40 results about "Replication (computing)" patented technology

A computing system performs real-time replication of database management system transactions. The computing system includes a source DBMS, a replication service and a data lakehouse. The source DBMS stores at least one source transaction table recording at least one source transaction and generates at least one recovery log indicating at least one modification in the at least one source transaction table. The replication service replicates the at least one source transaction recorded in the at least one source transaction table by generating at least one data file having a first data format. The data lakehouse stores at least one lakehouse table corresponding to the at least one source transaction table and having a second data format different from the first data format, and to modify the at least one lakehouse table based on the at least one data file.

In some examples, a first computing system receives a first metadata change made to metadata of a first instance of an object stored at a storage location of the first computing device. The first computing system associates, with the first metadata change, a time of the first metadata change. Based on receiving replication of a second instance of the object from a second computing system, the first computing system determines that a second metadata change was made to metadata of the second instance of the object. Further, the second instance of the object includes a time of the second metadata change. The first computing system at least one of: resolves a conflict between the first metadata change and the second metadata change based at least on comparing the times, or merges the first metadata change with the second metadata change into the metadata for the object.

Methods, systems, and devices for data management are described. A first computing node within a first set of computing nodes may receive, from a network gateway, a first replication request that is associated with a first replication job for the first cluster of computing nodes to provide copies of a first set of one or more computing snapshots to a second cluster of computing nodes. The first set of computing nodes may determine that a second computing node different from the first computing node will service the data replication request. A network bandwidth allocation may be increased for the first computing node that received the replication request directly from the network gateway while the network bandwidth allocation for the second computing node is maintained for the second computing node based on the second computing node receiving the request via an internal redirection of the request.

Verifying that data has been correctly replicated to a replication target, including: replicating a dataset stored at a first computing system to a second computing system; and determining, based at least on a comparison of a first hash and a second hash, validity of the dataset stored at the second computing system, wherein the first hash is generated by applying a hash function to a copy of the dataset that is stored at the first computing system and the second hash is generated by applying the hash function to a copy of the dataset that is stored at the second computing system.

A power state framework executes a workload in supported computing device types and configurations during a sequence of power state transitions to simulate user experiences. Workloads can be combined, e.g., default workload plus add-on workload(s). Device states are monitored, collected, reported, and analyzed to identify faults, which are prioritized for repair, e.g., based on the projected impact of the faults on users, e.g., if the users were to experience the faults. Computing device states include states of attached peripheral devices and wireless or wired connectivity before and after power state transitions. Device state indicates device stability throughout supported device power state transitions. Devices under test can be on debuggers during execution of workload and power state transitions. Automated, accelerated simulation of the execution of end user workloads and power state transitions in actual computing devices saves time and avoids replication errors compared to manual user-driven device testing.

The invention discloses a switching method and system of a redundant server, and solves the problems of detection delay and data inconsistency in traditional switching through cooperative work of hardware layer redundancy (redundant components, multi-path architecture and power supply protection), intelligent fault detection (multi-dimensional monitoring and second-level heartbeat) and dynamic load balancing and data consistency guarantee (synchronous replication and seamless switching technology). Experiments show that the scheme shortens the annual downtime of the system, improves the availability, is suitable for key business scenes such as financial transactions and cloud computing, and ensures the service continuity and the data reliability.

According to some embodiments, a digital core for a cloud computing environment is associated with an enterprise (e.g., the digital core may comprise enterprise resource planning software optimized for an in-memory database). Moreover, an integration inspector framework may establish a checkpoint between an application system (e.g., a source application system to provide information via the digital core and / or a target application system to receive information via the digital core) and the digital core. The integration inspector framework then automatically validates a communication arrangement associated with the application system and performs a data replication test, in connection with the application system and the digital core, using test data. A result of the data replication test is then evaluated and indicated (e.g., to a user via a user interface).

A power state framework executes a workload in supported computing device types and configurations during a sequence of power state transitions to simulate user experiences. Workloads can be combined, e.g., default workload plus add-on workload(s). Device states are monitored, collected, reported, and analyzed to identify faults, which are prioritized for repair, e.g., based on the projected impact of the faults on users, e.g., if the users were to experience the faults. Computing device states include states of attached peripheral devices and wireless or wired connectivity before and after power state transitions. Device state indicates device stability throughout supported device power state transitions. Devices under test can be on debuggers during execution of workload and power state transitions. Automated, accelerated simulation of the execution of end user workloads and power state transitions in actual computing devices saves time and avoids replication errors compared to manual user-driven device testing.

The application discloses a method for automatically discovering block devices and copying changed data in real time and a storage medium, belongs to the technical field of cloud computing, and aims to solve the technical problem of how to realize automatic discovery of block devices and real-time copying of data of newly added disks, reduce the invasiveness to user servers, and increase the continuity of user services. The technical scheme is as follows: the method is to load a block device data real-time copying kernel module into a linux kernel, register block device information, poll and scan or actively trigger by listening to block device events to query newly mounted / dismounted disks, enable / disable the real-time copying function of the corresponding disks, and simultaneously, in the case of not interrupting read-write I / O, real-time copy the disks in use.

Disclosure is made of methods, apparatus and system for migrating virtual machines (VMs) between source and destination in a computing environment and, more specifically, to replication based migration. VMs migration is controlled so as to manage transferal of data associated with one or more VMs from a source location to a destination location to meet certain user definable or system constraints. Dynamic control and adjustment of system parameters associated with the migration is also disclosed.

A disclosed computing device capable of achieving high data rates in link diagnostic testing by replicating traffic may include (1) a packet forwarding device configured to generate, as part of a link diagnostic test, traffic that falls below a certain threshold by a deficit and (2) a packet replication device configured to (A) receive at least a portion of the traffic from the packet forwarding device and (B) replicate the portion of traffic to generate supplemental traffic that compensates for the deficit, and (3) an interface configured to transmit the traffic and the supplemental traffic to a remote device via a link as part of the link diagnostic test. Various other apparatuses, systems, and methods are also disclosed.

The application belongs to the technical field of artificial intelligence, graph data processing and high-performance computing, and discloses a communication-efficient memory fusion method and system for super-memory graph neural network training and a medium. A task-specific memory management scheme, zero-copy CPU-GPU transmission, a layout-aware GPU intercommunication pipeline and an out-neighbor gradient aggregation mechanism are used to significantly reduce the overall data communication overhead while ensuring training accuracy. In the fusion memory-aware graph partitioning, a same subgraph is maintained by the storage layout of multiple GPUs. For P task partitioning characteristics and T task partitioning vertex dimensions, the number of graph partitions required by the system is reduced, thereby reducing neighbor replication, and the overall CPU-GPU communication overhead of the system is successfully reduced by 50% to 64%. The back propagation of the out-neighbor gradient aggregation of the application changes the back propagation into an out-neighbor aggregation process in the preprocessing stage.

A computing system performs real-time replication of database management system transactions. The computing system includes a source DBMS, a replication service and a data lakehouse. The source DBMS stores at least one source transaction table recording at least one source transaction and generates at least one recovery log indicating at least one modification in the at least one source transaction table. The replication service replicates the at least one source transaction recorded in the at least one source transaction table by generating at least one data file having a first data format. The data lakehouse stores at least one lakehouse table corresponding to the at least one source transaction table and having a second data format different from the first data format, and to modify the at least one lakehouse table based on the at least one data file.

Methods, systems, and devices for data management are described. A first computing node within a first set of computing nodes may receive, from a network gateway, a first replication request that is associated with a first replication job for the first cluster of computing nodes to provide copies of a first set of one or more computing snapshots to a second cluster of computing nodes. The first set of computing nodes may determine that a second computing node different from the first computing node will service the data replication request. A network bandwidth allocation may be increased for the first computing node that received the replication request directly from the network gateway while the network bandwidth allocation for the second computing node is maintained for the second computing node based on the second computing node receiving the request via an internal redirection of the request.

A high-throughput data transformation and upserting system for distributed computing environments, comprising: a data ingestion module configured to receive structured, semi-structured, and unstructured data from one or more distributed data sources, including databases, cloud platforms, enterprise systems, streaming infrastructures, and external computing devices; a transformation module configured to convert the received data into predefined processing formats through data mapping, normalization, filtering, aggregation, and schema matching operations performed across multiple distributed computing nodes; and an upsertion module configured to insert new records and update existing records in one or more distributed databases while maintaining synchronization and consistency of the records across distributed computing resources.a synchronization module configured to coordinate the management of distributed transactions, replication activities, and consistency operations between interconnected databases and compute nodes; a workload management module configured to distribute processing workloads across distributed compute resources to improve throughput efficiency, scalability, and resource utilization for data-intensive processing operations; and a monitoring module configured to detect duplicate records, synchronization conflicts, operational failures, processing delays, and data inconsistencies related to distributed transformation and upsertion activities.

The present application discloses an expert-parallel processing method and system, an electronic device, and a storage medium, which solve the problem of load imbalance on computing cards that carry expert execution, thereby improving the processing efficiency of the computing cards. In the expert-parallel processing system, W experts are deployed from a host node to N computing cards by means of expert replication, wherein each of the N computing cards comprises at least two experts having different expert weights. In an embodiment of the present application, on the basis of text processing units respectively activated by the W experts, M text processing units are assigned to at least two experts on each of the N computing cards, so as to obtain text processing units respectively activated by the at least two experts on each of the N computing cards. The total number of text processing units activated by the at least two experts and corresponding to a same expert weight across different computing cards is equal to the number of text processing units corresponding to the same expert weight among the W experts obtained by routing partitioning.

A computer-implemented method, according to one approach, includes using a determined classification of a predetermined file to determine a data compliance ruleset that applies to data of the predetermined file. In response to a determination that the data of the predetermined file is scheduled to be moved and / or replicated from a source node to a target node of an edge computing environment, the determined data compliance ruleset is applied to the data of the predetermined file. The method further includes preventing the scheduled movement and / or replication of the data of the predetermined file from occurring, in response to a determination that the scheduled movement and / or replication of the data of the predetermined file violates at least one rule of the determined data compliance ruleset.

Embodiments of the present invention involve composing synchronization structures that enable increased synchronization control. A virtual synchronization structure enables synchronization using structure traversal without replicating the full data set. This results in computing systems that are at a desired level of synchronization without full data replication. This method has the advantage of enabling limited sharing between systems and also speed.

A source computing system may transmit to a target computing system a replication request to replicate data. The replication request may comprise a checksum derived from a portion of the data to be replicated. The target system may use the checksum to obtain data from a local computing system corresponding to a matching checksum in a table corresponding to the local computing system. The target system may unicast a request to the local computing system based on a checksum associated with the local system in a master table matching the checksum received in the replication request. The target system may broadcast, to local computing systems, a request that comprises the checksum indicated in the replication request. The target system may construct the data to be replicated using data received from the local computing system(s), responsive to the unicast or broadcast, instead of replicating the data from the source computing system.

Disclosed is an improved approach to implement high availability for file servers in a virtualized computing environment. A high availability solution is provided that creates a global file system namespace across two clusters located at separate sites, where a synchronous storage replication is used to support the stretched container, allowing VMs and files stored in the container to be replicated in real-time between the two clusters.

The present application discloses an expert-parallel processing method and system, an electronic device, and a storage medium, which solve the problem of load imbalance on computing cards that carry expert execution, thereby improving the processing efficiency of the computing cards. In the expert-parallel processing system, W experts are deployed from a host node to N computing cards by means of expert replication, wherein each of the N computing cards comprises at least two experts having different expert weights. In an embodiment of the present application, on the basis of text processing units respectively activated by the W experts, M text processing units are assigned to at least two experts on each of the N computing cards, so as to obtain text processing units respectively activated by the at least two experts on each of the N computing cards. The total number of text processing units activated by the at least two experts and corresponding to a same expert weight across different computing cards is equal to the number of text processing units corresponding to the same expert weight among the W experts obtained by routing partitioning.

Methods, apparatuses, and systems for migrating a virtual machine (VM) from a source to a target in a computing environment, and more specifically, for migration-based replication are disclosed herein. To meet user-defined or system constraints, a VM is controlled to manage the transfer of data from a source location to a target location in relation to one or more VMs. Dynamic control and adjustment of system parameters related to migration are also disclosed herein.

The invention provides a DPU capacity expansion method and device, equipment, a storage medium and a program product, relates to the technical field of communication, solves the problem of insufficient computing resources of a single DPU, and improves the overall resource utilization rate of the DPU. The method is applied to a replication platform, and comprises the following steps: monitoring a first actual value of a performance index of a source DPU in a running process of a target application deployed on the source DPU; when the first actual value of the performance index does not meet the performance reference of the target application, determining a target capacity expansion DPU; the performance benchmark is used for reflecting performance demand parameters of the target application in a stable operation state; and configuring an application copy of the target application on the target capacity expansion DPU, and scheduling to-be-processed data of the target application to the target capacity expansion DPU, so that the application copy on the target capacity expansion DPU processes the to-be-processed data.

The invention provides a file storage method and device, electronic equipment and a storage medium, and relates to the technical field of Internet, in particular to the technical field of cloud computing, big data and the like. According to the scheme, a plurality of replication domains of the distributed storage cluster are obtained; wherein the replication domain is used for indicating a fault isolation unit, the replication domain comprises a plurality of logic data buckets, and the plurality of logic data buckets are distributed to different cache nodes of the distributed storage cluster; segmenting the to-be-stored file into a plurality of data blocks according to a set block size, and generating a plurality of copies of any data block; and respectively mapping the plurality of copies of any data block to different logic data buckets in the same copy domain, so that the plurality of copies are stored on a plurality of different cache nodes corresponding to the same copy domain.

Methods, apparatuses, and systems are disclosed herein for migrating a virtual machine (VM) from a source to a target in a computing environment, and more specifically, for migration-based replication. To meet user-defined or system constraints, a VM is controlled to manage the transfer of data from a source location to a target location in relation to one or more VMs. Dynamic control and adjustment of system parameters related to migration are also disclosed herein.

Provided are a computer program product, system, and method for a tape device to replicate data to a plurality of remote storage devices. A replication request is received from the computing device to replicate data to a replication number of at least one remote storage device of the remote storage devices over a network. The data to replicate is transmitted to the replication number of at least one remote storage device. At least one device acknowledgment from the at least one remote storage device indicating complete of writing the data to replicate is received. A replication acknowledgment is transmitted to the computing device indicating that the data in the replication request has been replicated in response to receiving the at least one device acknowledgment from the replication number of the at least one remote storage device.

Provided herein are systems and methods for maintaining integrity of data during clinical trials of digital therapeutic applications. A computing system can receive a data element generated based on interactions by a user with a digital therapeutic application during a clinical trial. The computing system can send the data element to each data store of a plurality of data stores. The computing system can access a first data store to retrieve a first instance of the data element. The computing system can identify the first instance of the data element as invalid. The computing system can access a second data store to retrieve a second instance of the data element, responsive to the identifying the first instance of the data element in the first data store as invalid. The computing system can store the second instance of the data element onto a data repository for the clinical trial.

Embodiments provide systems, methods and computer program products for cloud replication of data. One embodiment includes, accessing a virtual table definition and a data collection definition, the virtual table definition comprising a definition of a set of virtual table fields and a mapping of the set of virtual table fields to a set of target data types; automatically creating a virtual table according to the virtual table definition, the virtual table comprising the set of virtual table fields populated with the data of interest according to the data collection definition; and sending the virtual table and the mapping to a cloud computing environment. One embodiment further includes, in the cloud computing environment, storing the set of virtual table fields as a set of physical table fields in a physical table in a cloud hosted database, the set of physical table fields having the set of target data types.

One example method includes performing replication operations and / or failure recovery operations for replication operations in a computing system using vector clocks. Each node or process in a computing system is associated with a vector clock that reflects streams available at nodes. As the replication operation is performed, the vector clocks are updated. When a failure is detected or when insufficient streams are available, the nodes may be throttled until sufficient streams are available. The vector clocks facilitate a recovery line such that replication can resume when sufficient streams are available.