Distributed Database Read-Pool Selection Using Replica Timestamps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed database systems face challenges in providing both fast response times and reliable data consistency, particularly in asynchronous schemes where replica nodes may not always have up-to-date data.
Innovation Solution
A server is equipped with a communication circuit, memory, and a processor that collects timestamp values from original and replica storages to configure a read pool list, selecting reliable storages to respond to read requests based on timestamp differences and confidence time values, ensuring data consistency and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If an asynchronous database scheme is used to improve response speed, then the response time is reduced, but data consistency and reliability deteriorate
Solution Approach 1:
The system implements feedback by continuously monitoring the synchronization status of replica nodes through timestamp comparison. The server collects timestamps from original and replica storages, compares them to determine synchronization levels, and uses this feedback information to dynamically manage the read pool list, ensuring only synchronized replicas are selected for read requests.
Solution Approach 2:
The read pool list is dynamically managed based on real-time synchronization status. The server periodically collects timestamps, determines which replica storages are synchronized, and updates the read pool list accordingly. This dynamic adjustment allows the system to adapt to changing data consistency states while maintaining fast response times.
2Reliability
If all replica nodes are included in read pool to increase data availability, then read operation reliability improves, but response time increases due to data unsynchronization
Solution Approach 1:
The system applies local quality by differentiating the status of different replica nodes based on their synchronization level. Instead of treating all replicas equally, the server identifies which replica storages have synchronized data (within a predetermined time threshold) and includes only those in the read pool list. This selective inclusion ensures data reliability while avoiding unnecessary delays from accessing unsynchronized replicas.
3Reliability
If the server waits for all replica nodes to synchronize before providing data, then data consistency is ensured, but response time increases
Solution Approach 1:
The server applies partial action by not waiting for all replica nodes to synchronize completely, but rather by collecting synchronization status information and making selective decisions based on individual replica performance. The server determines a read pool list containing only those replicas that have achieved sufficient synchronization (within the time threshold), allowing data provision to proceed with partial synchronization rather than complete waiting.
Data Source
AI summary
A server is provided. The server includes a communication circuit, memory, including one or more storage media, storing instructions, and a processor communicatively coupled to the communication circuit and the memory, wherein the instructions, when executed by the processor, cause the server to collect a timestamp value indicating a point in time at which data writing is completed, determine a member to belong to a read pool by using the timestamp value in each storage, configure a read pool list so as to include the member determined through the determination operation and store the read pool list in the memory, receive a read request from a client apparatus, select a storage to provide data in response to the read request, receive data through the communication circuit from the storage selected from the read pool list, and transmit the received data to the client apparatus through the communication circuit.