Method and System for Maintaining Storage Device Failure Tolerance in a Composable Infrastructure

A server, configured for use in a system having a composable infrastructure, where the system includes disk drives coupled to a communications subsystem, and the drives are organized into storage groups in accordance with at least one coupled failure mechanism that pertains to the drives, said server including:

• a server interface configured to couple the server to the communications subsystem;
• and
• at least one computing subsystem, coupled and configured to run at least one application, wherein the server has been allocated a subset of the drives in accordance with storage group membership of each drive in the subset, and the application allows the server to access the drives which have been allocated to said server but not any of the drives which has not been allocated to said server.

An administrator, configured for use in a system having a composable infrastructure, where the system includes disk drives and servers coupled to a communications subsystem, and the drives are allocable to the servers such that each of the servers can access each of the drives which has been allocated thereto but not any of the drives which has not been allocated thereto, said administrator including:

• a server interface configured to couple the administrator to the communications subsystem;
• and
• at least one computing subsystem, coupled and configured to run at least one administration application to allocate the drives to the servers, such that the drives are organized into storage groups in accordance with at least one coupled failure mechanism that pertains to the drives, and each of the servers is allocated a subset of the drives in accordance with storage group membership of each drive in each said subset of the drives.

A RAID controller, configured for use in a system having a composable infrastructure, where the system includes disk drives and servers coupled to a communications subsystem, the drives are organized into storage groups in accordance with at least one coupled failure mechanism that pertains to the drives, and the drives are allocable to the servers such that each of the servers can access each of the drives which has been allocated thereto but not any of the drives which has not been allocated thereto, said RAID controller including:

• a server interface configured to couple the RAID controller to the communications subsystem;
• and
• at least one computing subsystem, coupled and configured to implement a RAID function, including by determining a RAID set of the drives in accordance with storage group membership of each of the drives in the RAID set.

A method for storing data in a system having a composable infrastructure, wherein the system includes a communications subsystem, and disk drives and servers coupled to the communications subsystem, wherein the drives are organized into storage groups in accordance with at least one coupled failure mechanism that pertains to said drives, said method including:

• in accordance with an application running on one of the servers, performing a RAID function including by determining a RAID set of the drives in accordance with storage group membership of each of the drives in the RAID set, and storing data in redundant fashion on the drives of the RAID set.

A method for configuring a system having a composable infrastructure, wherein the system includes a communications subsystem, and disk drives and servers coupled to the communications subsystem, wherein the drives are organized into storage groups in accordance with at least one coupled failure mechanism that pertains to said drives, said method including:

• in accordance with an administration application running on one of the servers, allocating to each server of a set of at least some of the servers, a different subset of the drives in accordance with storage group membership of the drives.