Consistent input/output (IO) recovery for active/active cluster replication

A system for consistent input/output (IO) recovery in an active/active replication process for clusters of a storage system, the system comprising:

• a memory comprising computer-executable instructions;
• and
• a processor executing the computer-executable instructions, the computer-executable instructions when executed by the processor cause the processor to perform operations comprising:
• designating one of the clusters as a lock winner, a lock winner designation indicating the one of the clusters takes priority, with respect to another cluster of the clusters, over a lock;
• and
• for each write IO request received from an application, performing:
• determining an extent of pages to be modified by the write IO request;
• acquiring the lock by one of the clusters determined to be local to the application issuing the write IO request, and locally persisting data subject to the write IO request;
• atomically creating a replication journal entry for IO recovery in a replication journal that is local to the one of the clusters that acquired the lock;
• sending, by the local cluster, a data write request to a peer cluster of the clusters;
• upon receiving acknowledgement from the peer cluster that the data write request was successfully transmitted, clearing the replication journal entry at the local cluster, unlocking the extent, and returning a notification to the peer cluster;
• and
• in response to a failure event occurring at the peer cluster after persisting the data at the local cluster and creating the replication journal entry at the local cluster, sending, to the peer cluster upon recovery of the peer cluster, the replication journal entry with respect to an extent corresponding to a time of the failure event.

A computer program product for consistent input/output (IO) recovery in an active/active replication process for clusters of a storage system, the computer program product embodied on a non-transitory computer readable medium and including instructions that, when executed by a computer causes the computer to perform operations comprising:

• designating one of the clusters as a lock winner, a lock winner designation indicating the one of the clusters takes priority, with respect to another cluster of the clusters, over a lock;
• and
• for each write IO request received from an application, performing:
• determining an extent of pages to be modified by the write IO request;
• acquiring the lock by the one of the clusters determined to be local to the application issuing the write IO request, and locally persisting data subject to the write IO request;
• atomically creating a replication journal entry for IO recovery in a replication journal that is local to the one of the clusters that acquired the lock;
• sending, by the local cluster, a data write request to a peer cluster of the clusters;
• upon receiving acknowledgement from the peer cluster that the data write request was successfully transmitted, clearing the replication journal entry at the local cluster, unlocking the extent, and returning a notification to the peer cluster;
• and
• in response to a failure event occurring at the peer cluster after persisting the data at the local cluster and creating the replication journal entry at the local cluster, sending, to the peer cluster upon recovery of the peer cluster, the replication journal entry with respect to an extent corresponding to a time of the failure event.