原生XML存储方案的数据更新机制

Data update mechanism for native XML storage scheme

原生XML存储方案直接关系到查询处理和数据更新。目前的原生XML存储方案大多关注于查询处理而很少涉及对数据更新的支持。与关系表的更新不同,XML更新要考虑到节点的文档顺序。提出了一种新的原生XML存储更新机制,它既保持了节点的文档顺序,又使更新操作局限于一个页面之内,保证了更新的效率。通过引入前向链接记录和重定位记录,该更新机制使页面分裂时记录存储地址保持不变,避免了索引更新的I/O开销。通过实例说明该原生XML存储方案的数据更新机制是有效的。

A native XML storage scheme is directly related to query processing and data update.The current native XML storage schemes are mostly concerned with query processing and rarely involve the support of data update.Different from the update of relational tables,XML update needs to take the document order of nodes into account.A novel update mechanism for native XML storage was presented,which not only maintained the document order of nodes,but also restricted an update operation within one page to ensure the update efficiency.Through the introduction of forward link records and relocated records,the update mechanism kept record storage addresses unchanged when splitting a page to avoid the I/O overhead of index update.A case study was made to demonstrate that the data update mechanism for native XML storage scheme is effective.