Optimizing write IO bandwidth and latency in an active-active clustered system based on a single storage node having ownership of a storage object

1. A method of executing a storage input/output (IO) request for a storage object that has affinity to a storage node, comprising:

receiving a storage IO request for a storage object at a respective storage node from among a first storage node and a second storage node in an active-active clustered system, the second storage node being communicably connected to the first storage node;

determining that the respective storage node is a current owner of the storage object, an ownership of the storage object providing an indication of an affinity of the storage object to the respective storage node;

having determined that the respective storage node is the current owner of the storage object, executing, by the respective storage node, the storage IO request for the storage object while avoiding performing internode communications with another storage node from among the first storage node and the second storage node in the active-active clustered system;

inspecting, by the respective storage node, the storage IO request to identify an entity associated with the storage object; and

setting, by the respective storage node, a local lock on the identified entity while avoiding performing internode communications with the other storage node in the active-active clustered system.

2. The method of claim 1 wherein the storage IO request is a write IO request, and wherein the method further comprises:

having set the local lock on the entity of the storage object, allocating, by the respective storage node, space for a journal from a directly accessible buffer; and

writing page reference or description information pertaining to the write IO request to the journal.

3. The method of claim 2 further comprising:

sending, by the respective storage node, an acknowledgement message to a host computer that issued the write IO request; and

releasing the local lock on the entity associated with the storage object.

4. The method of claim 1 wherein the storage IO request is a read IO request, and wherein the method further comprises:

having set the local lock on the entity of the storage object, reading, by the respective storage node, data corresponding to the entity associated with the storage object.

5. The method of claim 4 further comprising:

sending, by the respective storage node, an acknowledgement message to a host computer that issued the read IO request; and

releasing the local lock on the entity associated with the storage object.

6. The method of claim 1 further comprising:

receiving a second storage IO request for a second storage object at the respective storage node;

determining that the respective storage node is not the current owner of the second storage object; and

having determined that the respective storage node is not the current owner of the second storage object, engaging in internode communications with another storage node from among the first storage node and the second storage node to perform one of (i) obtaining a lock on an entity of the second storage object after transferring ownership of the second storage object from the other storage node to the respective storage node, and (ii) obtaining the lock on the entity of the second storage object without transferring the ownership of the second storage object.

7. The method of claim 6 wherein the respective storage node and the other storage node correspond to peer storage nodes in the active-active clustered system, wherein the respective storage node has allocated space for a first journal, wherein the other storage node has allocated space for a second journal, and wherein the method further comprises:

performing one or more of updating and synchronizing page reference or description information pertaining to outstanding writes to at least the second storage object stored in the first journal and the second journal, the page reference or description information including, for each of one or more storage IO operations performed on the second storage object, a transaction identifier (ID) that has (i) a first sequence ID identifying a first peer storage node in the active-active clustered system as the current owner of the second storage object that performed a respective storage IO operation on the second storage object, and (ii) a second sequence ID identifying a second peer storage node in the active-active clustered system that performed an immediately prior storage IO operation on the second storage object.

8. An active-active clustered system including at least two storage nodes communicably connected to one another, each respective storage node from among the at least two storage nodes comprising:

a memory; and

processing circuitry configured to execute program instructions out of the memory to:

receive a storage IO request for a storage object;

determine that the respective storage node is a current owner of the storage object, an ownership of the storage object providing an indication of an affinity of the storage object to the respective storage node;

having determined that the respective storage node is the current owner of the storage object, execute the storage IO request for the storage object while avoiding performing internode communications with another storage node from among the at least two storage nodes in the active-active clustered system;

inspect the storage IO request to identify an entity associated with the storage object; and

set a local lock on the identified entity while avoiding performing internode communications with the other storage node in the active-active clustered system.

9. The system of claim 8 wherein the storage IO request is a write IO request, and wherein the processing circuitry of the respective storage node is further configured to execute the program instructions out of the memory to:

having set the local lock on the entity of the storage object, allocate space for a journal from a directly accessible buffer; and

write page reference or description information pertaining to the write IO request to the journal.

10. The system of claim 9 wherein the processing circuitry of the respective storage node is further configured to execute the program instructions out of the memory to:

send an acknowledgement message to a host computer that issued the write IO request; and

release the local lock on the entity associated with the storage object.

11. The system of claim 8 wherein the storage IO request is a read IO request, and wherein the processing circuitry of the respective storage node is further configured to execute the program instructions out of the memory to:

having set the local lock on the entity of the storage object, read data corresponding to the entity associated with the storage object.

12. The system of claim 11 wherein the processing circuitry of the respective storage node is further configured to execute the program instructions out of the memory to:

send an acknowledgement message with the read data to a host computer that issued the read IO request; and

release the local lock on the entity associated with the storage object.

13. The system of claim 8 wherein the processing circuitry of the respective storage node is further configured to execute the program instructions out of the memory to:

receive a second storage IO request for a second storage object;

determine that the respective storage node is not the current owner of the second storage object; and

having determined that the respective storage node is not the current owner of the second storage object, engage in internode communications with the other storage node from among the at least two storage nodes to perform one of (i) obtaining a lock on an entity of the second storage object after transferring ownership of the second storage object from the other storage node to the respective storage node, and (ii) obtaining the lock on the entity of the second storage object without transferring the ownership of the second storage object.

14. The system of claim 13 wherein the respective storage node and the other storage node correspond to peer storage nodes in the active-active clustered system, wherein the respective storage node has allocated space for a first journal, wherein the other storage node has allocated space for a second journal, and wherein the processing circuitry of the respective storage node is further configured to execute the program instructions out of the memory to:

perform one or more of updating and synchronizing page reference or description information pertaining to outstanding writes to the second storage object stored in the first journal and the second journal, the PA reference or description information including, for each of one or more storage IO operations performed on the second storage object, a transaction identifier (ID) that has (i) a first sequence ID identifying a first peer storage node in the active-active clustered system as the current owner of the second storage object that performed a respective storage IO operation on the second storage object, and (ii) a second sequence ID identifying a second peer storage node in the active-active clustered system that performed an immediately prior storage IO operation on the second storage object.

15. A computer program product including a set of non-transitory, computer-readable media having instructions that, when executed by processing circuitry of a respective storage node from among a first storage node and a second storage node communicably connected to the first storage node in an active-active clustered system, cause the processing circuitry to perform a method comprising:

receiving a storage IO request for a storage object at the respective storage node;

determining that the respective storage node is a current owner of the storage object, an ownership of the storage object providing an indication of an affinity of the storage object to the respective storage node;

having determined that the respective storage node is the current owner of the storage object, executing, by the respective storage node, the storage IO request for the storage object while avoiding performing internode communications with another storage node from among the first storage node and the second storage node in the active-active clustered system;

inspecting, by the respective storage node, the storage IO request to identify an entity associated with the storage object; and

setting, by the respective storage node, a local lock on the identified entity while avoiding performing internode communications with the other storage node in the active-active clustered system.

16. The computer program product of claim 15 wherein the method further comprises:

having set the local lock on the entity of the storage object, allocating, by the respective storage node, space for a journal from a directly accessible buffer; and

writing page reference or description information pertaining to the storage IO request to the journal.

17. The computer program product of claim 16 wherein the method further comprises:

sending, by the respective storage node, an acknowledgement message to a host computer that issued the storage IO request; and

releasing the local lock on the entity associated with the storage object.