Multicast Change Notifications for NVMe-oF Discovery Controllers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In existing NVMe-oF systems, unconnected end-devices are negatively impacted by configuration changes, such as namespace masking updates, and require manual intervention by administrators to detect changes, consuming resources on both host and subsystem sides, especially in environments without a Centralized Discovery Controller (CDC).
Innovation Solution
A Multicast Change Notification (MCN) system is introduced, allowing the discovery controller to automatically notify impacted end-devices of configuration changes without requiring an explicit persistent connection, using MCNs to communicate changes to unconnected hosts and subsystems, enabling them to initiate connections and detect changes autonomously.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If explicit persistent connections are established between end-devices and discovery controllers, then Asynchronous Event Notifications can be delivered reliably, but resource consumption increases and manual intervention is required for connection management
Solution Approach 1:
The system performs preliminary actions by establishing implicit multicast group memberships before configuration changes occur. End-devices pre-register interest in configuration changes through multicast group membership, allowing the discovery controller to push notifications without requiring explicit persistent connections to be established at the time of change detection.
Solution Approach 2:
Multicast groups serve as an intermediary mechanism between end-devices and discovery controllers. Instead of requiring direct persistent connections, the multicast group acts as a mediator that enables the discovery controller to reach multiple end-devices simultaneously and deliver configuration change notifications reliably without establishing individual connections.
2Reliability
If manual intervention is used to connect hosts and DDCs after configuration changes, then connectivity is ensured, but resource consumption increases and administrative burden increases
Solution Approach 1:
The system enables self-service by allowing end-devices to automatically detect configuration changes through received multicast notifications and initiate reconnection procedures autonomously. This eliminates the need for manual administrator intervention to restore connectivity after configuration changes, reducing both resource consumption and administrative burden while maintaining connectivity reliability.
3Adaptability or versatility
If unconnected end-devices are present in the network, then network flexibility is maintained, but these devices cannot detect configuration changes automatically
Solution Approach 1:
The multicast notification mechanism provides universal functionality by enabling configuration change information delivery to both connected and unconnected end-devices through the same mechanism. The discovery controller sends multicast notifications that can be received by any end-device listening on the multicast group, regardless of whether an explicit persistent connection exists, ensuring no loss of configuration change information while maintaining network flexibility.
Data Source
AI summary
System and methods for using a discovery controller to establish network connections in a network that comprises Non-Volatile Memory Express™ over fabrics (NVMe-oF™) entities that support multicast change notifications comprise: in response to a change that affects at an NVMe-oF™ entity that has not yet established a connection with the discovery controller, generating a multicast change notification (MCN) that notifies the unconnected entity of the change and automatically communicating the MCN to the unconnected entity without requiring an explicit persistent connection to the discovery controller, e.g., to access storage ports in the network.