International Business Machines Corporation

Kevin Leo Beck

Kevin James Cherkauer

Paul Michael Brett

Catherine Suzanne McArthur

William T O'Connell

Jeffrey James Goss

Dieu Quang La

Roger C Raphael

Paul Sherwood Taylor

Ronen Grosman

Scott David Lashley

Bingjie Miao

Gary Valentin

Keith G Billings

Paula N Collins

Vijay Gopalakrishnan

Tania M Zirn

The intermittent sampling of storage access frequency is performed by determining a duration of a collection window and a duration of an observation window within the collection window. A position of the observation window within the collection window is randomly selected, and frequencies of accesses of one or more storage objects during the observation window are observed. When a new access of a given storage object occurs, a delta time for the given storage object is calculated as the time of the observed access minus the timestamp of the most recent observed prior access of the given storage object. Optionally, the delta time of two sequential accesses of a given storage object in two different observation windows may be calculated as if the two different observation windows are immediately adjacent to each other.

Intermittent sampling of storage access frequency

In defining database objects for storage in a storage hierarchy, frequencies of accesses of a plurality of database objects over a predetermined time period are observed. A mean and standard deviation for the plurality of database objects are computed based on the observed frequencies of accesses of the plurality of database objects. A z-score for a given database object is determined based on a comparison of the frequency of access for the given database with the mean and standard deviation computed for the plurality of database objects by the computing processor; and a level in the storage hierarchy corresponding to the z-score of the given database object is determined. The given database object may then be stored at the level in the storage hierarchy corresponding to the z-score of the given database object.

Autonomically defining hot storage and heavy workloads

A system and a method for the deferral and incremental performance of integrity processing for data rolled into a table of a data warehouse and for the deferral and incremental performance of maintenance on data rolled out of a table of the data warehouse while permitting general users online access to other data contained in the table. The rolled in and rolled out data are contained within specific data partitions of the table. Each data partition in the table has an associated distinguishable partition ID. Each data partition can also have associated attributes corresponding to containing rolled-in and rolled-out data respectively. Table operation mechanisms are arranged such that general users of the table do not access rolled in data in a partition for which integrity processing has not yet been performed and do not access rolled out data.

Online incremental deferred integrity processing and maintenance of rolled in and rolled out data

A choose tree is generated at compile time to facilitate a determination if any partition of a range partitioned database table includes a particular key at runtime. To generate the choose tree at compile time, a query compiler receives a source code indicative of each partition of the range partitioned database table, and generates the choose tree from the source code at compile time where the choose tree is representative of a pre-computed binary search of the range partitioned database table. Thereafter, the choose tree can be interpreted/executed at runtime to facilitate a traversal of the choose tree for determining if any partition of the range partitioned database table includes the particular key.

Method for generating a choose tree for a range partitioned database table

A method and system are provided to enable restoration of a base table and associated catalog table from a database. Table names and table identifiers are copied to a header of a storage media during a backup of the database. When a restore of a designated table is requested, table identifier data is extracted from the header of the storage media housing the backup of the database. Accordingly, the designated base table and associated catalog table(s) are restored from the storage media based upon data stored in the header of the storage media during the backup of the database.

System to automate schema creation for table restore

A method, apparatus and article of manufacture, implementing the method, accesses rows in a range-clustered table in a database management system. A table stores data that is organized into rows and columns. At least one column is designated as a key column. A particular row in the table is accessed based on a relationship between a specified key value and a predetermined range of key values associated with the key column. The table has slots for storing rows, and the slots are organized in accordance with the values of the predetermined range.

Range-clustered tables in a database management system

Growing community of inventors

Portland, OR, United States of America

Kevin Leo Beck