会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 2. 发明专利
    • Retrieval program, retrieval method, retrieval device and node
    • 检索程序,检索方法,检索设备和节点
    • JP2012252658A
    • 2012-12-20
    • JP2011126764
    • 2011-06-06
    • Fujitsu Ltd富士通株式会社
    • WATANABE TAKASHITSUCHIYA YOSHIHIRONOGUCHI YASUO
    • G06F17/30G06F12/00
    • G06F17/30946
    • PROBLEM TO BE SOLVED: To speed up data retrieval and increase a manageable data amount.SOLUTION: Because retrieval object data Dx is stored in a data block db4 of a node N2, the node N2 indicates to a client C1 that the retrieval object data Dx exists. On the other hand, because the retrieval object data Dx is not stored in a data block db7 of a node N4, the node N4 indicates to the client C1 that the retrieval object data Dx does not exist. When the client C1 first transmits a request for the data block db4, it is not necessary to transmit a request to the node N4 because a retrieval result indicating that the retrieval object data Dx exits can be acquired. By this way, it is possible to speed up retrieval processing. By using a hierarchical type of a divided and transposed Bloom filter group like this sequence, it is possible to perform retrieval of object data at higher speed and in a wide range.
    • 要解决的问题:加快​​数据检索和增加可管理的数据量。 解决方案:由于检索对象数据Dx存储在节点N2的数据块db4中,所以节点N2向客户端C1指示检索对象数据Dx存在。 另一方面,由于检索对象数据Dx未存储在节点N4的数据块db7中,所以节点N4向客户端C1指示检索对象数据Dx不存在。 当客户端C1首先发送对数据块db4的请求时,不需要向节点N4发送请求,因为可以获取指示检索对象数据Dx退出的检索结果。 通过这种方式,可以加快检索处理。 通过使用像这种顺序的分层和转置的布隆过滤器组的分级类型,可以以更高的速度和更广的范围执行对象数据的检索。 版权所有(C)2013,JPO&INPIT
    • 4. 发明专利
    • Disk array device, method and program for managing data distribution
    • 磁盘阵列设备,用于管理数据分配的方法和程序
    • JP2010182087A
    • 2010-08-19
    • JP2009024951
    • 2009-02-05
    • Fujitsu Ltd富士通株式会社
    • NOGUCHI YASUOOGIWARA KAZUTAKATAMURA MASATOSHITSUCHIYA YOSHIHIROMARUYAMA TETSUTAROWATANABE TAKASHIKUMANO TATSUOOE KAZUICHI
    • G06F3/06
    • G06F11/1076G06F2211/104
    • PROBLEM TO BE SOLVED: To detect writing omission in reading of data even when data in one physical block are updated in updating data in a logical block. SOLUTION: In response to a write request, a reading means 1 reads the existing data and parity data. Then, a parity data generation means 2 generates new parity data. Moreover, a time-series information generation means 3 generates time-series information capable of discriminating a previous write request to the same storage area from this time write request. A time-series information application means 4 applies the generated time-series information to the update data and the parity data. Then, a writing means 5 writes the update data to which time-series information for authentication has been applied and the parity data to which time-series information for comparison has been applied in different storage devices. COPYRIGHT: (C)2010,JPO&INPIT
    • 要解决的问题:即使在更新逻辑块中的数据时更新一个物理块中的数据,也可以检测在读取数据时的写入省略。 解决方案:响应于写请求,读取装置1读取现有数据和奇偶校验数据。 然后,奇偶校验数据生成单元2生成新的奇偶校验数据。 此外,时间序列信息生成装置3从该时间写入请求生成能够将相同的存储区域的先前写入请求区分开的时间序列信息。 时间序列信息应用装置4将生成的时间序列信息应用于更新数据和奇偶校验数据。 然后,写入装置5写入已经应用了用于认证的时间序列信息的更新数据和用于比较的时间序列信息已被应用于不同存储装置的奇偶校验数据。 版权所有(C)2010,JPO&INPIT
    • 7. 发明专利
    • Method for restoring redundant configuration, data management system, and redundant configuration restoration program
    • 恢复冗余配置,数据管理系统和冗余配置恢复程序的方法
    • JP2005301594A
    • 2005-10-27
    • JP2004115742
    • 2004-04-09
    • Fujitsu Ltd富士通株式会社
    • OGIWARA KAZUTAKATAKE RIICHIROMARUYAMA TETSUTAROTAMURA MASATOSHITSUCHIYA YOSHIHIRONOGUCHI YASUOKAMOSHITA MINORU
    • G06F13/10G06F3/06G06F12/00
    • PROBLEM TO BE SOLVED: To rapidly restore a redundant configuration.
      SOLUTION: A redundant configuration restoration method for a data management system that has storage nodes having data storage means for storing data with data storage areas managed in predetermined units and control nodes connected to at least one or more of the storage nodes via a network to effect control related to the storage nodes and that manages the data by means of the redundant configuration of the data storage means includes a data copy starting step for starting the copying of the data from the data storage means to another data storage means in order to restore the redundant configuration if there exists any storage node detected to be abnormal; and a redundant configuration restoration step for restoring the redundant configuration using reusable data and the data storage means by stopping the copying of unnecessary data if data in the data storage area of the predetermined units in the storage node detected to be abnormal are reusable.
      COPYRIGHT: (C)2006,JPO&NCIPI
    • 要解决的问题:快速恢复冗余配置。 解决方案:一种用于数据管理系统的冗余配置恢复方法,所述冗余配置恢复方法具有存储节点,所述数据管理系统具有数据存储装置,用于存储与预定单元管理的数据存储区域的数据,以及经由存储节点连接到至少一个或多个存储节点 网络实现与存储节点相关的控制,并且通过数据存储装置的冗余配置来管理数据包括数据复制开始步骤,用于开始数据从数据存储装置复制到另一个数据存储装置 如果存在检测到异常的存储节点,则恢复冗余配置; 以及冗余配置恢复步骤,用于如果可检测到异常的存储节点中的预定单元的数据存储区域中的数据可重用,则通过停止不需要的数据的复制来恢复使用可重用数据的冗余配置和数据存储装置。 版权所有(C)2006,JPO&NCIPI
    • 8. 发明专利
    • Data management program, and multi-node storage system
    • 数据管理程序和多节点存储系统
    • JP2013008387A
    • 2013-01-10
    • JP2012197110
    • 2012-09-07
    • Fujitsu Ltd富士通株式会社
    • NOGUCHI YASUOOGIWARA KAZUTAKATAMURA MASATOSHITSUCHIYA YOSHIHIROMARUYAMA TETSUTAROWATANABE TAKASHIKUMANO TATSUOOE KAZUICHI
    • G06F3/06G06F13/10
    • PROBLEM TO BE SOLVED: To shorten a stop period of data access which occurs during the occurrence of a failure of a storage device.SOLUTION: If the operation of a storage device 1 is in a bad condition or if a lapse time (T) measured by response time calculation means 4a exceeds an operation bad condition detection time (T1), operation bad condition detection means 4b outputs operation bad condition information showing that the storage device 1 has the possibility of being in a bad condition to a control node 7. The control node 7 stores the operation bad condition information in operation bad condition information storing means 7b. When an access node 8 accesses the first slice 1a in the storage device 1 to cause an error after that, access relevant information showing that the access fails is transmitted to the control node 7. The control node 7 makes a recovery instruction of the slice 1a shown by the access relevant information to disk nodes 5, 6.
    • 要解决的问题:缩短在存储设备发生故障期间发生的数据访问的停止时段。 解决方案:如果存储装置1的操作处于不良状态,或者如果由响应时间计算装置4a测量的经过时间(T)超过操作不良状况检测时间(T1),则操作不良状况检测装置4b 输出表示存储装置1具有向控制节点7处于不良状态的可能性的操作不良状况信息。控制节点7将操作不良条件信息存储在操作不良条件信息存储装置7b中。 当访问节点8访问存储设备1中的第一片1a以引起错误之后,将访问失败的访问相关信息发送到控制节点7.控制节点7进行片1a的恢复指令 通过访问磁盘节点5,6的访问相关信息显示。版权所有:(C)2013,JPO&INPIT
    • 9. 发明专利
    • File processing program and device
    • 文件处理程序和设备
    • JP2012198786A
    • 2012-10-18
    • JP2011062832
    • 2011-03-22
    • Fujitsu Ltd富士通株式会社
    • NOGUCHI YASUOOZAWA TOSHIHIROOE KAZUICHITSUCHIYA YOSHIHIROOGIWARA KAZUTAKATAMURA MASATOSHIWATANABE TAKASHIKUMANO TATSUO
    • G06F12/00
    • PROBLEM TO BE SOLVED: To further remove redundancy from a storage device having a redundancy removal function.SOLUTION: A file (A) that is read from a first storage device is written onto a second storage device that has a redundancy removal function to avoid writing a file that has the same contents as another file onto another area. At the time of writing the file (A) onto the second storage device, data areas (410 to 440) that are separately re-writable are detected for the file, the file is separated into multiple different files for each of the data areas, and the separated multiple files are written onto the second storage device. According to such technology, existence of redundancy is not determined in units of electronic documents, but in units of its parts of which contents are separately re-writable. Therefore, the redundancy that could not be removed when existence of redundancy is determined in units of electronic documents is able to be removed.
    • 要解决的问题:进一步从具有冗余删除功能的存储设备中去除冗余。 解决方案:将从第一存储设备读取的文件(A)写入具有冗余删除功能的第二存储设备,以避免将具有与另一文件相同内容的文件写入另一区域。 在将文件(A)写入第二存储装置时,针对文件检测分开重新写入的数据区(410至440),将文件分成用于每个数据区的多个不同文件, 并将分离的多个文件写入第二存储设备。 根据这种技术,冗余的存在不是以电子文件为单位确定的,而是以其内容单独重写的单位为单位。 因此,以电子文档为单位确定冗余存在时无法消除的冗余能够被删除。 版权所有(C)2013,JPO&INPIT
    • 10. 发明专利
    • Management program, management device, and management method
    • 管理计划,管理设备和管理方法
    • JP2011233016A
    • 2011-11-17
    • JP2010104015
    • 2010-04-28
    • Fujitsu Ltd富士通株式会社
    • WATANABE TAKASHITSUCHIYA YOSHIHIRONOGUCHI YASUO
    • G06F17/30
    • PROBLEM TO BE SOLVED: To increase efficiency of search by a Bloom filter.SOLUTION: A second transposed Bloom filter row tBF(p)s, tBFa(p)s are divided per word into words W1-1 to W1-12 and W2-1 to W2-12, and the words W1-1 to W1-12 and W2-1 to W2-12 are grouped in accordance with the identical arrangement position. Specifically, words having the same # in the words W1-# and W2-# are arranged in order of the word arrangement. For example, since both of the word W1-1 and the word W2-1 have 1 at the tail, the words W1-1 and W2-1 are arranged in this order. As a result, an original first transposed Bloom filter tbf(p-1) is restored. The other words are transposed in the same way, so that a first transposed Bloom filter row new-tBF(p) is restored.
    • 要解决的问题:提高Bloom过滤器的搜索效率。 解决方案:将第二个转置的布隆过滤器行tBF(p)s,tBFa(p)s每个字划分为字W1-1至W1-12和W2-1至W2-12,并且字W1-1 根据相同的布置位置对W1-12和W2-1至W2-12进行分组。 具体而言,在字W1-#和W2-#中具有相同#的字按照排列顺序排列。 例如,由于字W1-1和字W2-1两者在尾部具有1,所以字W1-1和W2-1按此顺序排列。 结果,恢复原始的第一转置布隆过滤器tbf(p-1)。 以相同的方式转置其他单词,以便恢复第一个转置的布隆过滤器行new-tBF(p)。 版权所有(C)2012,JPO&INPIT