A method, system, computer system, and computer-readable medium to maintain consistent, up-to-date copies of primary data that enable quick resumption of operations upon failure of the primary node, the secondary node, and/or the network connecting nodes. The current status of individual storage locations is tracked, indicating whether data in each individual storage location is synchronized between a primary node and a secondary node. Transaction logging by a transaction-logging application, such as a file system or database system, is used to track individual storage locations updated with each transaction. Acknowledgements of these updates are tracked during replication to determine whether individual storage locations containing primary and secondary data are synchronized. This coordinated transaction logging and acknowledgement tracking enables write operations and replication to be performed efficiently while maintaining the information necessary to resynchronize only individual storage locations when primary and secondary data become unsynchronized.