Abstract: The present disclosure involves systems, software, and computer implemented methods for deploying changes to key patterns in database systems. One example method includes receiving a new shared container that includes a new key pattern for a logical split table and an updated read-only-portion physical table for a logical split table. The updated read-only-portion physical table includes records that match a complement of the new key pattern. Records that match the new key pattern are identified in the read-only-portion physical table of the logical split table in a current shared container. The identified records are moved from the read-only-portion physical table in the current shared container to a writable-portion physical table in a tenant container. Records not matching the new key pattern are deleted from the writable-portion physical table of the logical split table in the tenant container.

Abstract: Various examples are directed to systems and methods of executing a web application. A web browser application may receive web application code for executing the web application and service worker code for executing a service worker. The web browser may execute a first portion of the web application code to register the service worker and may execute the service worker. The web application may send a first communication request describing a first recipient computing device. The service worker may determine that the first communication request complies with set of filter rules and send the first communication request to the first recipient computing device.

Abstract: Some embodiments provide a non-transitory machine-readable medium that stores a program. The program transmits via a multicast communication protocol a message specifying a set of services offered by the device to a plurality of computing devices. The program further establishes a connection with a computing device in the plurality of computing devices. The program also receives, through the connection, a set of data for the set of services. The program further applies the set of services to the set of data.

Abstract: Lock table management is provided for a lock manager of a database system, in which lock management is provided in a manner that is fast and efficient, and that conserves processing, memory, and other computational resources. For example, the lock table management can use a hashmap in which keys and values are stored in separate arrays, which can be loaded into separate CPU cache lines.

Abstract: The disclosure generally describes methods, software, and systems for providing features in a user interface supporting gestures for navigating analytic data. A client device prepares a presentation including, for each data value, a data value display element and an association between the data value display element and a display location in a display area. User interaction information is received identifying a user interaction with the presentation of the data values and a user interaction location in the display area in which the user interaction occurred. A UI operation is identified to apply to the data values based on associations between the data value display elements and the display location. Identification is based on the received user interaction information. An updated presentation of the data values is prepared using the UI operation on the set of data values. The preparing occurs without obtaining additional data values from a server.

Abstract: A method for optimizing memory access for database operations is provided. The method may include identifying an access pattern associated with a database operation. The access pattern may include data required to perform the database operation. One or more memory pages may be generated based at least on the access pattern. The one or more memory pages may include at least a portion of the data required to perform the database operation. The one or more memory pages including at least the portion of the data required to perform the database operation may be stored in a main memory. The database operation may be performed by at least loading, from the main memory and into a cache, the one or more memory pages including at least the portion of the data required to perform the database operation. Related systems and articles of manufacture, including computer program products, are also provided.

Abstract: A system and method of collecting debugging information for code running in a multi-threaded computing system. An association record is generated in an associative registry. The association record has a timestamp, an object pointer and/or further information. Upon detection of a bug, the computing system generates a debug report based on the association record and the detected bug, the debug reference comprising a reference type that references a portion of the code running in the multi-threaded computing system. Based on the debug reference, a lifecycle check of the portion of the code running in the multi-threaded computing system is generated, the lifecycle check annotating the bug within the portion of the code running in the multi-threaded computing system.

Abstract: Some embodiments provide a non-transitory machine-readable medium that stores a program. The program receives a log entry instance for creating a definition for a log entry type. The program further generates a markup associated with the log entry instance. The program also generates a document based on the markup associated with the log entry instance. The program further retrieves a set of reference documents. Each reference document is associated with a particular type of event. The program also generates a ranked list of types of events based on the document and the set of reference documents.

Abstract: Methods, systems, and computer-readable storage media for receiving a set of document-level training data including a plurality of documents, each document having a sentiment label associated therewith, receiving a set of aspect-level training data including a plurality of aspects, each aspect having a sentiment label associated therewith, training the aspect-level sentiment classifier including a long short-term memory (LSTM) network, and an output layer using one or more of pretraining, and multi-task learning based on the document-level training data and the aspect-level training data, pretraining including initializing parameters based on pretrained weights that are fine-tuned during training, and multi-task learning including simultaneous training of document-level classification and aspect-level classification, and providing the aspect-level sentiment classifier for classifying one or more aspects in one or more sentences of one or more input documents based on sentiment classes.

Abstract: A system includes determination and storage of usage data based on the audit data associated with events within the system, determination of a role associated with a user based on metadata of system objects stored in a system database and instance data of the system objects stored in a file repository, determination of usage information associated with the user based on the stored usage data, determination of one or more metrics based on the role and the usage information, and providing of a home page dashboard to the user including the determined one or more metrics.

Abstract: A system includes reception of a first query to search two or more heterogeneous data providers, generation of a first query language excerpt to define a combined key column based on key columns associated with a first data provider of the two or more heterogeneous data providers, for each of two or more columns of the first data provider, generation of a second query language excerpt to map the column of the first data provider to a respective result set key column and a respective result set value column, generation of a third query language excerpt to identify two or more of the result set value columns to search, and generation of a first query language statement based on the first, second and third query language excerpts.

Abstract: Node information for a set of nodes to display on a graph is identified. The node information includes metadata and relationship information. The relationship information indicates, for each relationship, a source node and a target node for the relationship. Weights for the nodes are determined based on the metadata associated with the nodes. The nodes are sorted based on the determined weights. A first node with a highest weight is assigned to a first layer of the graph. For each given node other than the first node: relationships of the given node to assigned nodes are determined, an assigned layer is determined for the given node based on the determined weights and the relationships of the given node to assigned nodes, and the given node is assigned to the assigned layer.

Abstract: A computer implemented system and method of memory management for an in-memory database. The system implements a paged data vector using non-uniform compression of its chunks. In this manner, the system achieves greater compression than systems that use uniform compression.

Abstract: Implementations are directed to providing a data custodian region within a public cloud, the data custodian region being specific to a customer of an enterprise having services hosted on the public cloud, the public cloud including regional data centers, through which customer data passes and/or is stored, each data center being at a location within a region, storing at least one union definition that is used to control access, transfer, and storage of customer data within respective regional data centers, the at least one union definition being provided by a data custodian associated with the customer, monitoring a plurality of actions of respective workflows executed using the one or more computer-implemented services hosted on the public cloud, for each action, logging a data event within a repository of the data custodian region, and determining whether the data event complies with the at least one union definition.

Abstract: In an example, data in a non-flat format and metadata corresponding to the data are obtained from a first database. The data is flattened into flat data and augmented with the metadata. One or more pieces of the flat data are scanned to locate a first piece of flat data having a first attribute with attribute values that are a subset of attribute values of a second attribute of a second piece of flat data. A link is then created between the first attribute of the first piece of flat data and the second attribute of the second piece of flat data. A graph structure is generated, the graph structure containing a plurality of nodes, each node corresponding to a data type of the flat data and corresponding to one or more pieces of data in the flat data of the corresponding data type.

Abstract: The present disclosure involves systems, software, and computer implemented methods for preventing reader starvation in a distributed system. One example method includes receiving, at a node that manages a data stream, a write-intent request from a writer for a first partition, the data stream associated with a graph that includes levels of partitions including a current level; identifying the first partition for the writer; providing partition information for the first partition to enable the writer to write to the first partition; determining that the writer has previously written to a second partition associated with the current level; in response to determining that the writer has previously written to the second partition: creating a new current level; associating the first partition with the new current level; identifying partitions associated with the current level; and providing a close-level instruction to each local node that includes a partition associated with the current level.

Abstract: A compiler for a software development kit is disclosed. In some embodiments, a computer-implemented method comprises receiving source code comprising an object node having a data type definition for a code value, detecting that the data type definition of the object node includes a context-dependent data type, where the context-dependent data type is dependent on at least one other data type, determining a corresponding value for each one of the other data type(s) based on the detecting that the data type definition includes the context-dependent data type, and generating a runtime instance of a software solution based on the source code, where the generating the runtime instance comprises assigning the corresponding value for each other data type as a single list ID for the object node, and the single list ID for the object node is configured to be used to determine a dedicated value of the code value.

Abstract: In one respect, there is provided a method that includes converting, into text, audio that includes a speech-based query. A first portion, a second portion, and a third portion of the text can be identified based on a semantic rule. The first portion of the text can be an operation specified by the speech-based query. The second portion of the text can be an object specified by the speech-based query. The third portion of the text can be a parameter specified by the speech-based query. A database query can be formed to include the operation being performed with respect to the object and in accordance with the parameter. Furthermore, the database query can be executed at a database. Related systems and articles of manufacture, including computer program products, are also disclosed.

Abstract: The subject matter disclosed herein provides methods for the forced ordering of a dictionary based on row identifier values. A plurality of concurrent transactions can be handled in an in-memory database. At least one of the transactions can include at least one write operation to a dictionary. Each write operation can be assigned a row identifier (ID). At least one of the write operations can be written to the dictionary out of sequence. The sequence can be based on the row ID. Each row ID in the dictionary can be mapped to a corresponding value identifier in the dictionary. The dictionary can position the value identifiers so that the corresponding row IDs are in a sorted sequential order based on the row ID. Related apparatus, systems, techniques, and articles are also described.

Abstract: Techniques and solutions are described for facilitating the conversion of tables, having primary data and an append specifying a format of append data, from a first format to a second format. A runtime object and a data dictionary entry can be created for a primary table in the second format, and at least a portion of primary data transferred to the primary table in the second format. A database layer object corresponding to the first table in the first format can be associated with a temporary runtime object and a temporary data dictionary entry so that append data can be transferred to an append table in the second format. Operations directed to the primary table in the second format can be modified to also be directed to the append table, such as using a JOIN view.