Algorithm to find LOB value in a relational table after key columns have been modified
    1.
    发明授权
    Algorithm to find LOB value in a relational table after key columns have been modified 有权
    在关键列修改后,在关系表中查找LOB值的算法

    公开(公告)号:US07203712B2

    公开(公告)日:2007-04-10

    申请号:US10789625

    申请日:2004-02-26

    IPC分类号: G06F17/30 G06F12/00

    摘要: A method and system for asynchronously replicating data changes in a database system captures update information for a data update in a source table from a database log, where the data value from the data update is not in the database log. A query is built to obtain the data value from the source table using the update information, including a key column value that identifies the row of the source table in which the data resides. When the key column is also updated in the same or different transaction, the data update information is stored, and the updated key column value is replaced with a later captured key column update value. The query is built and executed, and the data update is replicated at a target table. In this manner, data objects whose values are not recorded in a database log can be asynchronously replicated in a database system.

    摘要翻译: 用于在数据库系统中异步复制数据更改的方法和系统从数据库日志中捕获源表中的数据更新的更新信息,其中来自数据更新的数据值不在数据库日志中。 构建查询以使用更新信息从源表获取数据值,包括标识数据所在的源表的行的键列值。 当密钥列也在相同或不同的事务中被更新时,存储数据更新信息,并且更新的密钥列值被替换为稍后捕获的密钥列更新值。 构建并执行查询,并在目标表中复制数据更新。 以这种方式,可以在数据库系统中异步复制其值未记录在数据库日志中的数据对象。

    Fault tolerant mechanism to handle initial load of replicated object in live system
    3.
    发明授权
    Fault tolerant mechanism to handle initial load of replicated object in live system 有权
    在系统中处理复制对象的初始负载的容错机制

    公开(公告)号:US07330860B2

    公开(公告)日:2008-02-12

    申请号:US10789326

    申请日:2004-02-26

    IPC分类号: G06F12/00 G06F17/30

    摘要: An automatic initial data load to a new table copy concurrently with active replication to existing table copies in an asynchronous replication group includes: initiating a data load of the new table copy; creating spill queue(s) at the new table copy; loading the data from a source table copy to the new table copy, where changes for the active table copies are applied during the loading, but changes to the new table copy are stored in a spill storage area; applying the changes in the spill storage area to the new table copy after the loading of the data is done; and removing the spill storage area after the changes have been applied to the new table copy. Thus, a new table copy is initialized into the replication group without requiring that replication be suspended at the source or any of the existing table copies during the initializing process.

    摘要翻译: 在异步复制组中,将活动复制与现有表副本并发的新表复制的自动初始数据加载包括:启动新表副本的数据加载; 在新表副本上创建溢出队列; 将数据从源表副本加载到新的表副本,其中在加载期间应用活动表副本的更改,但对新表副本的更改将存储在溢出存储区中; 在完成数据的加载之后,将溢出存储区域中的更改应用于新的表格副本; 并且在将更改应用于新的表格副本之后移除溢出存储区域。 因此,将新的表副本初始化为复制组,而不需要在初始化过程中将源复制或任何现有表复制暂停。