Partition management for a scalable, structured storage system is provided. The storage system provides storage represented by one or more tables, each of which includes rows that represent data entities. A table is partitioned into a number of partitions, each partition including a contiguous range of rows. The partitions are served by table servers and managed by a table master. Load distribution information for the table servers and partitions is tracked, and the table master determines to split and/or merge partitions based on the load distribution information.

Partition management in a partitioned, scalable, and available structured storage

Various embodiments and implementations include a system and method for determining distinct occurrences of a pattern in a sequence of time-stamped event instances by determining a maximum cardinality of disjoint occurrences of the pattern in the one or more sequences. The present disclosure also includes estimating an expected quantity of distinct occurrences of a pattern in a sequence of time-stamped events assigned to event categories. The present disclosure further includes identifying a surprise pattern within a sequence of time-stamped events.

System and method for analyzing a pattern in a time-stamped event sequence

Techniques are disclosed for expanding the concept of partitioning in variety of ways. In particular techniques are provided for performing multiple-dimension partitioning. In multiple-dimension partitioning, a database object is divided into partitions based on one criteria, and each of those resulting partitions is divided into sub-partitions based on a second criteria. The process of partitioning partitions based on different criteria may be repeated across any number of dimensions. Entirely different partitioning techniques may be used for each level of partitioning. The database server takes advantage of partitions when processing queries by selectively accessing a subset of partitions on disk or reducing the number of internal join operations.

Partition pruning with composite partitioning

A grid-based contact center and method of managing tasks within such a contact center is provided. Nodes within the contact center are adapted to perform multiple tasks associated with the contact center, including the management of task assignment within the contact center. Accordingly, a truly distributed contact center is provided whereby the building blocks of the contact center include the use of nodes with a conference call SRTP based Operations, Administration, and Management system.

Grid-based contact center

Techniques for eliminating one or more portions of a data object from any join step of an operation that joins multiple data objects include determining that an operation joins a first data object and a second data object. The second data object includes multiple portions. Each of multiple data units of the first data object is scanned. Based on data in the data units of the first data object, information is generated. The information indicates a portion of the second data object for exclusion. The indicated portion is excluded from an output of the operation. Only one or more portions of the second data object that are not indicated for exclusion in the information are included in a particular join step involving the second data object. By pruning a large second table, such as a fact table, the computational resources consumed by the joins are substantially reduced.

Techniques for pruning a data object during operations that join multiple data objects

A method and system for generating a decision-tree classifier in parallel in a shared-memory multiprocessor system is disclosed. The processors first generate in the shared memory an attribute list for each record attribute. Each attribute list is assigned to a processor. The processors independently determine the best splits for their respective assigned lists, and cooperatively determine a global best split for all attribute lists. The attribute lists are reassigned to the processors and split according to the global best split into the lists for child nodes. The split attribute lists are again assigned to the processors and the process is repeated for each new child node until each attribute list for the new child nodes includes only tuples of the same record class or a fixed number of tuples.

Mohammed J. Zaki

Papers

Parallel classification for data mining in a shared-memory multiprocessor system