Single Node Message Broadcasting in Clustered Computing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In clustered computing environments, message broadcasting often results in multiple identical messages being delivered to and processed by multiple nodes, leading to unnecessary resource consumption and inefficiency.
Innovation Solution
A message is received in a message queue and sent to a single node within the environment, which processes the message, identifies relevant listeners based on the message topic, and broadcasts it to those listeners, tracking and routing delivery confirmations and responses, thus avoiding redundant processing across multiple nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a message is delivered to all nodes in the clustered computing environment, then all nodes can process the message and publish to listeners, but multiple identical messages are delivered to the same listener and resource consumption is unnecessarily high
Solution Approach 1:
The patent extracts the message processing function from all nodes and concentrates it in a single selected node. The message is delivered to only one node which then publishes to all listeners, eliminating redundant processing on other nodes while maintaining complete message delivery to all listeners.
Solution Approach 2:
The selected node acts as an intermediary between the message queue and the listeners. It receives the message from the queue, identifies the listeners, publishes the message to them, and tracks deliveries and responses, thereby mediating the message flow and preventing redundant deliveries.
2Productivity
If a message is delivered to all nodes in the clustered computing environment, then message processing can be distributed, but multiple substantially identical messages are delivered to the same listener
Solution Approach 1:
The patent extracts the message publishing function from multiple nodes and concentrates it in a single selected node. This node publishes the message to all listeners, ensuring each listener receives the message exactly once while maintaining overall message processing throughput.
Solution Approach 2:
Instead of having each node publish messages independently (which causes duplication), the patent inverts the approach by having a single selected node publish to all listeners. This reversal of the traditional publish/subscribe model eliminates message duplication while maintaining productivity.
3Productivity
If multiple nodes process the message independently, then processing can be parallelized, but unnecessary resource consumption occurs
Solution Approach 1:
The patent extracts the message processing function from all nodes and assigns it to a single selected node. This eliminates redundant processing energy consumption across multiple nodes while maintaining processing throughput through efficient single-node execution and listener broadcast.
Data Source
AI summary
Approaches are provided for message broadcasting within a clustered computing environment such as a Cloud computing environment. Specifically, a message is received in a message queue (e.g., on a message queue server). From the message queue, the message is sent to a single node of a plurality of nodes within the clustered computing environment. The single node can be selected based on factors such as a best available node within the plurality of nodes. This single node will process the message and identify a set of message listeners to whom the message should be broadcast (e.g., based on a topic of the message and an association of the topic to the set of message listeners). The single node will then publish the message to the set of message listeners. Delivery confirmations, responses, etc. associated with the message will then be tracked, processed, and/or routed by the single node.