations to your project – using the scope “provided” is a good practice as this avoids dragging in transitive dependencies.

Your public API should be in a separate Java package than your implementation – again this is a general good style and not tied to OSGi at all. If you want to export a package to be used by others, add a package-info.java file within your package:

package org.osoco.software.samples.guessinggame;

If it’s the first version of your package, simply use 1.0 as the version – from now on once you have released this API as version 1.0, follow semantic versioning. Build your project again and you will see a header similar to this:

Export-Package                org.osoco.software.samples.guessinggame;version="1.0"

As you can see creating the export is really easy, just use the above annotation and done. And the imports are even easier. The plugin analyses your classes and calculates the imports for you. For example for my sample project it looks like this:

Import-Package       javax.servlet;version="[3.1,4)",

And that’s it, your jar is now a bundle and can be used in any OSGi installation as a first class citizen. It’s really simple – and in 96% of your cases the automatic calculation by the tooling is sufficient. There are only rare and special cases where you want to have them differently. In that case, you can configure the plugin accordingly.

Migrating from the Apache Felix SCR Annotations to the OSGi Declarative Services Annotations

The Java annotations of the Apache Felix SCR Plugin were one of the first options to use annotations to create the descriptors for OSGi Declarative Services components and OSGi metatype descriptions for the configuration of such components. With the OSGi R6 release from 2015, the annotations of the OSGi specifications provide the same functionality and go even beyond.

Migrating from the Apache Felix SCR Annotations

Whenever there are competing solutions for the same problem, the question of which one to use arises. In this case, the answer is clear: use the official annotations from the OSGi specifications – for one, they are defined in a standard, but equally important these annotations support all features of Declarative Services R6 – while the Apache Felix annotations only support R5 and it is very unlikely that they will be updated. And that’s the other reason why you should not use the Apache Felix annotations – they are in maintenance mode. Of course, should any bug or problem arise, this will be fixed.

The second question usually is: Should I migrate existing code? There is no need to bulk migrate existing code. As said, the Apache Felix plugin is still maintained, is open source and simply works. However, if you have to touch your code, migrating the annotations might be a good idea. It also gives you the chance to simplify your code by leveraging the R6 features. That said, if you do so, this of course ties your implementation to an R6 implementation of Declarative Services, like Apache Felix SCR, version 2.0 or higher.

Mapping to OSGi Declarative Services Annotations

The below table gives you an overview of how to map the annotations.

Annotation Mapping

Apache Felix SCR AnnotationDescriptionOSGi Declarative Services AnnotationDescription
@ComponentThe @Component annotation marks a Java class to be used as a component.@ComponentThis is more or less a strict one-to-one replacement. Only difference is the default behavior for services. See below.
@ServiceMarks the component as a service and optionally lists the provided services (classes)@ComponentThe OSGi annotation has a service attribute which should be used to list the provided services. Be careful, if your component should not provide any service, set this attribute to an empty array.
@ReferenceReference to services, can be used on unary fields or on class level with bind/unbind methods.@ReferenceField references can directly be migrated, for event based references (methods), the @Reference annotation must be put on the bind method. In addition, more options for field references exist.
@Activate, @Deactivate, and @ModifiedMark a method being the activation, deactivation or modified method@Activate, @Deactivate, and @ModifiedStraight one-to-one migration.
@PropertyConfiguration properties and metatype.Component Property Type and metatype annotationsInstead of using a set of @Property annotations, the configuration is described through an annotation (component property type) and OSGi Metatype annotations can be used to add the metatype info.
Mapping between Apache Felix SCR Annotations and OSGi Declarative Services Annotations

The above table is of course just a short reference. You’ll find more information in the OSGi specifications and in my Declarative Services Tutorials: Part I, Part II, and Part III.

Example OSGi Web Application

In some of my OSGi talks I use the “Guessing Game” as the next great web application to conquer the world. It’s a web app that generates a random number and let’s you guess it, telling you whether you made it, are too high or too low.

You can find the latest sample code on github.

Some slides from the last OSGi Community event featuring this app:

Http Whiteboard – Simply Simple – Part I

Developing web components with OSGi can be very simple if you’re using the OSGi Http Whiteboard Specification (see OSGi Compendium Chapter 140). For example the Apache Felix Http Implementation supports this standard.

In this part of the tutorial we start pretty simple and just look at how to develop a servlet and register it as an active web component. In the next parts we’ll cover more details.

For the example we use Declarative Services (DS) to develop our components. This is the easiest and most elegant way of developing OSGi services and is a standard for OSGi component development (see OSGi Compendium Chapter 112). For a tutorial for Declarative Services see OSGi Components Tutorial.

Registering A Servlet with the Http Whiteboard

Let’s have a look at the typical Hello World Example. The below code registers a servlet which simply prints out Hello World if invoked.

@Component(service = Servlet.class, 
           property= "osgi.http.whiteboard.servlet.pattern=/hello")
public class HelloWorldServlet extends javax.servlet.http.HttpServlet {

    protected void doGet(HttpServletRequest req, HttpServletResponse resp)
    throws ServletException, IOException {
        resp.getWriter().println("Hello World");

With the @Component annotation the HelloWorldServlet is registered as a service of type Servlet with a prototype scope. Using the prototype scope for servlets is recommended and we’ll see in another part why. The more important property is osgi.http.whiteboard.servlet.pattern which defines the path under which the servlet is available, /hello in this case. Once deployed into an OSGi application which has support for the OSGi Http Whiteboard specification, this servlet is reachable using /hello from your browser.

And that’s all you have to do. So getting a servlet up and running comes down to using the @Component annotations and specifying the path. Everything else is taken care of and you can focus on implementing your business logic. Of course all the nice features of Declarative Services can be used within your servlet, like the @Reference annotation to use other services.

OSGi Components – Simply Simple – Part III

In the previous two parts (Part I and Part II) you learned how to use Declarative Services to develop your own components and services for OSGi. The topics included:

  • Creating components
  • Component lifecycle
  • Component configuration
  • Using services through references
  • Providing services

In this part we’ll cover another interesting topic:

Metatype Generation

The metatype specification (OSGi Compendium Chapter 105) provides a way to describe the configuration for a component. In general an OSGi configuration is just a dictionary with arbitrary key-value pairs. By defining a metatype description for the configuration, the developer of a component can define which properties together with their type the component expects. Such a metatype information can be used at runtime to generate forms to edit/create the configuration of a component. For example the Apache Felix Web Console does exactly that.

Now if you look at Part II of this series, if a component is configurable it contains already a component annotation for its properties:

public @interface MyComponentConfig {
    String welcome_message() default "Hello World!";
    int welcome_count() default 3;
    boolean output_goodbye() default true;

The above annotation already defines all possible properties together with their default values. Therefore the only thing you have to do is provide some human readable information like a label and a description for both, the configuration as a whole and each property. This can be done with some annotations from the metatype specification:

import org.osgi.service.metatype.annotations.AttributeDefinition;
import org.osgi.service.metatype.annotations.ObjectClassDefinition;

@ObjectClassDefinition(name = "Hello World Configuration",
		description = "The configuration for the hello world component.")
public @interface MyComponentConfig {
    @AttributeDefinition(name="Welcome Message", description="This message is displayed on startup of the component.")
    String welcome_message() default "Hello World!";
    @AttributeDefinition(name="Welcome Message Count", description="This is the number of times, the welcome message will be displayed. " +
                              "If less than one, no message will be displayed.")
    int welcome_count() default 3;
    @AttributeDefinition(name="Output Goodbye", description="Set this if the component should output a goodbye message.")
    boolean output_goodbye() default true;

The @ObjectClassDefinition annotation is used on the whole configuration annotation and for each attribute, the @AttributeDefinition annotation is used. With these annotations a metatype description for the MyComponentConfig class will be generated. However, we want to bind this configuration to our component which is a different class. Therefore you have to specify the @Designate annotation on your component class:

import org.osgi.service.metatype.annotations.Designate;

@Designate( ocd = MyComponentConfig.class )
public class MyFirstComponent {
    protected void activate( MyComponentConfig config ) {

And this will generate the metatype information for your component. It’s really easy to do such and it helps the user of your component to easily figure out how your component can be configured.

OSGi Components – Simply Simple – Part II

In part I I motivated that developing OSGi components is plain and simple by Declarative Services and the provided annotations. We learned how to create a component, add lifecycle methods and reference services. If you haven’t read the first part yet, make sure to go there first.

Usually a component takes some configuration. So let’s have a look at this next.

Component Configurations

If you want to make your component configurable, the best way of doing this is to leverage the OSGi Configuration Admin (another spec from the OSGi compendium with again a great implementation at Apache Felix). Configuration Admin is a service persisting configurations – these configurations are dictionaries where the key is a string and the value can be one of the simple Java types or an array or collection of such a type. This is usually sufficient for most components.

Configuration Admin nicely abstracts managing the configuration from your component. As a component developer you don’t need to know where the actual configuration is stored and how it ends up there. Without a component container, you would ask the configuration admin for “your” configuration (though there is some injection support through the ManagedService interface) – but fortunately with DS the container takes care of this and provides your component with the configuration. The component might get the configuration as a parameter for the activate method.

As mentioned Configuration Admin stores configurations as dictionaries. Whereas the names are of type String, the value can be any type. While your component might expect an integer, e.g. for a port, the actual value stored might be of type String holding an integer value. Therefore it is good style to not assume a specific type but try to convert whatever you get into the expected type. To avoid putting the burden of doing this on the component developer and to support type safe configurations, the configuration for a component can be described in Java. While the following might look a little bit out of the ordinary, stick with me, you’ll get used to it pretty soon and simply love it.

The way to describe a configuration is by defining a so called Component Property Type: an annotation describing all configuration properties. Let’s assume our component has three configuration properties, then the following annotation would do:

public @interface MyComponentConfig {
    String welcome_message() default "Hello World!";

    int welcome_count() default 3;

    boolean output_goodbye() default true;

Defining an annotation instead of a simple interface has the advantage that we can directly specify default values for each property. If no value for a property is stored in configuration admin, the default value is used. Let’s see how to use this:

import org.osgi.service.component.annotations.Activate;
import org.osgi.service.component.annotations.Component;

public class MyFirstComponent {

    protected void activate( MyComponentConfig config ) {
        for(int i = 0; i < config.welcome_count(); i++ ) {

When the above component is activated, SCR tries to get a configuration from the configuration admin and converts it into the used annotation type. Therefore the passed in object returns the value in the correct type for each property. As mentioned if there is no value, the default is returned. This makes handling of configurations very simple and avoids all the usual boilerplate code. The OSGi specification for Declarative Services explains in detail how the type conversion is done and what happens if a value can’t be converted. It also defines the conversion rule from the name of a property in the annotation to a property in configuration admin. In the example above we see the underscore in the names, this is acutally converted to a dot in the property name.

Taking configurations is as easy as this: define your set of configuration properties as an annotation and pass it as an argument to the activate method. Of course, if you don’t need configuration, leave out these steps and use the zero argument activate method signature – or if you don’t need the activate method at all, leave it out completely.

With this knowledge we can already develop configurable components capable of using other services. But how do you provide a service for others to be used?


Offering a service usually consists of two steps. First you define the service API and second you implement this API. Of course, in some cases the API might already exist as someone else defined the interface. If you define your own interface, put it into a public package and export this package – however, as already noted above for the component, a service implementation should never be public but in a private package. It depends on your use case, if you put the interface and the implementation in the same bundle or create an API bundle and an implementation bundle. If you’re implementing an existing interface, this usually is already in another bundle and you can just import it. And if you create public API, don’t forget to use proper versioning for these packages. Have a look at the semantic versioning whitepaper!

Offering a service is implementing the corresponding interface and registering the component in the OSGi service registry as the service. With the annotations you can just use the @Component annotation. Assume there is an interface EventHandler and your component should be registered as a service for this interface:

import org.osgi.service.component.annotations.Component;

public class MyFirstService implements EventHandler {

   ....// implement the EventHandler interface


Without further specifying anything at the @Component annotation, a component is registered as a  service for all interfaces it directly implements. In this case this is the EventHandler. While this is very convenient, it comes with the problem, that as soon as your component is implementing an interface directly, it gets registered as this service. In some cases this is not what you want. Therefore I suggest to:

  1. Always explicitely list the service interfaces the component implements:
    @Component(service = EventHandler.class}
  2. As a default always set the service attribute of the component annotations to the empty array. This prevents the automatic registration:
    @Component(service = {})

If you follow these simple rules, you can see directly by looking at the @Component annotation which services this component implements and you avoid accidental service registrations, e.g. when refactoring your implementation.

Lifecycle of a Service

A component which does not provide a service is active for as long as it is satisfied (all referenced services are available and some other conditions we get to later). In contrast, services are instantiated lazy or on demand by DS. This means, as long as no one is using your service, your service is never created nor activated! In most cases this is fine. But there is a catch with this approach one should be aware of: if someone else is starting to use your service, it gets created and activated. As soon as your service is not used anymore, it gets deactivated and destroyed.

The OSGi spec does not mandate this behavior. The implementation of Declarative Service is free to keep your service around for some time until it is disposed. By default the current Apache Felix implementation immediately disposes such components. However it is possible to configure such a detailed disposal of components. But this leaves you with the problem of configuring this correctly which might not be that easy.

On the other hand, frequent creation and disposal of an “expensive” service might create a performance bottleneck. For example if this happens being triggered by an event, a request and/or if your service is doing some computation in the activate method.

In many cases, your actual service is combining a “component” and a “service” part. While the “component” part is the expensive one and should only be done once, the “service” part is lightweight and might simply use the component part. In such cases you might think about splitting your implementation accordingly.

Or you can either think about holding the service by someone else and therefore keep a reference to it around (which in general sounds hacky though there are valid use cases). Or you can use the immediate flag on the @Component:

@Component(service=EventHandler.class, immediate=true)

With immediate set to true, the component is activated as soon as possible and kept as long as possible. Obviously, this increases things like startup time, memory consumption etc. So it should really be used with care and maybe only after problems are encountered in this area.

In the past the only situation where we encountered this was implementing an event handler. But with today’s event admin implementations this isn’t even true anymore either. But for completeness lets have a look at that problem: an implementation of the event admin as defined in the OSGi Event Admin Specification might fetch an event handler (service org.osgi.service.event.EventHandler) each time an event should be delivered to this handler. Clearly, this has the advantage that event handlers are only used if there is an event for such a handler. While this might work with a few events, with very frequent events, especially in a multi threaded environment, the same event handler might receive quiet a lot events, even “in parallel”.

In the past we suggest to make your event handler implementation immediate as otherwise it is potentially created/destroyed with each event for this handler! However, even this is up to the implementation of the event admin. The latest Apache Felix event admin implementation and the Equinox event admin implementation do not create/destroy the instance on each request, they rather create it once the first event for this component arrives and keep a reference to it from that point on. But as this is implementation specific, the above advice might be handy. And again, only use immediate if really required.

And with this the simple introduction to Declarative Services ends: you now know:

  • how to create components
  • how to configure components
  • how to reference services
  • how to provide services

Of course with most of this we only touched the simple case, which is really sufficient for most use cases. I’ll continue this series with more advanced stuff in the future.