公开(公告)号：US07676453B2

公开(公告)日：2010-03-09

申请号：US10830510

申请日：2004-04-22

申请人： Cetin Ozbutun

发明人： Cetin Ozbutun

IPC分类号： G06F7/00 ,  G06F17/30

CPC分类号： G06F17/30457 ,  Y10S707/99933

摘要： When a query is performed an execution plan is formed, which is made up of a plurality of plan operators. Before executing any given plan operator, a storage area is checked. If a results set has been stored corresponding to the plan operator, then the cached results set is used instead of the executing the plan operator. If no results set exists that corresponds to the particular plan operator, the plan operator is executed, and the results set that is output may be stored for use by another plan operator of another execution plan.

摘要翻译： 当执行查询时，形成由多个计划运算符组成的执行计划。 在执行任何给定的计划运算符之前，检查存储区域。 如果对应于计划运算符存储结果集，则使用缓存的结果集而不是执行计划运算符。 如果没有与特定计划运算符相对应的结果集，则执行计划运算符，并且输出的结果集可以存储供另一计划运算符使用。

公开(公告)号：US07653826B1

公开(公告)日：2010-01-26

申请号：US12356434

申请日：2009-01-20

申请人： Paul Reuben Day ,  Randy L. Egan ,  Roger Alan Mittelstadt ,  Brian Robert Muras

发明人： Paul Reuben Day ,  Randy L. Egan ,  Roger Alan Mittelstadt ,  Brian Robert Muras

IPC分类号： G06F1/32

CPC分类号： G06F3/0634 ,  G06F3/0625 ,  G06F3/0676 ,  G06F17/30457 ,  Y02D10/154 ,  Y02D10/45

摘要： The invention, in an embodiment, relates to optimizing database performance by taking advantage of sleepy drives. The invention analyzes queries to indexes based on usage frequency, and determines whether to access an index stored on a sleepy drive, or perform an alternate access plan based on green optimization. Furthermore, the invention transfers data, such as, for example, indexes, files, data areas, journals, and statistics, between active and sleepy drives based on usage frequency. In another embodiment, the invention provides an improved sleepy drive which optimizes energy savings by a graduated degree based on a speed setting. In yet another embodiment, the invention controls when read-write heads are unloaded from disks to reduce drag and power consumption by a disk drive.

摘要翻译： 在一个实施例中，本发明涉及通过利用睡眠驱动来优化数据库性能。 本发明基于使用频率分析对索引的查询，并且确定是否访问存储在睡眠驱动器上的索引，或者基于绿色优化来执行替代访问计划。 此外，本发明基于使用频率，在活动和睡眠驱动器之间传送数据，例如索引，文件，数据区域，期刊和统计信息。 在另一个实施例中，本发明提供一种改进的睡眠驱动器，其基于速度设置优化节能度量级。 在另一个实施例中，本发明控制读写头从盘上卸载以减少磁盘驱动器的阻力和功耗。

公开(公告)号：US20100017436A1

公开(公告)日：2010-01-21

申请号：US12504695

申请日：2009-07-17

申请人： Hakan WOLGE

发明人： Hakan WOLGE

IPC分类号： G06F17/30

CPC分类号： G06F17/3048 ,  G06F17/30457

摘要： Information is extracted from a database using a computer-implemented method that involves a sequential chain of main calculations, in which a first main calculation (P1) operates a first selection item (S1) on a data set (R0) that represents the database to produce a first result (R1), and a second main calculation (P2) operates a second selection item (S2) on the first result (R1) to produce a second result (R2). The first and second results (R1, R2) are cached in computer memory (10) for re-use in subsequent iterations of the method, thereby reducing the need to execute the first and/or second main calculations (P1, P2) for extracting the information. The caching involves calculating a first selection identifier value (ID1) as a function of at least the first selection item (S1), and a second selection identifier value (ID3) as a function of at least the second selection item (S2) and the first result (R1), and storing the first selection identifier value (ID1) and the first result (R1), and the second selection identifier value (ID3) and the second result (R2), respectively, as associated objects in a data structure (12). Each of the identifier values are generated as a statistically unique digital fingerprint by a hash function (f). The re-use involves calculating the first and second selection identifier values (ID1, ID3) during a subsequent iteration and accessing the data structure (12) to potentially retrieve the first and/or second result (R1, R2).

摘要翻译： 使用涉及主计算的顺序链的计算机实现的方法从数据库中提取信息，其中第一主计算（P1）在表示数据库的数据集（R0）上操作第一选择项（S1） 产生第一结果（R1），并且第二主计算（P2）对第一结果（R1）操作第二选择项目（S2）以产生第二结果（R2）。 第一和第二结果（R1，R2）被缓存在计算机存储器（10）中，以在该方法的后续迭代中重新使用，从而减少执行用于提取的第一和/或第二主计算（P1，P2）的需要 信息。 缓存涉及根据至少第一选择项目（S1）计算第一选择标识符值（ID1），以及作为至少第二选择项目（S2）的函数的第二选择标识符值（ID3）和 第一结果（R1），并分别存储第一选择标识符值（ID1）和第一结果（R1）和第二选择标识符值（ID3）和第二结果（R2）作为数据结构中的关联对象 （12）。 通过散列函数（f）将每个标识符值生成为统计唯一的数字指纹。 重用涉及在后续迭代期间计算第一和第二选择标识符值（ID1，ID3）并访问数据结构（12）以潜在地检索第一和/或第二结果（R1，R2）。

公开(公告)号：US20090271379A1

公开(公告)日：2009-10-29

申请号：US12455664

申请日：2009-06-04

申请人： Reuven Bakalash ,  Guy Shaked ,  Joseph Caspi

发明人： Reuven Bakalash ,  Guy Shaked ,  Joseph Caspi

IPC分类号： G06F17/30 ,  G06F7/00

CPC分类号： G06F17/30592 ,  G06F17/30457 ,  G06F17/30477 ,  G06F17/30489 ,  G06F17/3061 ,  Y10S707/954 ,  Y10S707/957 ,  Y10S707/99932 ,  Y10S707/99934 ,  Y10S707/99935 ,  Y10S707/99943

摘要： A relational database management system (RDBMS) for servicing query statements from one or more client machines, using a relational data store and a multi-dimensional database (MDDB). The RDBMS includes a query interface which is adapted to receive query statements from the client machines, and to generate one or more query requests. The RDBMS includes a query handling mechanism which receives each request from the query interface, extracts a set of dimensions associated with the request, uses the dimensions to retrieve aggregated fact data from the MDDB, and forwards retrieved aggregated fact data to the query processing mechanism for subsequent processing. When the query processing mechanism of the RDBMS determines that servicing one or more query requests requires data stored in the relational tables, then the query handling mechanism automatically routes the requests to the relational data tables, so that data can be accessed from the relational tables and forwarded to the query processing mechanism for use in servicing the requests, in a manner transparent to the client machine. When the query processing mechanism determines that servicing one or more query requests requires aggregated data from the MDDB, then the query handling mechanism automatically routes the requests to the MDDB, so that data can be accessed from the MDDB and forwarded to the query processing mechanism for use in servicing the requests, in a manner transparent to the client machine.

摘要翻译： 用于使用关系数据存储和多维数据库（MDDB）从一个或多个客户端机器处理查询语句的关系数据库管理系统（RDBMS）。 RDBMS包括一个查询接口，适用于从客户机接收查询语句，并生成一个或多个查询请求。 RDBMS包括查询处理机制，从查询接口接收每个请求，提取与请求相关联的维度集合，使用维度从MDDB检索汇总事实数据，并将检索到的汇总事实数据转发到查询处理机制 后续处理。 当RDBMS的查询处理机制确定服务一个或多个查询请求需要存储在关系表中的数据时，查询处理机制自动将请求路由到关系数据表，从而可以从关系表访问数据， 以对客户端机器透明的方式转发到用于服务请求的查询处理机制。 当查询处理机制确定服务一个或多个查询请求需要来自MDDB的聚合数据时，查询处理机制自动将请求路由到MDDB，从而可以从MDDB访问数据并将其转发到查询处理机制 用于以对客户端机器透明的方式来处理请求。

公开(公告)号：US07533115B2

公开(公告)日：2009-05-12

申请号：US10664200

申请日：2003-09-16

申请人： Kehsing J. Chou ,  Basuki N. Soetarman ,  Robert Nelson Summers ,  Mimi Phuong-Thao Vo ,  Alan Tsu-I Yaung

发明人： Kehsing J. Chou ,  Basuki N. Soetarman ,  Robert Nelson Summers ,  Mimi Phuong-Thao Vo ,  Alan Tsu-I Yaung

IPC分类号： G06F7/00

CPC分类号： G06F17/30566 ,  G06F17/30457 ,  Y10S707/99933 ,  Y10S707/99943

摘要： A persistent federated folder within a federated content management system can be used to save federated search results, to act as a container in a workflow process, or a container for a set of documents or folders with similar characteristics. The federated folder and its associated storage structure have the capability to save its instances persistently. The federated folder operations can selectively create, retrieve, update, and/or delete an instance of a federated folder in addition to adding, retrieving updating and deleting its parts and its members. Seamless integration with the federated query is provided to operate on the federated folder transparently regardless of the locality of the user or the datastores the user is searching. Access rights for federated folders are enforced according to user's general privileges.

摘要翻译： 联合内容管理系统中的永久联合文件夹可用于保存联合搜索结果，充当工作流过程中的容器，或用于一组具有相似特征的文档或文件夹的容器。 联合文件夹及其关联的存储结构具有持久保存其实例的能力。 除了添加，检索更新和删除其部件及其成员之外，联合文件夹操作可以有选择地创建，检索，更新和/或删除联合文件夹的实例。 提供与联合查询的无缝集成，可以透明地对联合文件夹进行操作，而不管用户的位置或用户正在搜索的数据存储区。 联合文件夹的访问权限根据用户的一般权限执行。

公开(公告)号：US20090089266A1

公开(公告)日：2009-04-02

申请号：US11863045

申请日：2007-09-27

申请人： Sreenivas Gollapudi ,  Rina Panigrahy

发明人： Sreenivas Gollapudi ,  Rina Panigrahy

IPC分类号： G06F17/30

CPC分类号： G06F17/30693 ,  G06F17/30457

摘要： A method is disclosed for identifying queries stored in a log which are semantically related to an input query that may include a large number of terms. A set of one or more subsequences are generated for each query stored in the log, and these sets of subsequences are stored in a lookup table. A set of one or more subsequences are also generated for the input query. The subsequences in the lookup table and of the input query are generated by hashing of the respective query terms to a value between 0 and 1 using a known technique of min-hashing. The present system then constructs the subsequences of the query using the k-min hashes of the query, where k is an integer based on the number of terms in the query.

摘要翻译： 公开了一种用于识别存储在日志中的查询与可能包括大量项目的输入查询语义相关的查询的方法。 为存储在日志中的每个查询生成一组或多个子序列，并将这些子序列存储在查找表中。 还为输入查询生成一组或多个子序列。 查询表和输入查询中的子序列通过使用最小哈希算法的已知技术将相应的查询项散列为0到1之间的值来生成。 本系统然后使用查询的k-min哈希构建查询的子序列，其中k是基于查询中的项数的整数。

公开(公告)号：US07440963B1

公开(公告)日：2008-10-21

申请号：US10337714

申请日：2003-01-07

申请人： Randall G. Bello ,  Bhujanga Panchapagesan ,  Tsae-Feng Yu ,  John D. Raitto

发明人： Randall G. Bello ,  Bhujanga Panchapagesan ,  Tsae-Feng Yu ,  John D. Raitto

IPC分类号： G06F17/30

CPC分类号： G06F17/30457 ,  Y10S707/99933

摘要： An approach for processing a query in a database system is disclosed. An original query is determined to be partially satisfied by a materialized view. The original query is transformed using a set operation (i.e., union and intersection) on the materialized view and either another materialized view or a base table.

摘要翻译： 公开了一种在数据库系统中处理查询的方法。 原始查询被确定为通过物化视图部分满足。 原始查询使用物化视图和另一物化视图或基表上的设置操作（即联合和交集）进行转换。

公开(公告)号：US07401084B1

公开(公告)日：2008-07-15

申请号：US09881505

申请日：2001-06-14

申请人： Bipul Binit Sinha ,  Amit Ganesh ,  Lakshminarayanan Chidambaran

发明人： Bipul Binit Sinha ,  Amit Ganesh ,  Lakshminarayanan Chidambaran

IPC分类号： G06F17/30

CPC分类号： G06F17/30575 ,  G06F17/30457

摘要： An optimization of the two-phase commit protocol employed in distributed systems. Each cohort component of the distributed system augments messages that are accessible to the coordinator component with state information indicating whether the cohort is read-only with regard to a transaction that the coordinator is coordinating. The coordinator retains the most recent state information. The coordinator reads the retained state information for the cohorts and when the transaction terminates, the coordinator reads the retained state information. Where the cohort's state is read-only, the coordinator simply sends an abort message instead of performing the full two-phase commit protocol with regard to the read-only cohort. In the trees of distributed system components that are defined by transactions, any cohort which has children in the tree is a local coordinator for its children. The general technique of cohorts augmenting messages accessible to a coordinator with state information that is relevant to an action to be performed by the coordinator and the coordinator acting in accordance with state retained from the messages has other uses as well.

摘要翻译： 分布式系统中采用的两阶段提交协议的优化。 分布式系统的每个队列组件增加协调器组件可访问的消息，状态信息指示该协调器正在协调的关于事务的队列是否为只读。 协调员保留最新的状态信息。 协调器读取队列的保留状态信息，当事务终止时，协调器读取保留的状态信息。 在队列状态为只读状态的情况下，协调器只需发送中止消息，而不是针对只读队列执行完整的两阶段提交协议。 在由事务定义的分布式系统组件的树中，树中有子节点的任何队列都是其子节点的本地协调器。 队列的一般技术将协调器可访问的消息扩展到状态信息，该状态信息与由协调器和协调器执行的动作相关，该协调器和协调器根据从消息保留的状态来执行，还有其他用途。

公开(公告)号：US20080082493A1

公开(公告)日：2008-04-03

申请号：US11537587

申请日：2006-09-29

申请人： Ricardo Polo-Malouvier ,  Gregoire Jean Antoine Cacheux ,  Gilles Vergnory-Mion ,  Mark Allerton

发明人： Ricardo Polo-Malouvier ,  Gregoire Jean Antoine Cacheux ,  Gilles Vergnory-Mion ,  Mark Allerton

IPC分类号： G06F17/30

CPC分类号： G06F17/30592 ,  G06F17/30457

摘要： A method of receiving a report includes submitting a search query, receiving a proposed abstract query and report search results, and accepting a choice between submitting the proposed abstract query to generate a report and designating an existing report from the report search results. A report is then received in accordance with the choice.

摘要翻译： 接收报告的方法包括提交搜索查询，接收所提出的抽象查询和报告搜索结果，以及在提交所提出的抽象查询以生成报告并从报告搜索结果指定现有报告之间接受选择。 然后根据选择收到报告。

公开(公告)号：US20070255683A1

公开(公告)日：2007-11-01

申请号：US11413231

申请日：2006-04-28

申请人： Shaun Cox

发明人： Shaun Cox

IPC分类号： G06F17/30

CPC分类号： G06Q10/10 ,  G06F17/30457 ,  H04L51/04

摘要： A system, apparatus, method, and computer-readable media are provided for efficient database lookup operations. According to method, procedural checks for generating a response to a database query are converted into a set of database operations. The database operations are performed the first time a query of the database is received to generate a set of possible results for the query. The set of possible results for the query are then stored. A specific answer to the query is generated from the set of possible results. Future database queries are responded to by generating an answer to the query from the stored set of possible results.

摘要翻译： 提供了一种用于有效的数据库查找操作的系统，装置，方法和计算机可读介质。 根据方法，用于生成对数据库查询的响应的过程检查被转换成一组数据库操作。 数据库操作首次在接收到数据库的查询时执行，以生成查询的一组可能的结果。 然后存储查询的可能结果集。 查询的具体答案是从一组可能的结果生成的。 通过从存储的可能结果集中生成查询的答案来响应未来的数据库查询。