In equinox OSGi, I am using a service (DS) from several different bundles and I would like to know in the service who is using it, each time.
The service write to the database, and I want to know which bundle writes what.
The buildin LogService must be able to do this, since it knows who wrote each log line, but I can't figure out how.
The simplest solution is to put the name of the bundle in each method to the service, but I hope for a more elegant solution.
This is precisely what a ServiceFactory is for, see OSGi Core R6 specification Section 5.9. "Service Factory".
Updated Below after question clarified to specify DS usage.
This can be achieved by using scope=ServiceScope.BUNDLE on your #Component annotation. Then you can access the calling bundle by allowing the ComponentContext to be injected into your activate method and calling getUsingBundle(). For example:
#Component(scope = ServiceScope.BUNDLE)
public class MyComponent implements MyService {
private Bundle usingBundle;
#Activate
void activate(ComponentContext context) {
this.usingBundle = context.getUsingBundle();
}
// ...
}
At the low-level this works by registering the service as an instance of ServiceFactory rather than a plain service object. When OSGi obtains the service on behalf of a consumer, it invokes the getService method, which passes the consumer Bundle object to the provider of the service. This happens entirely transparently for the consumer, i.e. they don't have to change their code at all.
Add a bundle name or a Bundle object to the object you pass from service clients to the service. By doing so, you can find out the client for each service call.
You mentioned LogService. It uses a LogEntry class to pass a Bundle object with the log message to the service.
Related
I've spent some time experimenting with and studying the OSGi enRoute site. The Quick Start, and Base tutorials were really good. Now as a learning exercise, I'm creating my own example following the principles in those tutorials.
I've decided to reproduce the StageService from the blog post "Making JavaFX better with OSGi". Rather than using the org.apache.felix.dm and org.apache.felix.dm.annotation.api packages I want to use the OSGi standard SCR packages (org.osgi.service.component.*) along with the enRoute provider template.
So far everything has worked out nicely. But I'm stuck on one point. In the "Making JavaFX better with OSGi" tutorial the service is programmatically registered into the service registry using the org.apache.felix.dm.DependencyManager like this:
#Override
public void start(Stage primaryStage) throws Exception {
BundleContext bc = FrameworkUtil.getBundle(this.getClass()).getBundleContext();
DependencyManager dm = new DependencyManager(bc);
dm.add(dm.createComponent()
.setInterface(StageService.class.getName(), null)
.setImplementation(new StageServiceImpl(primaryStage)));
}
My assumption is that in this example the DependencyManager is an Apache Felix specific feature rather than an OSGi standard. I would like to have my enRoute provider depend only on OSGi standard features.
So my question is simply:
How would one register a service in the service registry programmatically using only OSGi standard features? (I know from following the enRoute tutorials that if my component implements the exported service that SCR will automatically register my component in the service registry when my component is activated. The problem with this solution though is that when my component is activated it has to launch the JavaFX application in a different thread so as to not block the thread in use by the SCR until the JavaFX application terminates. Because of this, my component must programmatically register the service in the service registry. Otherwise it won't be guaranteed to be available upon registration.)
For reference, here is what I currently have:
private void registerService(Stage stage) {
DependencyManager dm = new DependencyManager(bundle().getBundleContext());
dm.add(
dm.createComponent()
.setInterface(StageService.class.getName(), null)
.setImplementation(new StageServiceImpl(primaryStage))
);
}
But instead I want to replace it with this:
private void registerService(Stage stage) {
// How to register service in service registry using only OSGi standard features? (not the apache felix dependency manager)
}
UPDATE 1
Following BJ Hargrave's recommendation I tried to register the service directly from the bundle context as follows:
FrameworkUtil
.getBundle(getClass())
.getBundleContext()
.registerService(StageService.class, new StageServiceImpl(primaryStage), null);
After doing this and trying to resolve the enRoute application project the following error occurs:
org.osgi.service.resolver.ResolutionException: Unable to resolve
<> version=null: missing requirement
com.github.axiopisty.osgi.javafx.launcher.application
-> Unable to resolve com.github.axiopisty.osgi.javafx.launcher.application
version=1.0.0.201608172037: missing requirement
objectClass=com.github.axiopisty.osgi.javafx.launcher.api.StageService]
I have uploaded the project to github so you can reproduce the error.
Update 2
The build tab in the bnd.bnd file in the provider module shows the following warning:
The servicefactory:=true directive is set but no service is provided, ignoring it
Might this have something to do with the application module not being able to be resolved?
In rare cases it is necessary to register a 'service by hand' using the standard OSGi API. Try very hard to avoid this case because if you start to register (and maybe depend) on services that you manually register you get a lot of responsibility that is normally hidden from view. For example, you have to ensure that the services you register are also unregistered.
One of the rare cases where this is necessary is when you have to wait for a condition before you can register your service. For example, you need to poll a piece of hardware before you register a service for the device. You will need to control the CPU but at that moment you cannot yet register a service. In that case you create an immediate component and register the service manually.
To register a service manually you require a BundleContext object. You can get that objectvia the activate method, just declare a Bundle Context in its arguments and it is automatically injected:
#Activate
void activate( BundleContext context) {
this.context = context;
}
You can now register a service with the bundle context:
void register(MyService service) {
Hashtable<String,Object> properties = new Hashtable<>();
properties.put("foo", "bar");
this.registration = context.registerService( MyService.class, service, properties );
}
However, you now have the responsibility to unregister this service in your deactivate. If you do not clean up this service then your component might be deactivated while your service still floats around. Your service is unmanaged. (Although when the bundle is stopped it will be cleaned up.)
#Deactivate
void deactivate() {
if ( this.registration != null)
this.registration.unregister();
}
If you create the service is a call back or background thread then you obviously have to handle the concurrency issues. You must ensure that there is no race condition that you register a service while the deactivate method has finished.
This text has also been added to the DS Page of OSGi enRoute
Reading the OSGi spec would help you understand the service API.
But this should do it:
ServiceRegistration<StageService> reg = bc.registerService(StageService.class, new StageServiceImpl(primaryStage), null);
I am developing an application in Eclipse RCP. I need help with a design decision concerning the design of a service.
I have some bundles which are used to provide an REngine object to other modules. The REngine is an interface to a calculation engine, which can be implemented in multiple ways. The bundles provide instances of REngine by connecting to a remote server or starting a local calculation thread. Some bundles require configuration by the GUI (but also need to be available on a headless platform). A client Bundle may request multiple REngine objects for parallel calculations.
I currently register these modules to provide an REngine service. The service is created by a ServiceFactory, which either launches a local calculation instance or a remote (server) instance. The client is responsible for trying out all Service registrations of the REngine class and selecting the right one.
The code to do this can be summarised as follows:
class API.REngine { ... }
class REngineProvider.Activator {
public void start(BundleContext ctx) {
ctx.registerService(REngine.class.getName(), new REngineFactory(), null);
}
}
class REngineProvider.REngineFactory implements ServiceFactory {
public Object getService(Bundle bundle, ServiceReference reference) {
return new MyREngineImplementation();
}
public void ungetService(REngine service) {
service.releaseAssociatedResources();
}
}
class RConsumer.Class {
REngine getREngine() {
ServiceReference[] references = bundleContext.getAllServiceReferences(REngine.class.getName(), null);
for(ServiceReference ref: references) {
try {
return bundleContext.getService(ref);
} catch (Exception e) {} // too bad, try the next one
}
}
}
I would like to keep this model. It is nice that the OSGi service spec matches my business requirement that REngine objects are living objects which should be released when they are not needed anymore.
However, a registered service can only provide one service instance per bundle. The second time the service is requested, a cached instance is returned (instead of creating a new one). This does not match my requirement; a bundle should be able to get multiple REngine objects from the same provider.
I have already looked at other OSGi framework classes, but nothing seems to help. The alternative is the whiteboard model, but it seems strange to register an REngineRequestService that is used by the REngineProvider bundle to give out a live REngine.
How do I implement this in OSGi? As a reminder, here is my list of requirements:
Easy enabling and disabling of REngineProvider bundles. Client code will just use another provider instead.
Configuration of REngineProvider bundles.
Multiple REngine instances per client bundle.
Explicit release of REngine instances
REngine creation can fail. The client module should be able to know the reason why.
Just to add the solution I have chosen as future reference. It seems the OSGi Services platform is not made for "requesting a service". It is the provider bundle that creates a service, and the client bundle that can find and use the services. It is not possible to provide an automatic "Factory" for services per user request.
The solution chosen involves the OSGi whiteboard model. On first sight, this may seem very difficult to manage, but Blueprint can help a lot!
Provider blueprint.xml file:
<reference-list interface="org.application.REngineRequest"
availability="optional">
<reference-listener
bind-method="bind" unbind-method="unbind">
<bean class="org.provider.REngineProvider"/>
</reference-listener>
The class REngineRequest is a shared API class allowing the provider to input his REngine object, or set an Exception explaining why the creation did not work.
For the client, using an REngine is now as easy as doing:
REngineRequest req = new REngineRequest();
ServiceRegistration reg = bundleContext.registerService(req, REngineRequest.class.getName(), engineCreationProperties);
req.getEngine().doSomeStuff();
reg.unregister();
We make the assumption that the provider will never stop while the client is using the REngine. If it does, the REngine becomes invalid.
ComponentFactory from the Declarative Services is what you need. Most of the time you should rather use DS instead of manually registering and looking up the services.
The provider side should register REngine factory service (you don't have to implement the factory itself, DS will do that for you). The consmer should declare one-to-many dependency to the REngine service. At runtime, all available factories will be injected and consumer can go through them to create actual REngine instances.
Two years ago I tried to create Genuine Service Factories what later became Parameterized Services. However, after analysis it turned out that nothing was needed, just register the factory as the service.
However.
I do not know enough about your service but it sounds very much that you could significantly simplify things by removing control from the client bundle, the client bundle should just use whatever REngine service is available in the service registry, maybe with a property signalling its use type if there are multiple bundles that need REngines and they should not share the same REngine (which should rarely be the case).
If that model is possible, it usually significantly simplifies. I generally then use DS with Configuration Admin configurations that drive the instances (one of the most useful aspects of DS, see http://www.aqute.biz/Bnd/Components). With the metatype integration, you even get a user interface to edit your configuration properties.
One solution would be to register the REngineFactory as the service rather than the REngine implementation itself and return the factory from the getService method. That way the clients can look up the factory and, on successfully finding one, use it to get a new REngine implementation.
I want to create a new system that will be built completely using services. I want to expose these services thru REST for client applications. But for performance reasons, I also want to make sure that other services can call a given service using local calls without paying the penalty of a remote call. Is there a framework that can help me do that.
Well, the way we have implemented this is by using something like Spring MVC where the controller just calls out to a Service class - our notion of Model. The Controller thus acts as "exposing the services" as RESTful services. The rest of the codebase accesses these Services just like any other object. Since we use spring, we leverage the IOC massively.
For example, we would have something like:
public class BillingService {
public void doSomething(String someParam) {}
}
public class BillingController {
#Autowired private BillingService billingService;
public void doSomething(#RequestParam String someParam) {
billingService.doSomething(someParam);
}
}
In the above examples, the annotations are all from Spring, but you get the picture. Any other class which wants to access the BillingService method, they can do so just by accessing that class's method.
I am not sure of any framework which is targeted at exactly this problem, but my guess is, you do not need one.
I have a large web project in Java EE 6 so far everything is working great.
Now I'm adding a new class that takes twitter information and returns a string. So far the strings have been extracted from the JSON file from twitter and are ready to be persisted in my database. My problem is I'm not sure how to pass information from the EJB that normally handles all of my database calls. I'm using JPA and have a DAO class that managers all database access. I already have a method there for updateDatabase(String). I'd like to be able to call updateDatabase(String) from the class that has the strings to add but I don't know if it's good form to instantiate a stateless bean like that. Normally you inject beans and then call just their class name to access their methods. I could also maybe try and reference the twitter string generating class from inside of the EJB but then I'd have to instantiate it there and mess with main() method calls for execution. I'm not really sure how to do this. Right now my Twitter consuming class is just a POJO with a main method. For some reason some of the library methods did not work outside of main in face IOUtils() API directly says "Instances should NOT be constructed in standard programming".
So on a higher level bottom line, I'm just asking how POJO's are normally "mixed" into a Java EE project where most of your classes are EJBs and servlets.
Edit: the above seems confusing to me after rereading so I'll try to simplify it. basically I have a class with a main method. I'd like to call my EJB class that handles database access and call it's updateDatabase(String) method and just pass in the string. How should I do this?
Edit: So it looks like a JNDI lookup and subsequence reference is the preferred way to do this rather than instantiating the EJB directly?
Edit: these classes are all in the same web project. In the same package. I could inject one or convert the POJO to an EJB. However the POJO does have a main method and some of the library files do not like to be instantiated so running it in main seems like the best option.
My main code:
public class Driver {
#EJB
static RSSbean rssbean;
public static void main(String[] args) throws Exception {
System.setProperty("http.proxyHost", "proxya..com");
System.setProperty("http.proxyPort", "8080");
/////////////auth code///////////////auth code/////////////////
String username = System.getProperty("proxy.authentication.username");
String password = System.getProperty("proxy.authentication.password");
if (username == null) {
Authenticator.setDefault(new ProxyAuthenticator("", ""));
}
///////////////end auth code/////////////////////////////////end
URL twitterSource = new URL("http://search.twitter.com/search.json?q=google");
ByteArrayOutputStream urlOutputStream = new ByteArrayOutputStream();
IOUtils.copy(twitterSource.openStream(), urlOutputStream);
String urlContents = urlOutputStream.toString();
JSONObject thisobject = new JSONObject(urlContents);
JSONArray names = thisobject.names();
JSONArray asArray = thisobject.toJSONArray(names);
JSONArray resultsArray = thisobject.getJSONArray("results");
JSONObject(urlContents.substring(urlContents.indexOf('s')));
JSONObject jsonObject = resultsArray.getJSONObject(0);
String twitterText = jsonObject.getString("text");
rssbean.updateDatabase("twitterText");
}
}
I'm also getting a java.lang.NullPointerException somewhere around rssbean.updateDatabase("twitterText");
You should use InitialContext#lookup method to obtain EJB reference from an application server.
For example:
#Stateless(name="myEJB")
public class MyEJB {
public void ejbMethod() {
// business logic
}
}
public class TestEJB {
public static void main() {
MyEJB ejbRef = (MyEJB) new InitialContext().lookup("java:comp/env/myEJB");
ejbRef.ejbMethod();
}
}
However, note that the ejb name used for lookup may be vendor-specific. Also, EJB 3.1 introduces the idea of portable JNDI names which should work for every application server.
Use the POJO as a stateless EJB, there's nothing wrong with that approach.
From the wikipedia: EJB is a server-side model that encapsulates the business logic of an application.
Your POJO class consumes a web service, so it performs a business logic for you.
EDIT > Upon reading your comment, are you trying to access an EJB from outside of the Java EE container? Because if not, then you can inject your EJB into another EJB (they HAVE to be Stateless, both of them)
If you have a stand alone program that wishes to access an EJB you have a couple of options.
One is to simply use JNDI to look up the EJB. The EJB must have a Remote interface, and you need to configure the JNDI part for you container, as well as include any specific container jars within your stand alone application.
Another technique is to use the Java EE artifact know as the "application client". Here, there is a container provider wrapper for your class, but it provides a run time environment very similar to running the class within the container, notably you get things like EJB injection.
You app still runs in a separate JVM, so you still need to reference Remote EJBs, but the app client container handles a bunch of the boiler plate in getting your app connected to the server. This, too, while a Java EE artifact, is also container dependent in how to configure and launch an app client application.
Finally, there is basically little difference in how a POJO interact with the EJB container this way in contrast to a POJO deployed within the container. The interface is still a matter of getting the EJB injected (more easily done in Java EE 6 than before) or looking up a reference via JNDI. The only significant difference being that a POJO deployed in the container can use a Local interface instead of the Remote.
I am wondering if it possible to do the equivalent of
public void start(BundleContext context)
{
String filter = "filter for my specific service";
context.addServiceListener(new MyServiceListener(), filter);
}
with Spring DM. I have found reference material for adding listeners for lifecycle management. But this relates to the lifecycle of the service being registered. What I need is to be made aware of the lifecycle of any service that is registered/unregistered of a specific service interface.
The bundle that needs to know this is not actually creating the service, but will be using it. Since there will multiple instances of this service that will change at runtime, I cannot simply create a service reference in my Spring configuration.
Spring DM includes support for referencing a collection of services. Unfortunately, as a new user, I can't post links, but this functionality is described in section 7.2.2 of the Spring DM 1.2 documentation. Spring DM automatically updates the contents of this collection at runtime as services are added to and removed from the service registry.
When you declare a reference upon a service, be it as a single reference, or a collection of services, you can also declare a listener bean that will be notified as services come and go. This is documented in section 7.2.3.
I'd recommend making use a List or Set of services, and using a listener bean so that you're notified of changes to the collection's contents.