SpringBoot Undertow : how to dispatch to worker thread - spring-mvc

i'm currently have a look a springboot undertow and it's not really clear (for me) how to dispatch an incoming http request to a worker thread for blocking operation handling.
Looking at the class UndertowEmbeddedServletContainer.class, it look like there is no way to have this behaviour since the only HttpHandler is a ServletHandler, that allow #Controller configurations
private Undertow createUndertowServer() {
try {
HttpHandler servletHandler = this.manager.start();
this.builder.setHandler(getContextHandler(servletHandler));
return this.builder.build();
}
catch (ServletException ex) {
throw new EmbeddedServletContainerException(
"Unable to start embdedded Undertow", ex);
}
}
private HttpHandler getContextHandler(HttpHandler servletHandler) {
if (StringUtils.isEmpty(this.contextPath)) {
return servletHandler;
}
return Handlers.path().addPrefixPath(this.contextPath, servletHandler);
}
By default, in undertow all requests are handled by IO-Thread for non blocking operations.
Does this mean that every #Controller executions will be processed by a non blocking thread ? or is there a solution to chose from IO-THREAD or WORKER-THREAD ?
I try to write a workaround, but this code is pretty uggly, and maybe someone has a better solution:
BlockingHandler.class
#Target({ElementType.TYPE})
#Retention(RetentionPolicy.RUNTIME)
#Documented
public #interface BlockingHandler {
String contextPath() default "/";
}
UndertowInitializer.class
public class UndertowInitializer implements ApplicationContextInitializer<ConfigurableApplicationContext> {
#Override
public void initialize(ConfigurableApplicationContext configurableApplicationContext) {
configurableApplicationContext.addBeanFactoryPostProcessor(new UndertowHandlerPostProcessor());
}
}
UndertowHandlerPostProcessor.class
public class UndertowHandlerPostProcessor implements BeanDefinitionRegistryPostProcessor {
#Override
public void postProcessBeanDefinitionRegistry(BeanDefinitionRegistry beanDefinitionRegistry) throws BeansException {
ClassPathScanningCandidateComponentProvider scanner = new ClassPathScanningCandidateComponentProvider(false);
scanner.addIncludeFilter(new AnnotationTypeFilter(BlockingHandler.class));
for (BeanDefinition beanDefinition : scanner.findCandidateComponents("org.me.lah")){
try{
Class clazz = Class.forName(beanDefinition.getBeanClassName());
beanDefinitionRegistry.registerBeanDefinition(clazz.getSimpleName(), beanDefinition);
} catch (ClassNotFoundException e) {
throw new BeanCreationException(format("Unable to create bean %s", beanDefinition.getBeanClassName()), e);
}
}
}
#Override
public void postProcessBeanFactory(ConfigurableListableBeanFactory configurableListableBeanFactory) throws BeansException {
//no need to post process defined bean
}
}
override UndertowEmbeddedServletContainerFactory.class
public class UndertowEmbeddedServletContainerFactory extends AbstractEmbeddedServletContainerFactory implements ResourceLoaderAware, ApplicationContextAware {
private ApplicationContext applicationContext;
#Override
public EmbeddedServletContainer getEmbeddedServletContainer(ServletContextInitializer... initializers) {
DeploymentManager manager = createDeploymentManager(initializers);
int port = getPort();
if (port == 0) {
port = SocketUtils.findAvailableTcpPort(40000);
}
Undertow.Builder builder = createBuilder(port);
Map<String, Object> handlers = applicationContext.getBeansWithAnnotation(BlockingHandler.class);
return new UndertowEmbeddedServletContainer(builder, manager, getContextPath(),
port, port >= 0, handlers);
}
#Override
public void setApplicationContext(ApplicationContext applicationContext) throws BeansException {
this.applicationContext = applicationContext;
}
}
...
override UndertowEmbeddedServletContainer.class
public UndertowEmbeddedServletContainer(Builder builder, DeploymentManager manager,
String contextPath, int port, boolean autoStart, Map<String, Object> handlers) {
this.builder = builder;
this.manager = manager;
this.contextPath = contextPath;
this.port = port;
this.autoStart = autoStart;
this.handlers = handlers;
}
private Undertow createUndertowServer() {
try {
HttpHandler servletHandler = this.manager.start();
String path = this.contextPath.isEmpty() ? "/" : this.contextPath;
PathHandler pathHandler = Handlers.path().addPrefixPath(path, servletHandler);
for(Entry<String, Object> entry : handlers.entrySet()){
Annotation annotation = entry.getValue().getClass().getDeclaredAnnotation(BlockingHandler.class);
System.out.println(((BlockingHandler) annotation).contextPath());
pathHandler.addPrefixPath(((BlockingHandler) annotation).contextPath(), (HttpHandler) entry.getValue());
}
this.builder.setHandler(pathHandler);
return this.builder.build();
}
catch (ServletException ex) {
throw new EmbeddedServletContainerException(
"Unable to start embdedded Undertow", ex);
}
}
set initializer to the application context
public static void main(String[] args) {
new SpringApplicationBuilder(Application.class).initializers(new UndertowInitializer()).run(args);
}
finaly create a HttpHandler that dispatch to worker thread
#BlockingHandler(contextPath = "/blocking/test")
public class DatabaseHandler implements HttpHandler {
#Autowired
private EchoService echoService;
#Override
public void handleRequest(HttpServerExchange httpServerExchange) throws Exception {
if(httpServerExchange.isInIoThread()){
httpServerExchange.dispatch();
}
echoService.getMessage("my message");
}
}
As you can see, my "solution" is really heavy, and i would really appreciate any help to simplify it a lot.
Thank you

You don't need to do anything.
Spring Boot's default Undertow configuration uses Undertow's ServletInitialHandler in front of Spring MVC's DispatcherServlet. This handler performs the exchange.isInIoThread() check and calls dispatch() if necessary.
If you place a breakpoint in your #Controller, you'll see that it's called on a thread named XNIO-1 task-n which is a worker thread (the IO threads are named XNIO-1 I/O-n).

Related

Axon: Sending multiple commands to the same aggregate

If I send multiple commands to the same aggregate, only the first is handled.
Is this a configuration problem, or am I missing something?
The message I am getting after the 2nd command is send:
org.springframework.web.util.NestedServletException: Request processing failed; nested exception is org.axonframework.commandhandling.CommandExecutionException: Cannot invoke "Object.hashCode()" because "key" is null
The service method where I do my sending of the command is:
public void maakAanvraag() {
UUID aanvraagId = UUID.randomUUID();
commandGateway.sendAndWait(
VerwerkAanvraag.builder()
.aanvraagId(aanvraagId)
.build()
);
commandGateway.sendAndWait(
VerwerkPersoonsgegevensVastgesteld.builder()
.aanvraagId(aanvraagId)
.build()
);
commandGateway.sendAndWait(
VerwerkOrganisatiegegevensVastgesteld.builder()
.aanvraagId(aanvraagId)
.organisatieId(organisatieView.getOrganisatieId())
.rolOrganisatie(rolOrganisatie)
.build()
);
commandGateway.sendAndWait(
VerwerkBeperkingErkenningsdoelGematcht.builder()
.aanvraagId(aanvraagId)
.build());
}
The aggregate I am using is:
#Aggregate
#Getter
#NoArgsConstructor
public class Aanvraag {
public static final String META_DATA_ZAAKNUMMER = "aanvraag_zaaknummer";
#AggregateIdentifier
private UUID aanvraagId;
#CommandHandler
public Aanvraag(VerwerkAanvraag command) {
AanvraagGeregistreerd aanvraagGeregistreerd =
AanvraagGeregistreerd.builder()
.aanvraagId(command.getAanvraagId())
.build();
apply(aanvraagGeregistreerd, MetaData.with(META_DATA_ZAAKNUMMER, "123456789"));
}
#EventSourcingHandler
public void on(AanvraagGeregistreerd event) {
aanvraagId = event.getAanvraagId();
}
#CommandHandler
public void verwerkOrganisatiegegevensVastgesteld(VerwerkOrganisatiegegevensVastgesteld command) {
OrganisatiegegevensVastgesteld persoonsgegevensVastgesteld =
OrganisatiegegevensVastgesteld.builder()
.aanvraagId(command.getAanvraagId())
.build();
apply(persoonsgegevensVastgesteld);
}
#EventSourcingHandler
public void on(OrganisatiegegevensVastgesteld event) {
aanvraagId = event.getAanvraagId();
}
#CommandHandler
public void verwerkPersoonsgegevensVastgesteld(VerwerkPersoonsgegevensVastgesteld command) {
PersoonsgegevensVastgesteld persoonsgegevensVastgesteld =
PersoonsgegevensVastgesteld.builder()
.aanvraagId(command.getAanvraagId())
.build();
apply(persoonsgegevensVastgesteld);
}
#EventSourcingHandler
public void on(PersoonsgegevensVastgesteld event) {
aanvraagId = event.getAanvraagId();
}
#CommandHandler
public void verwerkBeperkingErkenningsdoelGematcht(VerwerkBeperkingErkenningsdoelGematcht command) {
BeperkingErkenningsdoelGematcht beperkingErkenningsdoelGematcht =
BeperkingErkenningsdoelGematcht.builder()
.aanvraagId(command.getAanvraagId())
.build();
apply(beperkingErkenningsdoelGematcht);
}
#EventSourcingHandler
public void on(BeperkingErkenningsdoelGematcht event) {
aanvraagId = event.getAanvraagId();
}
}
The project uses Spring Boot 2.6.6 with axon-spring-boot-starter 4.5.9
It al runs with Java Temurin 17.0.3
We solved the problem.
Issue had nothing to do with Axon.
The problem was an Logging interceptor.
After removing this log interceptor the Axon worked as expected.

processing strategy of message in spring kafka listener

Just wanted to make sure that whether messages are processed in correct way or not. When the message gets received at listener, it will be always processed by a new thread( defined the processor bean as prototype). is this implementation correct ? (i have Considered the listener is not thread safe, so for this reason the prototype scope of bean to process the message has been used)
(Input : TestTopic- 5 partitions - 1 consumer) or (Input : TestTopic- 5 partitions - 5 consumers)
public class EventListener {
#Autowired
private EventProcessor eventProcessor;
#KafkaListener(topics = "TestTopic", containerFactory = "kafkaListenerContainerFactory",
autoStartup = "true")
public void onMessage(
#Payload List<ConsumerRecord<String, String>> consumerRecords, Acknowledgment acknowledgment) {
eventProcessor.processAndAcknowledgeBatchMessages(consumerRecords, acknowledgment);
}
}
//event processor
#Slf4j
#Component
#Scope(ConfigurableBeanFactory.SCOPE_PROTOTYPE)
#NoArgsConstructor
#SuppressWarnings("unused")
public class EventProcessorImpl implements EventProcessor {
#Autowired
private KafkaProducerTemplate kafkaProducerTemplate;
#Autowired
private ObjectMapper localObjectMapper;
#Autowired
private Dao dao;
public void processAndAcknowledgeBatchMessages(
List<ConsumerRecord<String, String>> consumerRecords, Acknowledgment acknowledgment) {
long start = System.currentTimeMillis();
consumerRecords.forEach( consumerRecord -> {
Event event = localObjectMapper.readValue(consumerRecord.value(), Event.class);
dao.save(process(event));
});
acknowledgment.acknowledge();
}
}
No it is not correct; you should not execute on another thread; it will cause problems with committing offsets and error handling.
Also, making the EventProcessorImpl a prototype bean won't help. That just means a new instance is used each time the bean is referenced.
Since it is #Autowired it is only referenced once, during initialization. To get a new instance for each request, you would need to call getBean() on the application context each time.
It is better to make your code thread-safe.
EDIT
There are (at least) a couple of ways to deal with a not thread-safe service defined in prototype scope.
Use a ThreadLocal:
#SpringBootApplication
public class So68447863Application {
public static void main(String[] args) {
SpringApplication.run(So68447863Application.class, args);
}
private static final ThreadLocal<NotThreadSafeService> SERVICES = new ThreadLocal<>();
#Autowired
ApplicationContext context;
#KafkaListener(id = "so68447863", topics = "so68447863", concurrency = "5")
void listen(String in) {
NotThreadSafeService service = SERVICES.get();
if (service == null) {
service = this.context.getBean(NotThreadSafeService.class);
SERVICES.set(service);
}
service.process(in);
}
#EventListener
void removeService(ConsumerStoppedEvent event) {
System.out.println("Consumer stopped; removing TL");
SERVICES.remove();
}
#Bean
NewTopic topic() {
return TopicBuilder.name("so68447863").partitions(10).replicas(1).build();
}
#Bean
#Scope(ConfigurableBeanFactory.SCOPE_PROTOTYPE)
NotThreadSafeService service() {
return new NotThreadSafeService();
}
}
class NotThreadSafeService {
void process(String msg) {
System.out.println(msg + " processed by " + this);
}
}
Use a pool of instances.
#SpringBootApplication
public class So68447863Application {
public static void main(String[] args) {
SpringApplication.run(So68447863Application.class, args);
}
private static final BlockingQueue<NotThreadSafeService> SERVICES = new LinkedBlockingQueue<>();
#Autowired
ApplicationContext context;
#KafkaListener(id = "so68447863", topics = "so68447863", concurrency = "5")
void listen(String in) {
NotThreadSafeService service = SERVICES.poll();
if (service == null) {
service = this.context.getBean(NotThreadSafeService.class);
}
try {
service.process(in);
}
finally {
SERVICES.add(service);
}
}
#Bean
NewTopic topic() {
return TopicBuilder.name("so68447863").partitions(10).replicas(1).build();
}
#Bean
#Scope(ConfigurableBeanFactory.SCOPE_PROTOTYPE)
NotThreadSafeService service() {
return new NotThreadSafeService();
}
}
class NotThreadSafeService {
void process(String msg) {
System.out.println(msg + " processed by " + this);
}
}

Logging MDC with #Async and TaskDecorator

Using Spring MVC, I have the following setup:
An AbstractRequestLoggingFilter derived filter for logging requests.
A TaskDecorator to marshal the MDC context mapping from the web request thread to the #Async thread.
I'm attempting to collect context info using MDC (or a ThreadLocal object) for all components involved in handling the request.
I can correctly retrieve the MDC context info from the #Async thread. However, if the #Async thread were to add context info to the MDC, how can I now marshal the MDC context info to the thread that handles the response?
TaskDecorator
public class MdcTaskDecorator implements TaskDecorator {
#Override
public Runnable decorate(Runnable runnable) {
// Web thread context
// Get the logging MDC context
Map<String, String> contextMap = MDC.getCopyOfContextMap();
return () -> {
try {
// #Async thread context
// Restore the web thread MDC context
if(contextMap != null) {
MDC.setContextMap(contextMap);
}
else {
MDC.clear();
}
// Run the new thread
runnable.run();
}
finally {
MDC.clear();
}
};
}
}
Async method
#Async
public CompletableFuture<String> doSomething_Async() {
MDC.put("doSomething", "started");
return doit();
}
Logging Filter
public class ServletLoggingFilter extends AbstractRequestLoggingFilter {
#Override
protected void beforeRequest(HttpServletRequest request, String message) {
MDC.put("webthread", Thread.currentThread().getName()); // Will be webthread-1
}
#Override
protected void afterRequest(HttpServletRequest request, String message) {
MDC.put("responsethread", Thread.currentThread().getName()); // Will be webthread-2
String s = MDC.get("doSomething"); // Will be null
// logthis();
}
}
I hope you have solved the problem, but if you did not, here comes a solution.
All you have to do can be summarized as following 2 simple steps:
Keep your class MdcTaskDecorator.
Extends AsyncConfigurerSupport for your main class and override getAsyncExecutor() to set decorator with your customized one as follows:
public class AsyncTaskDecoratorApplication extends AsyncConfigurerSupport {
#Override
public Executor getAsyncExecutor() {
ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
executor.setTaskDecorator(new MdcTaskDecorator());
executor.initialize();
return executor;
}
public static void main(String[] args) {
SpringApplication.run(AsyncTaskdecoratorApplication.class, args);
}
}
Create a bean that will pass the MDC properties from parent thread to the successor thread.
#Configuration
#Slf4j
public class AsyncMDCConfiguration {
#Bean
public Executor asyncExecutor() {
ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
executor.setTaskDecorator(new MDCTaskDecorator());//MDCTaskDecorator i s a custom created class thet implements TaskDecorator that is reponsible for passing on the MDC properties
executor.initialize();
return executor;
}
}
#Slf4j
public class MDCTaskDecorator implements TaskDecorator {
#Override
public Runnable decorate(Runnable runnable) {
Map<String, String> contextMap = MDC.getCopyOfContextMap();
return () -> {
try {
MDC.setContextMap(contextMap);
runnable.run();
} finally {
MDC.clear();
}
};
}
}
All Good now. Happy Coding
I have some solutions that roughly divided into Callable(for #Async), AsyncExecutionInterceptor(for #Async), CallableProcessingInterceptor(for controller).
1.The Callable solution for putting context infos into #Async thread:
The key is using the ContextAwarePoolExecutor to replace the default executor of #Async:
#Configuration
public class DemoExecutorConfig {
#Bean("demoExecutor")
public Executor contextAwarePoolExecutor() {
return new ContextAwarePoolExecutor();
}
}
And the ContextAwarePoolExecutor overwriting submit and submitListenable methods with ContextAwareCallable inside:
public class ContextAwarePoolExecutor extends ThreadPoolTaskExecutor {
private static final long serialVersionUID = 667815067287186086L;
#Override
public <T> Future<T> submit(Callable<T> task) {
return super.submit(new ContextAwareCallable<T>(task, newThreadContextContainer()));
}
#Override
public <T> ListenableFuture<T> submitListenable(Callable<T> task) {
return super.submitListenable(new ContextAwareCallable<T>(task, newThreadContextContainer()));
}
/**
* set infos what we need
*/
private ThreadContextContainer newThreadContextContainer() {
ThreadContextContainer container = new ThreadContextContainer();
container.setRequestAttributes(RequestContextHolder.currentRequestAttributes());
container.setContextMapOfMDC(MDC.getCopyOfContextMap());
return container;
}
}
The ThreadContextContainer is just a pojo to store infos for convenience:
public class ThreadContextContainer implements Serializable {
private static final long serialVersionUID = -6809291915300091330L;
private RequestAttributes requestAttributes;
private Map<String, String> contextMapOfMDC;
public RequestAttributes getRequestAttributes() {
return requestAttributes;
}
public Map<String, String> getContextMapOfMDC() {
return contextMapOfMDC;
}
public void setRequestAttributes(RequestAttributes requestAttributes) {
this.requestAttributes = requestAttributes;
}
public void setContextMapOfMDC(Map<String, String> contextMapOfMDC) {
this.contextMapOfMDC = contextMapOfMDC;
}
}
The ContextAwareCallable(a Callable proxy for original task) overwriting the call method to storage MDC or other context infos before the original task executing its call method:
public class ContextAwareCallable<T> implements Callable<T> {
/**
* the original task
*/
private Callable<T> task;
/**
* for storing infos what we need
*/
private ThreadContextContainer threadContextContainer;
public ContextAwareCallable(Callable<T> task, ThreadContextContainer threadContextContainer) {
this.task = task;
this.threadContextContainer = threadContextContainer;
}
#Override
public T call() throws Exception {
// set infos
if (threadContextContainer != null) {
RequestAttributes requestAttributes = threadContextContainer.getRequestAttributes();
if (requestAttributes != null) {
RequestContextHolder.setRequestAttributes(requestAttributes);
}
Map<String, String> contextMapOfMDC = threadContextContainer.getContextMapOfMDC();
if (contextMapOfMDC != null) {
MDC.setContextMap(contextMapOfMDC);
}
}
try {
// execute the original task
return task.call();
} finally {
// clear infos after task completed
RequestContextHolder.resetRequestAttributes();
try {
MDC.clear();
} finally {
}
}
}
}
In the end, using the #Async with the configured bean "demoExecutor" like this: #Async("demoExecutor")
void yourTaskMethod();
2.In regard to your question of handling the response:
Regret to tell that I don't really have a verified solution. Maybe the org.springframework.aop.interceptor.AsyncExecutionInterceptor#invoke is possible to solve that.
And I do not think it has a solution to handle the response with your ServletLoggingFilter. Because the Async method will be returned instantly. The afterRequest method executes immediately and returns before Async method doing things. You won't get what you want unless you synchronously wait for the Async method to finish executing.
But if you just want to log something, you can add those codes into my example ContextAwareCallable after the original task executing its call method:
try {
// execute the original task
return task.call();
} finally {
String something = MDC.get("doSomething"); // will not be null
// logthis(something);
// clear infos after task completed
RequestContextHolder.resetRequestAttributes();
try {
MDC.clear();
} finally {
}
}

springboot websocket with JPA NullPointerException

I'm wreting a SpringBoot web service use websocket to send message to client and using JPA to save data.
in project I use a thread to create a data ,and then use JPA to save it ,use Websocket to show data. But it doesn't work with a NullPointExecption.
this is my code:
#RestController
#ServerEndpoint("/websocket")
public class CrawlingController {
private final Logger logger = LoggerFactory.getLogger(CrawlingController.class);
#Autowired
private SimulateDataRepository simulateDataRepository;
private static GlobalStatus globalStatus = GlobalStatus.getInstance();
private Session session;
private static CopyOnWriteArraySet<CrawlingController> crawlingControllers =
new CopyOnWriteArraySet<>();
#OnOpen
public void onOpen(Session session) throws Exception {
this.session = session;
crawlingControllers.add(this);
String message = dataCrawling();
}
#OnMessage
public void onMessage(String message, Session session) throws IOException {
for (CrawlingController controller : crawlingControllers)
sendMessage(message);
}
#OnClose
public void onClose() {
crawlingControllers.remove(this);
if (crawlingControllers.size() == 0)
end();
session=null;
end();
}
private void sendMessage(String message) throws IOException {
this.session.getBasicRemote().sendText(message);
}
public void onError(Session session, Throwable error) {
logger.error(error.getMessage());
System.out.println("websocket found something error!");
}
#RequestMapping("/start")
public String dataCrawling() throws Exception {
if (!globalStatus.isDataCrawling) {
Process process = null;
synchronized (globalStatus) {
process = Utils.crawling();
globalStatus.crawlingProcss = process;
globalStatus.isDataCrawling = true;
}
BufferedReader bufferedReader = new BufferedReader(new InputStreamReader(process.getInputStream()));
new Thread(() -> {
try {
stuffSimulate(bufferedReader);
} catch (IOException | ParseException e) {
logger.error(e.getMessage());
}
}).start();
return "start Finish!";
} else return "faild";
}
public void stuffSimulate(BufferedReader reader) throws IOException, ParseException {
line = reader.readLine();
while (globalStatus.isDataCrawling) {
SimulateData simulateData = new SimulateData();
.... ...
simulateData.setData(data.toString());
this.saveAndShow(simulateData);
}
}
}
}
private void saveAndShow(SimulateData simulateData) throws IOException {
simulateData = simulateDataRepository.saveAndFlush(simulateData);
SummaryData summaryData = SimulateService.obtainSummary(simulateData);
onMessage(summaryData.printSummary(),this.session);
summaryData.sout();
}
}
simulateDataRepository can be work when i use http://localhost:8080/start.
the mysql can get data.
But it cna't be work when i use websocket!!!
The error meaasge said that:
Exception in thread "Thread-5" java.lang.NullPointerException
at cn.linhos.controller.CrawlingController.saveAndShow(CrawlingController.java:293)
at cn.linhos.controller.CrawlingController.stuffSimulate(CrawlingController.java:172)
at cn.linhos.controller.CrawlingController.lambda$dataCrawling$0(CrawlingController.java:104)
at java.lang.Thread.run(Thread.java:748)
why simulateDataRepository can't saveAndFlush when websocket but can do it when i use "/start"?
how should i do to make websocket work ?

#Context WebConfig not injected when using JerseyTest 2.0

I have a simple resource like:
#Path("/")
public class RootResource {
#Context WebConfig wc;
#PostConstruct
public void init() {
assertNotNull(wc);
}
#GET
public void String method() {
return "Hello\n";
}
}
Which I am trying to use with JerseyTest (2.x, not 1.x) and the GrizzlyTestContainerFactory.
I can't work out what I need to do in terms of config to get the WebConfig object injected.
I solved this issue by creating a subclass of GrizzlyTestContainerFactory and explicitly loading the Jersey servlet. This triggers the injection of the WebConfig object. The code looks like this:
public class ExtendedGrizzlyTestContainerFactory implements TestContainerFactory {
private static class GrizzlyTestContainer implements TestContainer {
private final URI uri;
private final ApplicationHandler appHandler;
private HttpServer server;
private static final Logger LOGGER = Logger.getLogger(GrizzlyTestContainer.class.getName());
private GrizzlyTestContainer(URI uri, ApplicationHandler appHandler) {
this.appHandler = appHandler;
this.uri = uri;
}
#Override
public ClientConfig getClientConfig() {
return null;
}
#Override
public URI getBaseUri() {
return uri;
}
#Override
public void start() {
if (LOGGER.isLoggable(Level.INFO)) {
LOGGER.log(Level.INFO, "Starting GrizzlyTestContainer...");
}
try {
this.server = GrizzlyHttpServerFactory.createHttpServer(uri, appHandler);
// Initialize and register Jersey Servlet
WebappContext context = new WebappContext("WebappContext", "");
ServletRegistration registration = context.addServlet("ServletContainer", ServletContainer.class);
registration.setInitParameter("javax.ws.rs.Application",
appHandler.getConfiguration().getApplication().getClass().getName());
// Add an init parameter - this could be loaded from a parameter in the constructor
registration.setInitParameter("myparam", "myvalue");
registration.addMapping("/*");
context.deploy(server);
} catch (ProcessingException e) {
throw new TestContainerException(e);
}
}
#Override
public void stop() {
if (LOGGER.isLoggable(Level.INFO)) {
LOGGER.log(Level.INFO, "Stopping GrizzlyTestContainer...");
}
this.server.stop();
}
}
#Override
public TestContainer create(URI baseUri, ApplicationHandler application) throws IllegalArgumentException {
return new GrizzlyTestContainer(baseUri, application);
}
Notice that the Jersey Servlet configuration is being loaded from the ApplicationHandler that is passed in as a parameter using the inner Application object's class name (ResourceConfig is a subclass of Application). Therefore, you also need to create a subclass of ResourceConfig for this approach to work. The code for this is very simple:
package com.example;
import org.glassfish.jersey.server.ResourceConfig;
public class MyResourceConfig extends ResourceConfig {
public MyResourceConfig() {
super(MyResource.class);
}
}
This assumes the resource you are testing is MyResource. You also need to override a couple of methods in your test like this:
public class MyResourceTest extends JerseyTest {
public MyResourceTest() throws TestContainerException {
}
#Override
protected Application configure() {
return new MyResourceConfig();
}
#Override
protected TestContainerFactory getTestContainerFactory() throws TestContainerException {
return new ExtendedGrizzlyTestContainerFactory();
}
#Test
public void testCreateSimpleBean() {
final String beanList = target("test").request().get(String.class);
Assert.assertNotNull(beanList);
}
}
Finally, for completeness, here is the code for MyResource:
#Path("test")
public class MyResource {
#Context WebConfig wc;
#PostConstruct
public void init() {
System.out.println("WebConfig: " + wc);
String url = wc.getInitParameter("myparam");
System.out.println("myparam = "+url);
}
#GET
#Produces(MediaType.APPLICATION_JSON)
public Collection<TestBean> createSimpleBean() {
Collection<TestBean> res = new ArrayList<TestBean>();
res.add(new TestBean("a", 1, 1L));
res.add(new TestBean("b", 2, 2L));
return res;
}
#POST
#Produces(MediaType.APPLICATION_JSON)
#Consumes(MediaType.APPLICATION_JSON)
public TestBean roundTrip(TestBean s) {
return s;
}
}
The output of running the test shows that the WebConfig is loaded and the init param is now available:
WebConfig: org.glassfish.jersey.servlet.WebServletConfig#107d0f44
myparam = myvalue
The solution from #ametke worked well but wasn't picking up my ExceptionMapper classes. To solve this I simplified the start() method to:
#Override
public void start() {
try {
initParams.put("jersey.config.server.provider.packages", "my.resources;my.config");
this.server = GrizzlyWebContainerFactory.create(uri, initParams);
} catch (ProcessingException | IOException e) {
throw new TestContainerException(e);
}
}
This was based on Problems running JerseyTest when dealing with HttpServletResponse

Resources