System and method for managing namespace of offloading card and processing input/output request

A method for managing a namespace of an offloading card, wherein the method runs in a system for managing namespace of an offloading card, and the system for managing the namespace of the offloading card comprises a host and the offloading card connected to the host, wherein the method comprises:

• sending, by the host, a namespace creation request to the offloading card, based on multiple applications issuing an input/output request running on the host;
• creating, by the offloading card, corresponding multiple namespaces for the multiple applications, according to the namespace creation request;
• and
• allocating, by the offloading card, multiple hardware queues corresponding to the created multiple namespaces according to the namespace creation request and binding the allocated multiple hardware queues respectively to the corresponding namespaces.

A method for processing an input/output request, wherein the method runs in a system for processing an input/output request, the system for processing the input/output request comprises a host, an offloading card, and a Non-Volatile Memory Host Controller Interface Specification (NVMe) device connected to the host via the offloading card,

• wherein the host has a virtual machine running thereon, the host comprises a central processing unit with multiple cores, the virtual machine comprises a user space and a kernel space, wherein the user space has multiple applications issuing an input/output request running therein, the applications are executed by corresponding cores, the kernel space comprises a general block layer and an NVMe driver, the general block layer comprises multiple software queues corresponding to the multiple cores, the virtual machine sends a namespace creation request to the offloading card via the NVMe driver,
• wherein the offloading card comprises an NVMe controller and a hardware accelerator, the NVMe controller creates respective namespaces for the multiple applications according to the namespace creation request, allocates multiple hardware queues corresponding to the created multiple namespaces from the hardware accelerator and binds the allocated multiple hardware queues respectively to the corresponding namespaces,
• wherein the NVMe driver is configured to manage the multiple hardware queues, the general block layer establishes a one-to-one correspondence relationship from the multiple software queues to the multiple hardware queues,
• wherein the method further comprises:
• issuing, by an application running in the user space, an input/output request to the general block layer;
• converting, by processing of the general block layer and the NVMe driver, the input/output request into an input/output request conforming with NVMe protocol, and sending the input/output request conforming with the NVMe protocol to the NVMe controller;
• transferring, by the NVMe controller, the input/output request conforming with the NVMe protocol to the offloading card via the hardware queues corresponding to the created namespaces;
• and
• sending, by the offloading card, the input/output request conforming with the NVMe protocol to the NVMe device.