Abstract: A method of distributed file deletion, performed by a storage system, is provided. The method includes receiving, at the storage system, a request to delete a directory and contents of the directory and adding the directory to a first set, listed in a memory in the storage system. The method includes operating on the first set, by examining each directory in the first set to identify subdirectories, adding each identified subdirectory to the first set as a directory, and adding each examined directory to a second set listed in the memory. The method includes deleting in a distributed manner across the storage system without concern for order, contents of directories, and the directories, listed in the second set.

Abstract: Data transformation caching in an artificial intelligence infrastructure that includes one or more storage systems and one or more graphical processing unit (‘GPU’) servers, including: identifying, in dependence upon one or more machine learning models to be executed on the GPU servers, one or more transformations to apply to a dataset; generating, in dependence upon the one or more transformations, a transformed dataset; storing, within one or more of the storage systems, the transformed dataset; receiving a plurality of requests to transmit the transformed dataset to one or more of the GPU servers; and responsive to each request, transmitting, from the one or more storage systems to the one or more GPU servers without re-performing the one or more transformations on the dataset, the transformed dataset.

Abstract: Data transformation offloading in an artificial intelligence infrastructure that includes one or more storage systems and one or more graphical processing unit (‘GPU’) servers, including: storing, within the storage system, a dataset; identifying, in dependence upon one or more machine learning models to be executed on the GPU servers, one or more transformations to apply to the dataset; and generating, by the storage system in dependence upon the one or more transformations, a transformed dataset.

Abstract: An artificial intelligence and machine learning infrastructure system, including: one or more storage systems comprising, respectively, one or more storage devices; and one or more graphical processing units, wherein the graphical processing units are configured to communicate with the one or more storage systems over a communication fabric; where the one or more storage systems, the one or more graphical processing units, and the communication fabric are implemented within a single chassis.

Abstract: An artificial intelligence and machine learning infrastructure system, including: one or more storage systems comprising, respectively, one or more storage devices; and one or more graphical processing units, wherein the graphical processing units are configured to communicate with the one or more storage systems over a communication fabric; where the one or more storage systems, the one or more graphical processing units, and the communication fabric are implemented within a single chassis.

Abstract: A method of processing transactions associated with a command in a storage system is provided. The method includes receiving, at a first authority of the storage system, a command relating to user data. The method includes sending a transaction of the command, from the first authority to a second authority of the storage system, wherein a token accompanies the transaction and writing data in accordance with the transaction as permitted by the token into a partition that is allocated to the second authority in a storage device of the storage system.

Abstract: A plurality of storage nodes is provided. Each of the plurality of storage nodes includes nonvolatile solid-state memory for user data storage. The plurality of storage nodes is configured to distribute the user data and metadata associated with the user data throughout the plurality of storage nodes such that the plurality of storage nodes maintain the ability to read the user data, using erasure coding, despite a loss of two of the plurality of storage nodes. The plurality of storage nodes is configured to initiate an action based on the redundant copies of the metadata, responsive to achieving a level of redundancy for the redundant copies of the metadata. A method for accessing user data in a plurality of storage nodes having nonvolatile solid-state memory is also provided.

Abstract: A plurality of storage nodes is provided. Each of the plurality of storage nodes includes nonvolatile solid-state memory for user data storage. The plurality of storage nodes is configured to distribute the user data and metadata associated with the user data throughout the plurality of storage nodes such that the plurality of storage nodes maintain the ability to read the user data, using erasure coding, despite a loss of two of the plurality of storage nodes. The plurality of storage nodes is configured to initiate an action based on the redundant copies of the metadata, responsive to achieving a level of redundancy for the redundant copies of the metadata. A method for accessing user data in a plurality of storage nodes having nonvolatile solid-state memory is also provided.

Abstract: Systems and methods of deduplication aware scalable content placement am described. A method may include receiving data to be stored on one or more nodes of a storage array and calculating a plurality of hashes corresponding to the data. The method further includes determining a first subset of the plurality of hashes, determining a second subset of the plurality of hashes of the first subset, and generating a node candidate placement list. The method may further include sending the first subset to one or more nodes represented on the node candidate placement list and receiving, from the nodes represented on the node candidate placement list, characteristics corresponding to the nodes represented on the candidate placement list. The method may further include identifying one of the one or more nodes represented on the candidate placement list m view of the characteristic and sending the data to the identified node.

Abstract: A method for proactively rebuilding user data in a plurality of storage nodes of a storage cluster is provided. The method includes distributing user data and metadata throughout the plurality of storage nodes such that the plurality of storage nodes can read the user data, using erasure coding, despite loss of two of the storage nodes. The method includes determining that one of the storage nodes is unreachable and determining to rebuild the user data for the one of the storage nodes that is unreachable. The method includes reading the user data across a remainder of the plurality of storage nodes, using the erasure coding and writing the user data across the remainder of the plurality of storage nodes, using the erasure coding. A plurality of storage nodes within a single chassis that can proactively rebuild the user data stored within the storage nodes is also provided.

Abstract: Ensuring reproducibility in an artificial intelligence infrastructure that includes one or more storage systems and one or more graphical processing unit (‘GPU’) servers, including: identifying, by a unified management plane, one or more transformations applied to a dataset by the artificial intelligence infrastructure, wherein applying the one or more transformations to the dataset causes the artificial intelligence infrastructure to generate a transformed dataset; storing, within the one or more storage systems, information describing the dataset, the one or more transformations applied to the dataset, and the transformed dataset; identifying, by the unified management plane, one or more machine learning models executed by the artificial intelligence infrastructure using the transformed dataset as input; and storing, within the one or more storage systems, information describing one or more machine learning models executed using the transformed dataset as input.

Abstract: Systems and methods of deduplication aware scalable content placement are described. A method may include receiving data to be stored on one or more nodes of a storage array and calculating a plurality of hashes corresponding to the data. The method further includes determining a first subset of the plurality of hashes, determining a second subset of the plurality of hashes of the first subset, and generating a node candidate placement list. The method may further include sending the first subset to one or more nodes represented on the node candidate placement list and receiving, from the nodes represented on the node candidate placement list, characteristics corresponding to the nodes represented on the candidate placement list. The method may further include identifying one of the one or more nodes represented on the candidate placement list in view of the characteristic and sending the data to the identified node.

Abstract: A method for proactively rebuilding user data in a plurality of storage nodes of a storage cluster is provided. The method includes distributing user data and metadata throughout the plurality of storage nodes such that the plurality of storage nodes can read the user data, using erasure coding, despite loss of two of the storage nodes. The method includes determining that one of the storage nodes is unreachable and determining to rebuild the user data for the one of the storage nodes that is unreachable. The method includes reading the user data across a remainder of the plurality of storage nodes, using the erasure coding and writing the user data across the remainder of the plurality of storage nodes, using the erasure coding. A plurality of storage nodes within a single chassis that can proactively rebuild the user data stored within the storage nodes is also provided.

Abstract: Data transformation caching in an artificial intelligence infrastructure that includes one or more storage systems and one or more graphical processing unit (‘GPU’) servers, including: identifying, in dependence upon one or more machine learning models to be executed on the GPU servers, one or more transformations to apply to a dataset; generating, in dependence upon the one or more transformations, a transformed dataset; storing, within one or more of the storage systems, the transformed dataset; receiving a plurality of requests to transmit the transformed dataset to one or more of the GPU servers; and responsive to each request, transmitting, from the one or more storage systems to the one or more GPU servers without re-performing the one or more transformations on the dataset, the transformed dataset.

Abstract: Data transformation offloading in an artificial intelligence infrastructure that includes one or more storage systems and one or more graphical processing unit (‘GPU’) servers, including: storing, within the storage system, a dataset; identifying, in dependence upon one or more machine learning models to be executed on the GPU servers, one or more transformations to apply to the dataset; and generating, by the storage system in dependence upon the one or more transformations, a transformed dataset.

Abstract: Data transformation caching in an artificial intelligence infrastructure that includes one or more storage systems and one or more graphical processing unit (‘GPU’) servers, including: identifying, in dependence upon one or more machine learning models to be executed on the GPU servers, one or more transformations to apply to a dataset; generating, in dependence upon the one or more transformations, a transformed dataset; storing, within one or more of the storage systems, the transformed dataset; receiving a plurality of requests to transmit the transformed dataset to one or more of the GPU servers; and responsive to each request, transmitting, from the one or more storage systems to the one or more GPU servers without re-performing the one or more transformations on the dataset, the transformed dataset.

Abstract: Ensuring reproducibility in an artificial intelligence infrastructure that includes one or more storage systems and one or more graphical processing unit (‘GPU’) servers, including: identifying, by a unified management plane, one or more transformations applied to a dataset by the artificial intelligence infrastructure, wherein applying the one or more transformations to the dataset causes the artificial intelligence infrastructure to generate a transformed dataset; storing, within the one or more storage systems, information describing the dataset, the one or more transformations applied to the dataset, and the transformed dataset; identifying, by the unified management plane, one or more machine learning models executed by the artificial intelligence infrastructure using the transformed dataset as input; and storing, within the one or more storage systems, information describing one or more machine learning models executed using the transformed dataset as input.

Abstract: Data transformation offloading in an artificial intelligence infrastructure that includes one or more storage systems and one or more graphical processing unit (‘GPU’) servers, including: storing, within the storage system, a dataset; identifying, in dependence upon one or more machine learning models to be executed on the GPU servers, one or more transformations to apply to the dataset; and generating, by the storage system in dependence upon the one or more transformations, a transformed dataset.

Abstract: Dynamically allocated thread storage in a computing device is disclosed. The dynamically allocated thread storage is configured to work with a process including two or more threads. Each thread includes a statically allocated thread-local slot configured to store a table. Each table is configured to include a table slot corresponding with a dynamically allocated thread-local value. A dynamically allocated thread-local instance corresponds with the table slot.

Abstract: Dynamically allocated thread storage in a computing device is disclosed. The dynamically allocated thread storage is configured to work with a process including two or more threads. Each thread includes a statically allocated thread-local slot configured to store a table. Each table is configured to include a table slot corresponding with a dynamically allocated thread-local value. A dynamically allocated thread-local instance corresponds with the table slot.