Asynchronous IO Splitter for Continuous Data Protection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data protection systems face challenges such as system shutdown during backups, limited recovery points, and lengthy data recovery processes, which hinder continuous data protection and minimize downtime in disaster scenarios.
Innovation Solution
The implementation of an asynchronous splitting method for IOs, where a storage array receives and stores IOs, and later provides them according to a record, allowing for efficient replication and rollback to any point in time, using a five-state journaling process and data protection appliances to manage data replication and journaling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If synchronous data replication is used to ensure data consistency, then data reliability is improved, but system downtime increases during recovery operations
Solution Approach 1:
The system performs preliminary actions by continuously replicating data to a standby storage system before failures occur. The standby system maintains synchronized data in advance, so when a failure happens, the failed system can immediately failover to the standby without waiting for data replication, thus reducing recovery downtime while maintaining data consistency.
Solution Approach 2:
The patent introduces a standby storage system as an intermediary between the primary storage system and the final recovery target. This intermediary continuously receives and buffers replicated data, allowing the primary system to failover quickly while the standby system handles the actual data restoration, separating the consistency requirement from the recovery speed requirement.
2Adaptability or versatility
If continuous data protection with journaling is implemented to enable rollback to any point in time, then adaptability is improved, but device complexity increases
Solution Approach 1:
The journaling system is segmented into discrete journal entries, each representing a specific data change event. These segmented entries can be independently managed, selected, and applied during recovery operations. This segmentation allows the system to provide flexible rollback to any point in time by selecting specific journal entries while keeping the overall system structure manageable through modular organization of journal data.
3Reliability
If data is replicated to a remote backup storage system, then reliability is improved, but data recovery time increases
Solution Approach 1:
The standby storage system continuously and preliminarily receives replicated data from the primary system before failures occur. This preliminary data preparation means that when a failure happens, the recovery process doesn't need to wait for data to be replicated from remote backup - the data is already prepared and available at the standby system, significantly reducing recovery time while maintaining reliable data protection.
Data Source
AI summary
Example embodiments of the present invention provide asynchronous splitting of IOs from a storage array, such as a Symmetrix by EMC Corporation of Hopkinton, Mass. to a replication appliance, such as RecoverPoint by EMC Corporation of Hopkinton, Mass. The storage array may receive a plurality of IOs from a host and may store a record (e.g., in an ordered table) associated with the IOs. At a later time, the storage array may provide the IOs according to the record associated with the IOs. For example, the storage array may receive a plurality of Small Computer System Interface (SCSI) Command Descriptor Blocks (CDBs) from one or more hosts and receive data associated with the CDBs. The storage array then may write the data to respective locations in storage and create entries in the ordered table according to the order in which the CDBs were received.