International Business Machines Corporation

Paula Kim Ta-Shma

Michael Edward Factor

Danny Harnik

Dmitry Sotnikov

Elliot Karl Kolodner

Guy Laden

Shmuel Ben-Yehuda

Shachar Fienblit

David Hadas

Yosef Moatti

Yaron Weinsberg

Avishay Traeger

Moshik Hershcovitch

Guy Khazma

Aviad Zlotnick

Matthew Albert Huras

Joseph S Glider

Ilias Iliadis

Dan S Pelleg

Aamer U Sachedina

Vinodh Venkatesan

Dalit Naor

Dean Har'el Lorenz

Sivan Tal

Shlomit Sarah Pinter

Demyn Lee Plantenberg

Eran Rom

Lisa F Spainhower

Roee Shlomo

Vladimir Lipets

Tomer Solomon

Naama Parush-Shear-Yashuv

Gal Lushi

Guy Gerson-Golan Guy Gerson-Golan

Arik Rinberg

A hub of a computing environment obtains a training set from an edge of the computing environment. The training set that is obtained includes data from the edge and is used to train a neural network at the hub. The training of the neural network provides a detector and a generator at the hub. A determination is made as to whether the training of the neural network is complete. Based on determining that the training of the neural network is complete, the detector is sent to the edge. The detector at the edge is to facilitate suppression of additional edge data to the hub based on the detector at the edge determining that the additional edge data is statistically similar, based on one or more selected criteria, to data used to train the neural network.

Reduction of data transmission and data storage using neural network technology

An example system includes a processor to receive deltas corresponding to concurrently executed operations of a number of other replicas on an element of a delta-based conflict free replicated data type (CRDT) array of a distributed computing system. The processor is to modify, at the first replica, the element of the delta-based CRDT array based on the deltas using a dot store that maps dots to a dot store.

Delta-based conflict-free replicated data type arrays using dot stores that map dots to dot stores

Techniques for data skipping that may be user defined and extensible may be provided. A method may include receiving a query, at a computer system comprising a processor, memory accessible by the processor, and computer program instructions stored in the memory and executable by the processor, modifying, at the computer system, the received query to utilize at least one data skipping index to evaluate at least one criterion of the query, wherein the at least one data skipping index comprises an index on at least one attribute of data that excludes index information for at least a portion of those data items that do not meet the at least one criterion, and wherein the at least one data skipping index is generated based on information received from an application programming interface, and evaluating, at the computer system, the query.

Extensible data skipping

A computer program product and a system comprising: a cluster of Secure Execution Platforms (SEPs) having connectivity to a data storage, each SEP of said cluster is configured to maintain, using a key, confidentiality of data while processing thereof; the key is shared among the SEPs of said cluster, the key is automatically generated by the cluster or portion thereof and is unavailable to any non-cluster entity; the data storage retains encrypted data that is encrypted using the key; a first SEP of the cluster is configured to encrypt client data using the key to obtain encrypted client data and store the encrypted client data in the data storage; and a second SEP of the cluster is configured to retrieve encrypted stored data from the data storage, decrypt the encrypted stored data using the key to obtain non-encrypted form of the encrypted stored data.

Cluster of secure execution platforms

Improved computer data query processing by a) identifying a cut point value (CPV) for a data set, where an attribute value of each data set record is normalized between first and second normal bounds, where the CPV is any value between the normal bounds that is associated with a maximum data skipping performance value (DSP) calculated by applying a DSP function based on the normalized values and the CPV, and where the DSP function relates to a selected data query distribution profile, b) partitioning the data set into multiple partitions, one partition including a data record subset whose corresponding normalized attribute values are greater than the CPV, and another partition including a data record subset whose corresponding normalized attribute values are less than the CPV, and c) storing the partitions on physical data storage media for access by a computer configured to perform data query processing.

Data partitioning for improved computer performance when processing

A globally distributed scan list is determined. A determination is made whether the first data replica in the first plurality of data stored on a first device is in sync with a second data replica in the second plurality of data on a second device. In response to determining that the first data replica is not in sync with the second data replica, the first data replica is added to an unsynced queue. The neighbor data of the first plurality of data is added to a suspect queue. The priority to check the neighbor data is increased if the neighbor data is already in the suspect queue. Unsynced neighbor data is added to the unsynced queue. The priority for recovery of the data in the unsynced queue is determined. The priority is based on the vulnerability of the data. A data replica in the unsynced queue is recovered.

Fully distributed intelligent rebuild

Machines, systems and methods for enhancing data recovery in a data storage system, the method comprising determining whether one or more data storage mediums in a data storage system are unavailable; determining data that are at a risk of loss, due to said one or more data storage mediums being unavailable; from among the data that is determined to be at the risk of loss, identifying data that is highly vulnerable to loss; and creating one or more temporary replicas of the data that is highly vulnerable to loss.

Reliability enhancement in a distributed storage system

Machines, systems and methods for optimizing data replication in a distributed storage network, the method comprising determining a need to create a replica for a data item in a remote failure zone in a data storage network; creating a temporary replica of the data item in a local failure zone defined in the data storage network, in response to determining that it is beneficial to create the temporary replica in the local failure zone based on a cost versus reliability improvement analysis; attempting to create the replica in the remote failure zone; and removing the temporary replica from the local failure zone, in response to successfully creating the replica in the remote failure zone.

Enhancing reliability of a storage system by strategic replica placement and migration

A storage appliance system is disclosed which may include at least one application server for locally executing an application, and one or more storage servers in communication with the application server for I/O transmission therebetween. Also disclosed are an application server, a method, and a computer program product.

Storage appliance, application server and method thereof

Methods, computing systems and computer program products implement embodiments of the present invention that include detecting multiple sets of storage objects stored in a data facility including multiple server racks, each of the server racks including a plurality of server computers, each of the storage objects in each set being stored in a separate one of the server racks and including one or more data objects and one or more protection objects. A specified number of the storage objects are identified in a given server rack, each of the identified storage objects being stored in a separate one of the server computers, and one or more server computers in the given server rack not storing any of the identified storage objects are identified. Finally, in the identified one or more server computers, an additional protection object is created and managed for the identified storage objects.

Intra-rack and inter-rack erasure code distribution

A method for processing data using snapshots is provided. The method comprises generating a primary snapshot of data stored in a first storage medium, wherein the primary snapshot is a copy of the data stored in the first storage medium at a point in time; generating one or more secondary snapshots of the primary snapshot, wherein the secondary snapshots are copies of the data stored on the first storage medium at the point in time; and extracting the data stored in the first storage medium at the point in time by accessing the data included in the secondary snapshots, wherein multiple processing units concurrently transform the extracted data and load the transformed data onto a second storage medium.

Concurrent data processing using snapshot technology

Growing community of inventors

Tel Aviv, Israel

Paula Kim Ta-Shma