Understanding publishers
- Last UpdatedApr 24, 2025
- 1 minute read
Publishers connect directly to the PubSub Engine running on the same computer by using the Local Delivery Layer (LDL) for communications. It is normal for one publisher to publish to many different topics. Typically, these are organized under a parent topic that is associated with the publishing application.
Publishers are free to publish without worrying about the existence of any subscribers and to choose when to publish their data, even if the data has not changed. It is not a requirement for the data to be different since the last publication. However, since persistence stores are aware of whether current subscribers exist for a topic, they have the option to publish only when there are known subscribers.
Within a group of connected systems, there should be only one publisher per unique topic name. In the case of redundant publishers, only the Hot publisher should publish the data. The PubSub system cannot guarantee the updates are published in the correct order if the updates come from more than one publisher.
System and network usage should be considered when choosing a publication methodology. It is important to avoid flooding the network with insignificant updates. If necessary, the publisher may apply a deadband before deciding to publish or reduce the update rate by ignoring the intermediate values and publishing periodically.
If a publisher disconnects for any reason, the PubSub Engine publishes a "no more publishers" message for subscriptions to topics associated with the publisher. If the engine is aware of any peer subscriptions, it will also broadcast a "no more publishers" publication.