com.lightbend.lagom.scaladsl.pubsub
PubSubRef
Related Doc:
package pubsub
Request if this topic has any known subscribers at this point.
Request if this topic has any known subscribers at this point.
The Future is completed with the currently known information
at this node, i.e. completion is not deferred until there are subscribers.
Note that the registry of subscribers is eventually consistent, i.e. new subscribers are not immediately visible at other nodes, but typically the information will be fully replicated to all other nodes after a few seconds.
This method is especially useful when writing tests that require that a subscriber is known before sending messages to a topic.
Publish messages from a stream to the topic.
Publish messages from a stream to the topic.
You have to connect a Source that produces the messages to
this Sink and then run the stream.
Consume messages from the topic via a stream.
Consume messages from the topic via a stream.
You can return this Source as a response in a ServiceCall
and the elements will be streamed to the client.
Otherwise you have to connect a Sink that consumes the messages from
this Source and then run the stream.
A
PubSubRefrepresents a publish/subscribe topic.Messages can be published to the topic via a stream by using the
Sourcereturned by the #publisher method. A single message can be published with the #publish method.Messages can be consumed from the topic via a stream by using the
Sinkreturned by the #subscriber method.The registry of subscribers is eventually consistent, i.e. new subscribers are not immediately visible at other nodes, but typically the information will be fully replicated to all other nodes after a few seconds.
New subscribers will not see old messages that were published to the topic, i.e. it is only a live stream of messages.
Messages are not guaranteed to be delivered, i.e. they may be lost. That can for example happen if there is a transient network partition.
The subscriber has a buffer of received messages, but messages will be dropped if that buffer is full and demand for more elements have not been requested from downstream. This can happen if a subscriber is slower than the publishers of the messages. When the buffer is full the oldest messages are dropped to make room for new messages.