Abstract: A direct-attached storage device software Redundant Array of Independent Disk (RAID) crash dump system includes a chassis housing a software RAID subsystem coupled to each of a plurality of physical storage devices, each of a plurality of controller devices, and an operating system. While in a runtime mode, the software RAID subsystem presents the operating system with a primary controller device from the plurality of controller devices as being connected to a logical storage device provided by the plurality of physical storage devices. In response to entering a crash dump mode from the runtime mode, the software RAID subsystem provides a storage driver subsystem for the primary controller device and uses it to initialize at least one secondary controller device in the plurality of controller devices that controls at least one of the plurality of physical storage devices that provides a boot logical storage subsystem in the logical storage device.

Abstract: A direct-attached storage device software RAID hibernation system includes a chassis having an operating system, a bus driver subsystem, controller devices coupled to physical storage devices, and a software RAID subsystem coupled to the operating system and the controller devices. While in a runtime mode, the software RAID subsystem presents the operating system a primary controller device as being connected to a logical storage device provided by the physical storage devices, and provides a filter subsystem in a secondary controller device that controls at least one of the physical storage devices. While in the runtime mode, the filter subsystem identifies a controller initialization request from the operating system that requests initialization of the secondary controller device and, in response, transmits a power-down prevention communication to the bus driver subsystem that is configured to prevent the bus driver subsystem from powering down the secondary controller device during a hibernation mode.

Abstract: A direct-attached storage device software Redundant Array of Independent Disk (RAID) crash dump system includes a chassis housing a software RAID subsystem coupled to each of a plurality of physical storage devices, each of a plurality of controller devices, and an operating system. While in a runtime mode, the software RAID subsystem presents the operating system with a primary controller device from the plurality of controller devices as being connected to a logical storage device provided by the plurality of physical storage devices. In response to entering a crash dump mode from the runtime mode, the software RAID subsystem provides a storage driver subsystem for the primary controller device and uses it to initialize at least one secondary controller device in the plurality of controller devices that controls at least one of the plurality of physical storage devices that provides a boot logical storage subsystem in the logical storage device.

Abstract: A direct-attached storage device software RAID hibernation system includes a chassis having an operating system, a bus driver subsystem, controller devices coupled to physical storage devices, and a software RAID subsystem coupled to the operating system and the controller devices. While in a runtime mode, the software RAID subsystem presents the operating system a primary controller device as being connected to a logical storage device provided by the physical storage devices, and provides a filter subsystem in a secondary controller device that controls at least one of the physical storage devices. While in the runtime mode, the filter subsystem identifies a controller initialization request from the operating system that requests initialization of the secondary controller device and, in response, transmits a power-down prevention communication to the bus driver subsystem that is configured to prevent the bus driver subsystem from powering down the secondary controller device during a hibernation mode.

Abstract: Methods, systems, and computer programs encoded on computer storage medium, for identifying a data storage environment including a plurality of disk drives; exposing for each of the disk drives of the data storage environment, a management interface of the disk drive to create a zero-capacity device for the disk drive; identifying a management command for a particular disk drive of the plurality of disk drives; identifying a particular zero-capacity device that corresponds to the particular disk drive; providing the management command to the particular zero-capacity device; and in response to providing the management command, receiving a response from the particular disk drive, the response including a processing result that is responsive to the management command.

Abstract: Methods, systems, and computer programs encoded on computer storage medium, for identifying a data storage environment including a plurality of disk drives; exposing for each of the disk drives of the data storage environment, a management interface of the disk drive to create a zero-capacity device for the disk drive; identifying a management command for a particular disk drive of the plurality of disk drives; identifying a particular zero-capacity device that corresponds to the particular disk drive; providing the management command to the particular zero-capacity device; and in response to providing the management command, receiving a response from the particular disk drive, the response including a processing result that is responsive to the management command.

Abstract: In accordance with embodiments of the present disclosure, an information handling system may include a processor and a plurality of physical storage resources communicatively coupled to the processor via a bus, the plurality of physical storage resources configured to, in order to facilitate booting from a virtual logical unit spanning the plurality of physical storage resources: (i) after enumeration of the plurality of physical storage resources, construct the virtual logical unit based on metadata of the plurality of physical storage resources; and (ii) prior to construction of the virtual logical unit, respond to input/output commands from an operating system of the information handling system with data associated with a dummy logical unit or with an indication that the dummy logical unit is busy.

Abstract: In accordance with embodiments of the present disclosure, an information handling system may include a processor and a plurality of physical storage resources communicatively coupled to the processor via a bus, the plurality of physical storage resources configured to, in order to facilitate booting from a virtual logical unit spanning the plurality of physical storage resources: (i) after enumeration of the plurality of physical storage resources, construct the virtual logical unit based on metadata of the plurality of physical storage resources; and (ii) prior to construction of the virtual logical unit, respond to input/output commands from an operating system of the information handling system with data associated with a dummy logical unit or with an indication that the dummy logical unit is busy.

Abstract: In accordance with embodiments of the present disclosure, a method may include receiving requirements for building a virtual storage resource from an array of physical storage resources, receiving performance metrics and power metrics of the physical storage resources of the array available for inclusion in the virtual storage resource, determining a plurality of unique combinations of the available physical storage resources that could be used to build the virtual storage resource, determining an effective performance, an effective performance penalty, a total power consumption, and an effective power penalty for each of the plurality of unique combinations, and selecting a single combination of the plurality of unique combinations for the virtual storage resource based on effective performances, effective performance penalties, total power consumptions, and effective power penalties of the plurality of unique combinations.

Abstract: An information handling system recognizes PCIe-based RAID volumes, including RAID virtual disks spanning two or more NVMe storage drives, while performing boot, hibernation, and crash-dump functions, rather than treating each storage device and its corresponding storage controller separately. The information handling system may perform a two-tiered discovery/initialization process during which a storage protocol function driver detects and initializes physical storage controllers, a storage controller bus driver virtualizes all physical storage controllers of a particular storage protocol and exposes a single virtualized storage controller for the protocol. A virtual miniport driver may then detect RAID volumes associated with the virtual storage controller and initialize the detected RAID stack(s).

Abstract: An information handling system recognizes PCIe-based RAID volumes, including RAID virtual disks spanning two or more NVMe storage drives, while performing boot, hibernation, and crash-dump functions, rather than treating each storage device and its corresponding storage controller separately. The information handling system may perform a two-tiered discovery/initialization process during which a storage protocol function driver detects and initializes physical storage controllers, a storage controller bus driver virtualizes all physical storage controllers of a particular storage protocol and exposes a single virtualized storage controller for the protocol. A virtual miniport driver may then detect RAID volumes associated with the virtual storage controller and initialize the detected RAID stack(s).

Abstract: In accordance with embodiments of the present disclosure, a method may include receiving requirements for building a virtual storage resource from an array of physical storage resources, receiving performance metrics and power metrics of the physical storage resources of the array available for inclusion in the virtual storage resource, determining a plurality of unique combinations of the available physical storage resources that could be used to build the virtual storage resource, determining an effective performance, an effective performance penalty, a total power consumption, and an effective power penalty for each of the plurality of unique combinations, and selecting a single combination of the plurality of unique combinations for the virtual storage resource based on effective performances, effective performance penalties, total power consumptions, and effective power penalties of the plurality of unique combinations.