DISTRIBUTED COMPUTE ARRAY IN A STORAGE SYSTEM

    公开(公告)号:US20190250852A1

    公开(公告)日:2019-08-15

    申请号:US15898062

    申请日:2018-02-15

    Abstract: Implementations disclosed herein include a storage system including a plurality of storage devices, wherein each of the plurality of storage devices includes compute resources, memory resources, and a storage device controller configured to perform application-specific data management operations using the compute resources and the memory resources of the storage device, and a storage system controller configured to distribute a workload across the plurality of storage devices based on a capability of each of the plurality of storage devices during an IDLE state. The capability of each of the plurality of storage devices may be specifications, current availability, and performance history of each of the plurality of storage devices. In some implementations, each of the plurality of storage devices communicate with each other via a peer-to-peer networking communications protocol (e.g., NVme, NVMof, PCIe, Ethernet, etc.).

    Distributed compute array in a storage system

    公开(公告)号:US10802753B2

    公开(公告)日:2020-10-13

    申请号:US15898062

    申请日:2018-02-15

    Abstract: Implementations disclosed herein include a storage system including a plurality of storage devices, wherein each of the plurality of storage devices includes compute resources, memory resources, and a storage device controller configured to perform application-specific data management operations using the compute resources and the memory resources of the storage device, and a storage system controller configured to distribute a workload across the plurality of storage devices based on a capability of each of the plurality of storage devices during an IDLE state. The capability of each of the plurality of storage devices may be specifications, current availability, and performance history of each of the plurality of storage devices. In some implementations, each of the plurality of storage devices communicate with each other via a peer-to-peer networking communications protocol (e.g., NVme, NVMof, PCIe, Ethernet, etc.).

    Proactive load balancer for data storage system

    公开(公告)号:US11445012B2

    公开(公告)日:2022-09-13

    申请号:US17105162

    申请日:2020-11-25

    Abstract: Load balancing technology is used to distribute tasks requested of a data storage network across the connected storage nodes within the data storage network. Conventional load balancing technology is not proactive. Storage nodes within a data storage network may have different performance characteristics and/or current loads created by various demands on the storage nodes and connected servers. The proactive load balancers and corresponding load balancer agents disclosed herein are used to detect these variations in performance characteristics and/or current loads in the storage nodes and connected servers to inform decisions made by the proactive load balancer in distributing incoming access requests across the storage nodes and connected servers.

    Data integrity verification optimized at unit level

    公开(公告)号:US11537463B1

    公开(公告)日:2022-12-27

    申请号:US17378177

    申请日:2021-07-16

    Abstract: The technology disclosed herein pertains to a system for data integrity verification that is optimized at unit level. One or more implementations of such system include a method including identifying a first unit boundary in data block received from a client at an application layer, the unit boundary indicating end of a first unit of data, generating an intermediate checksum for the first unit of data at the application layer, receiving a second unit of data of the data block at the application layer, and generating a final checksum with an intermediate checksum as a basis and the second unit of data.

    Heuristic based authentication protocol for stateless protocol

    公开(公告)号:US11349840B1

    公开(公告)日:2022-05-31

    申请号:US17091754

    申请日:2020-11-06

    Abstract: Heuristic based approach to authentication of requests in a stateless protocol environment. Heuristics may be applied to a request to calculate a trust level for the request. The trust level for a request may at least in part be based on request parameters for the request and one or more previously received requests in a user context profile. Thus, historical request metadata may be used in calculating a trust value of a received request. If the trust value for a request exceeds a trust threshold, the request may be authenticated without forwarding the request to an authentication server. Thus, for requests in which the trust value exceeds the trust threshold, system performance may be improved by not requiring computational overhead and/or network bandwidth to be used in authentication. In this regard, the format of the request may still comport with the stateless protocol, but authentication may be made more efficient.

    HYBRID STORAGE DEVICE PARTITIONS WITH STORAGE TIERS

    公开(公告)号:US20190286355A1

    公开(公告)日:2019-09-19

    申请号:US15920264

    申请日:2018-03-13

    Abstract: A hybrid storage device includes a first storage medium configured to store data at a first speed and a second storage medium configured to store data at a second speed. The first storage medium may be a NAND flash storage medium, and the second storage medium may be disc storage medium. Partitions of the first storage medium are associated with partitions of the second storage medium to form at least two storage tiers. Each of the storage tiers may include different NAND partition capacities. The storage device further includes a peer to peer communication channel between the first storage medium and the second storage medium for moving data between a NAND partition and HDD partition. The storage device is accessible via a dual port SAS or PCIe interface.

Patent Agency Ranking