following this stackoverflow post seems to have fixed it:
https://stackoverflow.com/questions/39080296/hazelcast-threads-prevent-tomee-from-stopping
creating a default producer:
@ApplicationScoped
public class NatsConnectionProducer {
@Resource(name = "baseAddressNats")
private String baseAddressNats;
@Produces
@ApplicationScoped
public StreamingConnection instance() throws IOException,
InterruptedException {
StreamingConnectionFactory cf = new
StreamingConnectionFactory(new Options.Builder().natsUrl(baseAddressNats)
.clusterId("cluster-id").clientId("client-id").build());
return cf.createConnection();
}
public void destroy(@Disposes final StreamingConnection instance)
throws IOException, TimeoutException, InterruptedException {
instance.close();
}
}
On 28/05/2021 11:35, Matthew Broadhead wrote:
i am trying to subscribe to a nats streaming server in a singleton.
the idea is that it will listen for any messages from the queue while
the server is running.
it looks like it might work ok in production but during development if
i make any code change then it blows up with
org.apache.catalina.loader.WebappClassLoaderBase.clearReferencesThreads
The web application [api] appears to have started a thread named
[jnats-streaming ack timeout thread] but has failed to stop it. This
is very likely to create a memory leak.
i have created a gist
https://gist.github.com/chongma/2a3ab451f2aeabc98340a9b897394cfe
is there any way to make sure the thread gets shut down when the web
application is stopped?
