Data access by virtual processors in a distributed system

A non-transitory machine-readable storage medium comprising instructions that upon execution cause a distributed system to:

• execute a virtual machine (VM) across a plurality of computer nodes that are connected by respective physical input/output (I/O) interfaces to a network-attached storage system;
• execute instances of a distributed virtual management program in respective computer nodes of the plurality of computer nodes;
• provide a virtual I/O interface between the distributed virtual management program and a guest operating system (OS) of the VM, the virtual I/O interface comprising queues useable by a plurality of virtual processors in the VM to request parallel access over the respective physical I/O interfaces of data in the network-attached storage system;
• receive, at the distributed virtual management program, a request from a first virtual processor in a first computer node to use a first queue in the virtual I/O interface;
• determine, with the distributed virtual management program, that the first queue is at a second computer node;
• migrate, with the distributed virtual management program, the first queue from the second computer node to the first computer node to allow the first virtual processor to add information of the request to the first queue to allow an access of data from the network-attached storage system that satisfies the request locally using the physical I/O interface of the first computer node;
• and
• maintain, with the distributed virtual management program, coherency of the first queue across the first computer node and the second computer node.

A method of a distributed system, comprising:

• executing a virtual machine (VM) across a plurality of computer nodes that are connected by respective physical input/output (I/O) interfaces to a network-attached storage system;
• executing instances of a distributed hypervisor in respective computer nodes of the plurality of computer nodes;
• providing a virtual I/O interface between the distributed hypervisor and a guest operating system (OS) of the VM, the virtual I/O interface comprising queues useable by a plurality of virtual processors in the VM to request parallel access over the respective physical I/O interfaces of data in the network-attached storage system;
• receiving, at the distributed hypervisor, a request from a first virtual processor in a first computer node to use a first queue in the virtual I/O interface;
• determining, by the distributed hypervisor, that the first queue is at a second computer node;
• determining, using a machine learning process by the distributed hypervisor, that the first queue is to be migrated from the second computer node to the first computer node, wherein a decision of the machine learning process is based on information regarding use of the first queue at the second computer node and an affinity of the first virtual processor to each computer node of the plurality of computer nodes;
• migrating, by the distributed hypervisor, the first queue from the second computer node to the first computer node to allow the first virtual processor to add information of the request to the first queue to allow an access of data specified by the request from the network-attached storage system using the physical I/O interface of the first computer node, wherein the access of the data specified by the request is performed locally at the first computer node without involving any other computer node in accessing the data specified by the request;
• and
• maintaining, by the distributed hypervisor, coherency of the first queue across the first computer node and the second computer node.