Application-Consistent Data Copy Scheduling Under RPO Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies fail to ensure application-level consistency in data recovery point objectives (RPO) and recovery time objectives (RTO) in hybrid cloud environments, particularly due to inconsistent network connectivity and lack of consideration for application quiescing times, leading to potential data inconsistencies during copying.
Innovation Solution
A system managing apparatus and method that calculates application quiescing times, copy data amounts, and copy periods to ensure consistent data recovery by optimizing copy execution timing based on recovery point objectives, application quiescing time periods, and network conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If copy execution is performed at intervals designated by RPO without considering application quiescing time, then copy frequency meets RPO requirements, but application consistency cannot be guaranteed
Solution Approach 1:
The system performs preliminary quiescing of the application before executing the copy operation. By temporarily suspending the application in a consistent state prior to copying, the system ensures that the copy captures a point-in-time consistent state of the application, thereby guaranteeing application-level consistency while meeting RPO requirements.
2Reliability
If entire data from memory is written to non-transitory storage before copying, then application consistency is maintained, but copy time increases
Solution Approach 1:
The system performs preliminary flushing of memory data to non-transitory storage before the copy operation begins. This ensures that all in-memory data is persisted to stable storage, guaranteeing consistency without requiring the entire data set to be copied again, thereby reducing overall copy preparation time.
3Adaptability or versatility
If copy is performed over non-dedicated network lines in hybrid cloud, then cloud resource utilization increases, but copy completion time becomes unpredictable
Solution Approach 1:
The system continuously monitors the actual copy progress and compares it against the predicted completion time during the copy operation. If network conditions cause deviations from the predicted timeline, the system adjusts the copy execution timing dynamically to compensate, ensuring that the final copy completion aligns with the required RPO timing despite using non-dedicated network lines.
4Reliability
If copy execution timing is calculated without considering application quiescing time per unit data amount, then calculation simplicity is maintained, but RPO guarantee fails
Solution Approach 1:
The system introduces a new parameter - application quiescing time per unit data amount - into the copy timing calculation. By incorporating this parameter that reflects the specific characteristics of the application being copied, the system achieves accurate RPO guarantee while maintaining a relatively simple calculation framework that can be extended to different applications.
Data Source
AI summary
A system managing apparatus decides a time optimum for executing data copying from a copy source volume to a copy destination volume, and manages an RPO requirement of a user and a quiescing time period for each application at the time of a copy start. The system managing apparatus calculates a copy execution time and a copy execution interval such that data in which application consistency is maintained can be acquired at an interval that satisfies the RPO. In addition, the system managing apparatus monitors the use status of an application and the use status of a copy line after the start of copy operation, and calculates a copy execution time and a copy execution interval that allow a return to an appropriate recovery point, taking into consideration an application quiescing time period and a period of time required for copy completion that change in real time.