Coordinator Ownership Authentication in Distributed Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In distributed storage systems across multiple local area network sites, authenticating and managing storage object coordinators leads to potential race conditions and split-brain scenarios due to dynamic election of primary and secondary coordinators, resulting in unintended connection interruptions.
Innovation Solution
Synchronizing metadata for primary and secondary coordinator ownership to prioritize secondary coordinators and mark coordinators as unhealthy upon disconnection, ensuring that only healthy coordinators establish connections and preventing simultaneous connection attempts by multiple coordinators.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple object owners or coordinators are deployed in the distributed storage system across multiple network sites, then performance in managing the object is improved, but difficulty in authenticating different coordinators with exclusive authorization increases
Solution Approach 1:
The system performs preliminary actions by establishing coordinator connections to component replicas before failures occur. Connection priorities are predetermined with the primary coordinator having first priority and secondary coordinators having subsequent priorities. This preliminary connection establishment and priority assignment prevents authentication conflicts when failures occur, as the system already has pre-established connections and known priority sequences ready to follow.
Solution Approach 2:
The patent introduces an intermediary mechanism in the form of a coordinated authentication system that mediates between multiple coordinators and component replicas. This intermediary system manages connection priorities and coordinates authentication sequences, preventing race conditions by ensuring that only the highest-priority available coordinator authenticates and connects to each replica at any given time.
2Reliability
If a new primary coordinator is elected dynamically once the previous primary coordinator is shut down, then system availability is improved, but probability of race condition increases
Solution Approach 1:
The system establishes connection priorities in advance before failures occur. The primary coordinator is designated with first priority and secondary coordinators are assigned subsequent priorities beforehand. When dynamic election occurs, this preliminary priority structure remains intact, allowing the newly elected coordinator to seamlessly assume its predetermined priority level without creating authentication conflicts or race conditions.
Solution Approach 2:
The system implements feedback mechanisms where coordinators continuously monitor the availability of other coordinators and adjust their authentication attempts accordingly. When a coordinator detects that a higher-priority coordinator is unavailable, it receives feedback to proceed with authentication. This feedback loop prevents race conditions by ensuring coordinators only attempt authentication when appropriate, based on real-time system state information.
3Adaptability or versatility
If both the new primary coordinator and old secondary coordinator attempt to connect to a replica simultaneously, then system adaptability is improved, but connection stability deteriorates
Solution Approach 1:
The patent introduces asymmetry in the authentication process by assigning different connection priorities to different coordinators. The primary coordinator has first priority while secondary coordinators have subsequent priorities. This asymmetric priority structure breaks the symmetry of simultaneous authentication attempts, ensuring that only one coordinator (the highest-priority available one) successfully connects to each replica at any given time, thus maintaining connection stability while allowing adaptability to coordinator changes.
Data Source
AI summary
Exemplary methods, apparatuses, and systems include a replica node storing a component of a storage object detecting that a primary coordinator for the storage object component is no longer available to serve as primary coordinator. The replica node is within a cluster of nodes storing components of the storage object. In response to detecting that the primary coordinator is no longer available, the replica node updates a first metadata entry indicating that a secondary coordinator for the storage object component is unhealthy. The replica node rejects connection requests from the secondary coordinator in response to the first metadata entry indicating that the secondary coordinator for the storage object component is unhealthy.