公开(公告)号：US5065311A

公开(公告)日：1991-11-12

申请号：US184075

申请日：1988-04-20

申请人： Kazuo Masai ,  Satoshi Wakayama ,  Shoji Yamamoto ,  Takashi Sumiyoshi ,  Masao Makino

发明人： Kazuo Masai ,  Satoshi Wakayama ,  Shoji Yamamoto ,  Takashi Sumiyoshi ,  Masao Makino

IPC分类号： G06F11/14

CPC分类号： G06F11/1402 ,  Y10S707/99953

摘要： In a composite system in which a plurality of data base systems and data communication are treated as respective subsystems. At the occurrence of a fault in one subsystem, other subsystems are allowed to operate continuously, thereby facilitating recovery after a faulty subsystem has started up. A controller performs as to which data base system a transaction in execution has accessed so that the range of failure is confined, thereby facilitating the fault recovery. The check points of two processings including updating of information in the memory and accumulation of the updated information in the journal are detected so that the need of journals for information occurring earlier than the check point is eliminated, and a check point dump is acquired without waiting for the end of a transaction which has been active at the check point.

摘要翻译： 在其中将多个数据库系统和数据通信视为相应子系统的复合系统中。 在一个子系统发生故障时，允许其他子系统连续运行，从而有助于在故障子系统启动后进行恢复。 控制器执行执行中的事务已访问哪个数据库系统，从而限制故障范围，从而便于故障恢复。 检测包括更新存储器中的信息和日志中更新的信息的累积的两个处理的检查点，以消除在检查点之前发生的信息的期刊的需要，并且获取检查点转储而不等待 对于在检查点活动的交易的结束。

公开(公告)号：US5333314A

公开(公告)日：1994-07-26

申请号：US701816

申请日：1991-05-17

申请人： Kazuo Masai ,  Satoshi Wakayama ,  Shoji Yamamoto ,  Takashi Sumiyoshi ,  Masao Makino

发明人： Kazuo Masai ,  Satoshi Wakayama ,  Shoji Yamamoto ,  Takashi Sumiyoshi ,  Masao Makino

IPC分类号： G06F11/14 ,  G06F15/40 ,  G06F11/00

CPC分类号： G06F11/1402 ,  Y10S707/99953

摘要： In a composite subsystem having a plurality of data base systems and data communication on a plurality of processors, a composite subsystem controller unifies other data base systems of the composite subsystem and distributed data base systems, and, at the occurrence of a fault in some subsystem, allows other subsystems to operate continuously, thereby facilitating the recovery after the faulty subsystem has started up, and makes management as to which data base system a transaction in execution has accessed so that the range of failure is confined, thereby facilitating the fault recovery. The check points of two processings including updating of information in the memory and accumulation of the updated information in the journal are detected so that the need of journals earlier than the check point is eliminated, and a check point dump is acquired without waiting for the end of transaction which has been active at the check point.

摘要翻译： 在具有多个数据库系统和多个处理器上的数据通信的复合子系统中，复合子系统控制器将复合子系统和分布式数据库系统的其他数据库系统统一起来，并且在某些子系统发生故障时 ，允许其他子系统连续运行，从而有助于在故障子系统启动后进行恢复，并对执行中的事务所访问的数据库系统进行管理，使故障范围受到限制，从而便于故障恢复。 检测包括更新存储器中的信息和日志中的更新信息的累积的两个处理的检查点，使得消除了比检查点早的期刊的需要，并且获取检查点转储而不等待结束 的交易活动在检查点。

公开(公告)号：US5274809A

公开(公告)日：1993-12-28

申请号：US940347

申请日：1992-09-03

申请人： Masaaki Iwasaki ,  Yoshifumi Takamoto ,  Shoji Yamamoto ,  Takashi Sumiyoshi ,  Kazuo Masai ,  Toshiharu Shinozaki ,  Tetsuo Saito

发明人： Masaaki Iwasaki ,  Yoshifumi Takamoto ,  Shoji Yamamoto ,  Takashi Sumiyoshi ,  Kazuo Masai ,  Toshiharu Shinozaki ,  Tetsuo Saito

IPC分类号： G06F9/46

CPC分类号： G06F9/526

摘要： Task execution control for a multiprocessor wherein at a time point when a post issue task ends the use of a shared resource, the shared resource is released, another task which is running on another processor is allowed to lock the shared resource, and thereafter the task is made ready. After that, the post procedure is initiated, and thereafter through the high-speed dispatch procedure, the processor is granted first to a task which has failed to lock the shared resource, and the task is prompted to retry to lock the shared resource.

摘要翻译： 对于多处理器的任务执行控制，其中在后发行任务结束共享资源的使用的时间点，共享资源被释放，允许在另一个处理器上运行的另一个任务锁定共享资源，之后任务 准备好了 之后，发起后程序，之后通过高速调度程序，将处理器首先授予未能锁定共享资源的任务，并提示该任务重试以锁定共享资源。

公开(公告)号：US5778388A

公开(公告)日：1998-07-07

申请号：US529176

申请日：1995-09-15

申请人： Nobuo Kawamura ,  Kazuo Masai ,  Nobuyuki Yamashita ,  Hiroshi Nagai

发明人： Nobuo Kawamura ,  Kazuo Masai ,  Nobuyuki Yamashita ,  Hiroshi Nagai

IPC分类号： G06F11/14 ,  G06F12/00 ,  G06F17/30

CPC分类号： G06F11/1466 ,  G06F11/1474 ,  G06F2201/80 ,  Y10S707/99932 ,  Y10S707/99953 ,  Y10S707/99954

摘要： In a database management method and a database management system in which transactions need not be stopped at a synchronization point (syncpoint) acquired at a fixed interval of time, when a point to acquire a syncpoint is reached in a database control procedure, a syncpoint acquisition start log is output and then a syncpoint acquisition flag is set ON in a page control table of each updated page existing in a buffer pool. In the operation, a list of update page control tables is also created to write in the database the pages which are indicated in the list thus prepared. However, for a transaction accessing the pertinent updated page prior to the write operation thereof in the database, the data write operation is performed before the page is referenced, thereby enabling the access to be performed to the database without interrupting the transaction.

摘要翻译： 在数据库管理方法和数据库管理系统中，其中事务不需要在以固定时间间隔获取的同步点（syncpoint）停止，当在数据库控制过程中达到获取同步点的点时，同步点获取 输出开始日志，然后在存储在缓冲池中的每个更新页面的页面控制表中将同步点获取标志设置为ON。 在操作中，还创建了更新页面控制表的列表，以便在数据库中写入由此准备的列表中指示的页面。 然而，对于在其在数据库中的写入操作之前访问相关更新页面的事务，在引用页面之前执行数据写入操作，从而使得能够对数据库执行访问而不中断事务。

公开(公告)号：US06584457B1

公开(公告)日：2003-06-24

申请号：US09662911

申请日：2000-09-15

申请人： Hideaki Kasao ,  Keisuke Toda ,  Kota Yamaguchi ,  Kazuo Masai ,  Kiyomi Hirohata

发明人： Hideaki Kasao ,  Keisuke Toda ,  Kota Yamaguchi ,  Kazuo Masai ,  Kiyomi Hirohata

IPC分类号： G06F1730

CPC分类号： G06F17/30445 ,  Y10S707/99931 ,  Y10S707/99933 ,  Y10S707/99937 ,  Y10S707/99944 ,  Y10S707/99945

摘要： A parallel data base management system with a data base divided into a plurality of external memory units is disclosed. A control CPU, upon receipt of a request for access to the data base, decides whether the access request is for storing or searching a record. For a record storage request, the control CPU determines an external memory unit to store the record using a hash function. In the case of a record search request, on the other hand, all the external memory units are searched in parallel without using the hash function thereby to obtain the desired record. The hash function is altered upon addition of an external memory unit.

摘要翻译： 公开了一种具有划分为多​​个外部存储器单元的数据库的并行数据库管理系统。 控制CPU在接收到访问数据库的请求时，决定访问请求是存储还是搜索记录。 对于记录存储请求，控制CPU使用散列函数确定外部存储单元来存储记录。 另一方面，在记录搜索请求的情况下，并行地搜索所有外部存储器单元，而不使用散列函数从而获得所需的记录。 添加外部存储单元后，哈希函数会被更改。

公开(公告)号：US5717918A

公开(公告)日：1998-02-10

申请号：US467278

申请日：1995-06-06

申请人： Hiroshi Takayama ,  Mitsuo Miyazaki ,  Kazuo Masai ,  Kuniaki Yamashita

发明人： Hiroshi Takayama ,  Mitsuo Miyazaki ,  Kazuo Masai ,  Kuniaki Yamashita

IPC分类号： G06F17/30

CPC分类号： G06F17/30492 ,  Y10S707/99938 ,  Y10S707/99953

摘要： A database system is provided for reading data from a database to a database buffer and performing a physical sequential scan for the data on the buffer. If the preceding request for physical sequential scan is under execution, the database system operates to keep the succeeding request for scan in a waiting state until the preceding requested scan is completed and then execute the succeeding request for scan after the preceding requested scan is completed. If there are more than one succeeding requests for physical sequential scan, they are executed at the same time. Another scanning method is also provided. If the preceding request for scan is under execution, the succeeding requested scan is executed for the data existing after the halfway location at which the preceding scan process now reads the data. The short data is read from its head for completing the subsequent scan.

摘要翻译： 提供了一种数据库系统，用于将数据从数据库读取到数据库缓冲区，并对缓冲区上的数据执行物理顺序扫描。 如果前面的物理顺序扫描请求正在执行，则数据库系统操作以将扫描的后续请求保持在等待状态，直到完成上述请求的扫描，然后在完成先前请求的扫描之后执行后续的扫描请求。 如果有多个后续的物理顺序扫描请求，它们将同时执行。 还提供了另一种扫描方法。 如果上述扫描请求正在执行中，则对先前扫描进程现在读取数据的中途位置之后的数据执行后续请求的扫描。 从头读取短数据以完成后续扫描。

公开(公告)号：US07599910B1

公开(公告)日：2009-10-06

申请号：US09665448

申请日：2000-09-19

申请人： Masashi Tsuchida ,  Kazuo Masai ,  Shunichi Torii

发明人： Masashi Tsuchida ,  Kazuo Masai ,  Shunichi Torii

IPC分类号： G06F17/30

CPC分类号： G06F9/5083 ,  G06F17/30584 ,  Y10S707/99932 ,  Y10S707/99933 ,  Y10S707/99936 ,  Y10S707/99953

摘要： A method and a system of database divisional management for use with a parallel database system comprising an FES (front end server), BES's (back end servers), an IOS (I/O server) and disk units. The numbers of processors assigned to the FES, BES's and IOS, the number of disk units, and the number of partitions of the disk units are determined in accordance with the load pattern in question. Illustratively, there may be established a configuration of one FES, four BES's, one IOS and eight disk units. The number of BES's is varied from one to four depending on the fluctuation in load, so that a scalable system configuration is implemented. When the number of BES's is increased or decreased, only the management information thereabout is transferred between nodes and not the data, whereby the desired degree of parallelism is obtained for high-speed query processing.

摘要翻译： 一种与并行数据库系统一起使用的数据库分区管理方法和系统，包括FES（前端服务器），BES（后端服务器），IOS（I / O服务器）和磁盘单元。 根据所讨论的负载模式确定分配给FES，BES和IOS的处理器的数量，磁盘单元的数量和磁盘单元的分区数。 说明性地，可以建立一个FES，四个BES，一个IOS和八个磁盘单元的配置。 根据负载波动，BES的数量从1到4不等，从而实现可扩展的系统配置。 当BES的数量增加或减少时，仅在节点之间传送管理信息而不是数据，从而获得高速查询处理所需的并行度。

公开(公告)号：US06510428B2

公开(公告)日：2003-01-21

申请号：US09805594

申请日：2001-03-14

申请人： Masashi Tsuchida ,  Kazuo Masai ,  Shunichi Torii

发明人： Masashi Tsuchida ,  Kazuo Masai ,  Shunichi Torii

IPC分类号： G06F1730

CPC分类号： G06F9/5083 ,  G06F17/30584 ,  Y10S707/99932 ,  Y10S707/99933 ,  Y10S707/99936 ,  Y10S707/99953

摘要： A method and a system of database divisional management for use with a parallel database system comprising an FES (front end server), BES's (back end servers), an IOS (I/O server) and disk units. The numbers of processors assigned to the FES, BES's and IOS, the number of disk units, and the number of partitions of the disk units are determined in accordance with the load pattern in question. Illustratively, there may be established a configuration of one FES, four BES's, one IOS and eight disk units. The number of BES's is varied from one to four depending on the fluctuation in load, so that a scalable system configuration is implemented. When the number of BES's is increased or decreased, only the management information thereabout is transferred between nodes and not the data, whereby the desired degree of parallelism is obtained for high-speed query processing.

公开(公告)号：US06169982A

公开(公告)日：2001-01-02

申请号：US09292606

申请日：1999-04-15

申请人： Hideaki Kasao ,  Keisuke Toda ,  Kota Yamaguchi ,  Kazuo Masai ,  Kiyomi Hirohata

发明人： Hideaki Kasao ,  Keisuke Toda ,  Kota Yamaguchi ,  Kazuo Masai ,  Kiyomi Hirohata

IPC分类号： G06F1730

CPC分类号： G06F17/30445 ,  Y10S707/99931 ,  Y10S707/99933 ,  Y10S707/99937 ,  Y10S707/99944 ,  Y10S707/99945

摘要： A parallel data base management system with a data base divided into a plurality of external memory units is disclosed. A control CPU, upon receipt of a request for access to the data base, decides whether the access request is for storing or searching a record. For a record storage request, the control CPU determines an external memory unit to store the record using a hash function. In the case of a record search request, on the other hand, all the external memory units are searched in parallel without using the hash function thereby to obtain the desired record. The hash function is altered upon addition of an external memory unit.

摘要翻译： 公开了一种具有划分为多​​个外部存储器单元的数据库的并行数据库管理系统。 控制CPU在接收到访问数据库的请求时，决定访问请求是存储还是搜索记录。 对于记录存储请求，控制CPU使用散列函数确定外部存储单元来存储记录。 另一方面，在记录搜索请求的情况下，并行地搜索所有外部存储器单元，而不使用散列函数从而获得所需的记录。 添加外部存储单元后，哈希函数会被更改。

公开(公告)号：US5987462A

公开(公告)日：1999-11-16

申请号：US840643

申请日：1997-04-25

申请人： Hideaki Kasao ,  Keisuke Toda ,  Kota Yamaguchi ,  Kazuo Masai ,  Kiyomi Hirohata

发明人： Hideaki Kasao ,  Keisuke Toda ,  Kota Yamaguchi ,  Kazuo Masai ,  Kiyomi Hirohata

IPC分类号： G06F12/00 ,  G06F17/30

CPC分类号： G06F17/30445 ,  Y10S707/99931 ,  Y10S707/99933 ,  Y10S707/99937 ,  Y10S707/99944 ,  Y10S707/99945

摘要： A parallel data base management system with a data base divided into a plurality of external memory units is disclosed. A control CPU, upon receipt of a request for access to the data base, decides whether the access request is for storing or searching a record. For a record storage request, the control CPU determines an external memory unit to store the record using a hash function. In the case of a record search request, on the other hand, all the external memory units are searched in parallel without using the hash function thereby to obtain the desired record. The hash function is altered upon addition of an external memory unit.