-
-
Notifications
You must be signed in to change notification settings - Fork 387
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Bug]: Loss of retained QoS 1 messages across cluster nodes #2251
Comments
@dlanzafame Thanks for your report. The retain store is eventually consistent; it has often been noted. But ultimately, the proper solution to this is to introduce consensus into the distribution of retained messages. This will lower performance of the retain store drastically, but make users more happy who observe wallclock time of events. |
Environment
Current Behavior
If the publisher and subscriber are connected to different cluster nodes, the last retained message is lost if the message is published less than 1 second before the subscription is completed by another client. When the publish and subscribe actions are over 1 second apart, 100% of messages are delivered. However, if the time interval is less than 1 second, the percentage of lost messages increases, reaching up to 70% when publish and subscribe actions are concurrent. If the publisher and subscriber are connected to the same node, no message loss occurs, even with concurrent publish and subscribe actions.
Steps to Reproduce:
Expected behaviour
A retained message published with QoS 1 should always be delivered to subscribers.
Configuration, logs, error output, etc.
Code of Conduct
The text was updated successfully, but these errors were encountered: