Abstract: A physical extent manager (PEM) receives a first message indicating a first PE of a first redundant array of independent disks (RAID) array of a disk has failed. The PEM determines a plurality of spare physical extents (PEs) available for reconstruction, the plurality of spare PEs including a first spare PE and a second spare PE. The PEM determines that the first spare PE has previously been assigned to a second PE of a second RAID array for reconstruction. The PEM determines that the first PE is able to use the first spare PE for reconstruction as the first spare PE meets the array row rule for the first PE and that the second PE is able to use the second spare PE for reconstruction as the second spare PE meets the array row rule for the second PE.

Abstract: A physical extent manager (PEM) receives a first message indicating a first PE of a first redundant array of independent disks (RAID) array of a disk has failed. The PEM determines a plurality of spare physical extents (PEs) available for reconstruction, the plurality of spare PEs including a first spare PE and a second spare PE. The PEM determines that the first spare PE has previously been assigned to a second PE of a second RAID array for reconstruction. The PEM determines that the first PE is able to use the first spare PE for reconstruction as the first spare PE meets the array row rule for the first PE and that the second PE is able to use the second spare PE for reconstruction as the second spare PE meets the array row rule for the second PE.

Abstract: A redundant array of independent disks (raid) array can include one or more virtual disks representing a raid configuration of the raid array. Each of the one or more virtual disks includes a number of physical extents (PEs), where a first PE and a second PE of the PEs have a same size. The first PE is located at a first storage disk of a first storage node of a cluster and the second PE is located at a second storage disk of a second storage node of the cluster.

Abstract: A storage node can include one or more processors and one or more storage disks, where the one or more storage disks include one or more local physical extents (PEs) that are local to the storage node. The storage node can include a protection pool driver executed by the one or more processors to run in a kernel space of the storage node, where the protection pool driver includes a local disk manager (LDM) and an array group module (AGRP). The LDM can be configured to manage the one or more local PEs at the one or more storage disks. The AGRP can include a number of storage arrays, where each of the storage arrays includes one or more virtual disks, where each of the one or more virtual disks is associated to at least a local PE or an external PE external to the storage node.

Abstract: A first node of a plurality of nodes is determined based on a node UID that is included in a physical extent (PE) universally unique identifier (UUID). Each of the plurality of nodes includes one or more groups of arrays. A local disk manager (LDM) that is associated with the first node is identified based on an LDM UID that is included in the PE UUID. The LDM manages a physical disk that includes a plurality of physical extents. A disk object is obtained based on a disk UID that is included in the PE UUID. The disk object is used to access a physical extent (PE) of the physical disk, using a PE offset that is included in the PE UUID. Other embodiments are described and claimed.

Abstract: A method, apparatus, and system for processing Redundant Array of Independent Disks (RAID) Input/Output (I/O) requests for a plurality of nodes in a cluster is disclosed. A file system request including a byte offset is received. Then, a Physical Extent (PE) row that matches the file system request and a RAID stripe within the identified PE row based on the byte offset is identified. Next, a plurality of RAID I/O requests to be routed to a physical disk is generated. Each of the plurality of the RAID I/O requests includes information associated with the PE and a type of operation. Thereafter, each of the RAID I/O requests is processed based on the information associated with the PE and the type of operation.

Abstract: A method of reassembling a local disk manager (LDM) and array group (AGRP) includes starting a physical extent manager (PEM) configured to run on a number of nodes. The PEM on each node is configured to manage an AGRP running on the same node. A number of LDMs are reassembled, and each LDM is configured to manage virtual disks on each of the nodes. Once enough LDMs are reassembled, an AGRP can be reassembled.

Abstract: A method of managing extents of a file system having a protection pool includes collecting and initializing physical extent manager (PEM) metadata, using a PEM daemon thread. The PEM is configured to run on each of a number of nodes. The method also includes creating a request queue, using the PEM daemon thread, for all requests submitted to the PEM. The method also includes scanning the request queue, using a PEM worker thread, to handle incoming requests submitted to the PEM. The method also includes listening for multicast messages, using a PEM multicast listener thread, to be handled by the PEM worker thread.

Abstract: A first node of a plurality of nodes is determined based on a node UID that is included in a physical extent (PE) universally unique identifier (UUID). Each of the plurality of nodes includes one or more groups of arrays. A local disk manager (LDM) that is associated with the first node is identified based on an LDM UID that is included in the PE UUID. The LDM manages a physical disk that includes a plurality of physical extents. A disk object is obtained based on a disk UID that is included in the PE UUID. The disk object is used to access a physical extent (PE) of the physical disk, using a PE offset that is included in the PE UUID. Other embodiments are described and claimed.

Abstract: A method, apparatus, and system for processing Redundant Array of Independent Disks (RAID) Input/Output (I/O) requests for a plurality of nodes in a cluster is disclosed. A file system request including a byte offset is received. Then, a Physical Extent (PE) row that matches the file system request and a RAID stripe within the identified PE row based on the byte offset is identified. Next, a plurality of RAID I/O requests to be routed to a physical disk is generated. Each of the plurality of the RAID I/O requests includes information associated with the PE and a type of operation. Thereafter, each of the RAID I/O requests is processed based on the information associated with the PE and the type of operation.

Abstract: A storage node can include one or more processors and one or more storage disks, where the one or more storage disks include one or more local physical extents (PEs) that are local to the storage node. The storage node can include a protection pool driver executed by the one or more processors to run in a kernel space of the storage node, where the protection pool driver includes a local disk manager (LDM) and an array group module (AGRP). The LDM can be configured to manage the one or more local PEs at the one or more storage disks. The AGRP can include a number of storage arrays, where each of the storage arrays includes one or more virtual disks, where each of the one or more virtual disks is associated to at least a local PE or an external PE external to the storage node.

Abstract: A method of reassembling a local disk manager (LDM) and array group (AGRP) includes starting a physical extent manager (PEM) configured to run on a number of nodes. The PEM on each node is configured to manage an AGRP running on the same node. A number of LDMs are reassembled, and each LDM is configured to manage virtual disks on each of the nodes. Once enough LDMs are reassembled, an AGRP can be reassembled.

Abstract: A method of managing extents of a file system having a protection pool includes collecting and initializing physical extent manager (PEM) metadata, using a PEM daemon thread. The PEM is configured to run on each of a number of nodes. The method also includes creating a request queue, using the PEM daemon thread, for all requests submitted to the PEM. The method also includes scanning the request queue, using a PEM worker thread, to handle incoming requests submitted to the PEM. The method also includes listening for multicast messages, using a PEM multicast listener thread, to be handled by the PEM worker thread.

Abstract: A redundant array of independent disks (raid) array can include one or more virtual disks representing a raid configuration of the raid array. Each of the one or more virtual disks includes a number of physical extents (PEs), where a first PE and a second PE of the PEs have a same size. The first PE is located at a first storage disk of a first storage node of a cluster and the second PE is located at a second storage disk of a second storage node of the cluster.