公开(公告)号：US20080147673A1

公开(公告)日：2008-06-19

申请号：US11613036

申请日：2006-12-19

申请人： George Candea ,  Anastasios Argyros ,  Mayank Bawa

发明人： George Candea ,  Anastasios Argyros ,  Mayank Bawa

IPC分类号： G06F7/00

CPC分类号： G06F17/30563

摘要： An event tap associated with a server, such as a Web server, at a machine can transform a server event into a tuple, select a database node for the tuple, and place the tuple in a queue for that database node, and then flush the queue periodically directly into database notes. The use of an event tap can thus reduce the computational burden on the database while keeping the server event data in the database relatively fresh.

摘要翻译： 与机器上的服务器（如Web服务器）关联的事件抽头可以将服务器事件转换为元组，为元组选择数据库节点，并将该元组放入该数据库节点的队列中，然后刷新 定期直接排入数据库备注。 因此，使用事件抽头可以减少数据库的计算负担，同时将数据库中的服务器事件数据保持在较新状态。

公开(公告)号：US08156107B2

公开(公告)日：2012-04-10

申请号：US11670936

申请日：2007-02-02

申请人： Mayank Bawa ,  Anastasios Argyros ,  George Candea

发明人： Mayank Bawa ,  Anastasios Argyros ,  George Candea

IPC分类号： G06F7/00 ,  G06F17/30

CPC分类号： G06F17/30545 ,  G06F17/30312 ,  G06F17/30584 ,  Y10S707/966 ,  Y10S707/971 ,  Y10S707/972

摘要： The present invention addresses the problem of partitioning database relations across a plurality of hosts in a shared-nothing cluster while minimizing communication costs. A new partitioning approach for database relations—join-partitioning—is introduced that focuses on relations and their dependencies to a priori partition the plurality of relations to the plurality of hosts such that each host can locally compute a large class of queries over its partitioned data in isolation at query time, independent of other hosts in the cluster. Such an approach thus overcomes the primary bottleneck, i.e., the network, in deploying database systems over shared-nothing clusters by allowing them to seamlessly scale linearly to tens of thousands of hosts and manage tens or hundreds of terabytes of data.

摘要翻译： 本发明解决了在无共享群集中的多个主机之间划分数据库关系同时最小化通信成本的问题。 引入了一种用于数据库关系的新的分区方法 - 加入 - 分割，其重点在于关系及其与先验分区的依赖关系到与多个主机的多个关系，使得每个主机可以在其分区数据上本地计算一大类查询 在查询时隔离，独立于集群中的其他主机。 因此，这种方法克服了主要瓶颈，即网络在通过无共享群集部署数据库系统时，允许它们无缝地线性地扩展到成千上万的主机并管理数十或数百TB的数据。

公开(公告)号：US08849746B2

公开(公告)日：2014-09-30

申请号：US11613036

申请日：2006-12-19

申请人： George Candea ,  Anastasios Argyros ,  Mayank Bawa

发明人： George Candea ,  Anastasios Argyros ,  Mayank Bawa

IPC分类号： G06F17/00 ,  G06F17/30

CPC分类号： G06F17/30563

摘要： An event tap associated with a server, such as a Web server, at a machine can transform a server event into a tuple, select a database node for the tuple, and place the tuple in a queue for that database node, and then flush the queue periodically directly into database nodes. The use of an event tap can thus reduce the computational burden on the database while keeping the server event data in the database relatively fresh.

摘要翻译： 与机器上的服务器（如Web服务器）关联的事件抽头可以将服务器事件转换为元组，为元组选择数据库节点，并将该元组放入该数据库节点的队列中，然后刷新 定期直接排入数据库节点。 因此，使用事件抽头可以减少数据库的计算负担，同时将数据库中的服务器事件数据保持在较新状态。

公开(公告)号：US20080244585A1

公开(公告)日：2008-10-02

申请号：US11692113

申请日：2007-03-27

申请人： George Candea ,  Mayank Bawa ,  Anastasios Argyros

发明人： George Candea ,  Mayank Bawa ,  Anastasios Argyros

IPC分类号： G06F9/46

CPC分类号： G06F11/008 ,  G06F11/0781

摘要： A system and method for using failure casting to manage failures in computer system. In accordance with an embodiment, the system uses a failure casting hierarchy to cast failures of one type into failures of another type. In doing this, the system allows incidents, problems, or failures to be cast into a (typically smaller) set of failures, which the system knows how to handle. In accordance with a particular embodiment, failures can be cast into a category that is considered reboot-curable. If a failure is reboot-curable then rebooting the system will likely cure the problem. Examples include hardware failures, and reboot-specific methods that can be applied to disk failures and to failures within clusters of databases. The system can even be used to handle failures that were hitherto unforeseen—failures can be cast into known failures based on the failure symptoms, rather than any underlying cause.

摘要翻译： 一种使用故障转移来管理计算机系统故障的系统和方法。 根据一个实施例，系统使用失败铸造层次结构将一种类型的故障投射到另一类型的故障中。 在这样做时，系统会将事件，问题或故障转化为（通常较小的）故障集，系统知道如何处理。 根据特定实施例，可以将故障转换为被认为可重新启动的类别。 如果故障是可重新启动的，那么重启系统可能会解决问题。 示例包括硬件故障和可重新启动的方法，可用于磁盘故障和数据库集群中的故障。 该系统甚至可以用于处理到目前为止不可预见的故障 - 故障可以基于故障症状而不是任何潜在的原因而被转化为已知故障。

公开(公告)号：US08359495B2

公开(公告)日：2013-01-22

申请号：US11692113

申请日：2007-03-27

申请人： George Candea ,  Mayank Bawa ,  Anastasios Argyros

发明人： George Candea ,  Mayank Bawa ,  Anastasios Argyros

IPC分类号： G06F11/00 ,  G06F11/22

CPC分类号： G06F11/008 ,  G06F11/0781

摘要： A system and method for using failure casting to manage failures in a computer system. In accordance with an embodiment, the system uses a failure casting hierarchy to cast failures of one type into failures of another type. In doing this, the system allows incidents, problems, or failures to be cast into a (typically smaller) set of failures, which the system knows how to handle. In accordance with a particular embodiment, failures can be cast into a category that is considered reboot-curable. If a failure is reboot-curable then rebooting the system will likely cure the problem. Examples include hardware failures, and reboot-specific methods that can be applied to disk failures and to failures within clusters of databases. The system can even be used to handle failures that were hitherto unforeseen—failures can be cast into known failures based on the failure symptoms, rather than any underlying cause.

摘要翻译： 一种使用故障转移来管理计算机系统故障的系统和方法。 根据一个实施例，系统使用失败铸造层次结构将一种类型的故障投射到另一类型的故障中。 在这样做时，系统会将事件，问题或故障转化为（通常较小的）故障集，系统知道如何处理。 根据特定实施例，可以将故障转换为被认为可重新启动的类别。 如果故障是可重新启动的，那么重启系统可能会解决问题。 示例包括硬件故障和可重新启动的方法，可用于磁盘故障和数据库集群中的故障。 该系统甚至可以用于处理到目前为止不可预见的故障 - 故障可以基于故障症状而不是任何潜在的原因而被转化为已知故障。

公开(公告)号：US20080189239A1

公开(公告)日：2008-08-07

申请号：US11670936

申请日：2007-02-02

申请人： Mayank Bawa ,  Anastasios Argyros ,  George Candea

发明人： Mayank Bawa ,  Anastasios Argyros ,  George Candea

IPC分类号： G06F17/30

CPC分类号： G06F17/30545 ,  G06F17/30312 ,  G06F17/30584 ,  Y10S707/966 ,  Y10S707/971 ,  Y10S707/972

摘要： The present invention addresses the problem of partitioning database relations across a plurality of hosts in a shared-nothing cluster while minimizing communication costs. A new partitioning approach for database relations—join-partitioning—is introduced that focuses on relations and their dependencies to a priori partition the plurality of relations to the plurality of hosts such that each host can locally compute a large class of queries over its partitioned data in isolation at query time, independent of other hosts in the cluster. Such an approach thus overcomes the primary bottleneck, i.e., the network, in deploying database systems over shared-nothing clusters by allowing them to seamlessly scale linearly to tens of thousands of hosts and manage tens or hundreds of terabytes of data. This description is not intended to be a complete description of, or limit the scope of, the invention. Other features, aspects, and objects of the invention can be obtained from a review of the specification, the figures, and the claims.

摘要翻译： 本发明解决了在无共享群集中的多个主机之间划分数据库关系同时最小化通信成本的问题。 引入了一种用于数据库关系的新的分区方法 - 加入 - 分割，其重点在于关系及其与先验分区的依赖关系到与多个主机的多个关系，使得每个主机可以在其分区数据上本地计算一大类查询 在查询时隔离，独立于集群中的其他主机。 因此，这种方法克服了主要瓶颈，即网络在通过无共享群集部署数据库系统时，允许它们无缝地线性地扩展到成千上万的主机并管理数十或数百TB的数据。 本说明书不是对本发明的完整描述或限制本发明的范围。 本发明的其它特征，方面和目的可以通过对说明书，附图和权利要求的评述来获得。

公开(公告)号：US08285709B2

公开(公告)日：2012-10-09

申请号：US12779040

申请日：2010-05-12

申请人： George Candea ,  Neoklis Polyzotis

发明人： George Candea ,  Neoklis Polyzotis

IPC分类号： G06F17/30

CPC分类号： G06F17/30442

摘要： In one embodiment, a method includes concurrently executing a set of multiple queries, through a processor, to improve a resource usage within a data warehouse system. The method also includes permitting a group of users of the data warehouse system to simultaneously run a set of queries. In addition, the method includes applying a high-concurrency query operator to continuously optimize a large number of concurrent queries for a set of highly concurrent dynamic workloads.

摘要翻译： 在一个实施例中，一种方法包括通过处理器同时执行一组多个查询，以改善数据仓库系统内的资源使用。 该方法还包括允许数据仓库系统的一组用户同时运行一组查询。 此外，该方法包括应用高并发查询运算符来持续优化大量并发的查询，用于一组高度并发的动态工作负载。

公开(公告)号：US09141354B2

公开(公告)日：2015-09-22

申请号：US13453973

申请日：2012-04-23

申请人： Volodymyr Kuznetsov ,  Johannes Kinder ,  Stefan Bucur ,  George Candea

发明人： Volodymyr Kuznetsov ,  Johannes Kinder ,  Stefan Bucur ,  George Candea

IPC分类号： G06F9/44 ,  G06F9/45

CPC分类号： G06F8/42

摘要： A determination is made regarding whether to merge two symbolic analysis states. A first state corresponds to a first path through a program to a program location and a second state corresponds to a second path through the program to the program location. A set of variables of the program at the program location is determined. For each variable in the set: a) a first value of the variable in the first state is determined; b) a second value of the variable in the second state is determined; and c) a determination is made, based on the first and second values, regarding whether merging the first and second states would be advantageous. A determination is made, responsive to determining that merging the first state and the second state would not be advantageous for at least one variable in the set, not to merge the first state and the second state.

摘要翻译： 确定是否合并两个符号分析状态。 第一状态对应于通过节目到节目位置的第一路径，第二状态对应于通过节目到节目位置的第二路径。 确定程序位置处的程序的一组变量。 对于集合中的每个变量：a）确定第一状态中变量的第一个值; b）确定第二状态中的变量的第二值; 以及c）基于第一和第二值，确定是否合并第一和第二状态将是有利的。 响应于确定合并第一状态而第二状态对于集合中的至少一个变量是不利的，而不是合并第一状态和第二状态，进行确定。

公开(公告)号：US08966453B1

公开(公告)日：2015-02-24

申请号：US13304041

申请日：2011-11-23

申请人： Cristian Zamfir ,  George Candea

发明人： Cristian Zamfir ,  George Candea

IPC分类号： G06F9/45

CPC分类号： G06F8/30 ,  G06F11/3604

摘要： An execution of a target software program that causes a particular bug to manifest itself is generated. A target software program is received. A bug report generated by an execution of the target software program is received. The bug report describes a particular bug within the target software program. A bug type of the particular bug is received. A synthesized execution file is generated based on the target software program, the bug report, and the bug type. The synthesized execution file specifies an execution of the target software program that causes the particular bug to manifest itself.

摘要翻译： 生成导致特定错误自身呈现的目标软件程序的执行。 收到目标软件程序。 收到由执行目标软件程序生成的错误报告。 错误报告描述了目标软件程序中的特定错误。 收到特定错误的错误类型。 基于目标软件程序，错误报告和错误类型生成合成执行文件。 合成执行文件指定导致特定错误自身显示的目标软件程序的执行。

公开(公告)号：US08863096B1

公开(公告)日：2014-10-14

申请号：US13345585

申请日：2012-01-06

申请人： Stefan Bucur ,  Cristian Zamfir ,  George Candea

发明人： Stefan Bucur ,  Cristian Zamfir ,  George Candea

IPC分类号： G06F9/44 ,  G06F9/45 ,  G06F9/46

CPC分类号： G06F11/3604 ,  G06F9/5088

摘要： A symbolic execution task is dynamically divided among multiple computing nodes. Each of the multiple computing nodes explores a different portion of a same symbolic execution tree independently of other computing nodes. Workload status updates are received from the multiple computing nodes. A workload status update includes a length of a job queue of a computing node. A list of the multiple computing nodes ordered based on the computing nodes' job queue lengths is generated. A determination is made regarding whether a first computing node in the list is underloaded. A determination is made regarding whether a last computing node in the list is overloaded. Responsive to the first computing node being underloaded and the last computing node being overloaded, a job transfer request is generated that instructs the last computing node to transfer a set of one or more jobs to the first computing node.

摘要翻译： 符号执行任务在多个计算节点之间动态划分。 多个计算节点中的每一个独立于其他计算节点探索相同符号执行树的不同部分。 从多个计算节点接收工作负载状态更新。 工作负载状态更新包括计算节点的作业队列的长度。 生成基于计算节点的作业队列长度排列的多个计算节点的列表。 确定列表中的第一计算节点是否欠载。 确定列表中的最后一个计算节点是否过载。 响应于第一计算节点被负载并且最后的计算节点被重载，生成作业传送请求，其指示最后的计算节点将一组一个或多个作业传送到第一计算节点。