每个索引分区仅从相应的数据分区取得索引行。
Each index partition indexes rows only from the corresponding data partition.
在我们的例子中，通过索引扫描得到6574个候选行。
在这种情况下，任何索引的第一行，不管简单或复杂，都是满足星型联接的一个候选者。
In this case, the first column of any index, simple or compound, is a candidate to satisfy the star join.
接下来，创建一个查询以生成要被索引的行。
然后，DB 2可以根据预期顺序，也就是索引顺序，向程序返回符合条件的行。
DB2 can then return qualified rows to the program in the desired order, the order of the index.
如果从一个带有10个索引的表删除一行，DB 2必须删除与被删除的行关联的所有索引条目。
Delete a row from a table with 10 indexes, and DB2 must remove all index entries associated with the deleted rows.
我们不是很乐意使用DPSI，但是我们必须跨多个分区探测多个索引树来寻找行。
We are not so pleased with the DPSI if we must probe multiple index trees to find rows across multiple partitions.
缓冲池是内存中的一块存储区域，用于临时读入和更改数据库页(包含表行或索引项)。
A buffer pool is an area of storage in memory into which database pages (containing table rows or index entries) are temporarily read and changed.
下面浏览一下表1中的行，来研究每个谓词的索引合格性。
Now step through the rows in Table 1 to explore index eligibility for each predicate.
IDS允许应用程序开发人员将锁放在不同的对象上，如数据库、表、页或行、以及索引。
IDS allows application developers to place locks on different objects, like databases, tables, pages or rows, and indexes.
优化程序采用系统目录，说明那里有多少行，以及构造了较好的索引来计算所采用的查询路径。
The optimizer uses system catalogs indicating how many rows there are and how well the indexes are formed to figure out what query path to use.
相关子查询将极大地降低数据检索的速度，对于拥有几百万个行但没有索引的大型表的非选择性查询更是如此。
Correlated subqueries will greatly degrade the speed of data retrieval, especially for non-selective queries on huge tables with millions of rows and with no index.
数据库服务器首先通过查找索引中的行，定位表中的相关行。
The database server locates relevant rows in the table by first finding the row in the index.
将需要读取的索引页和数据行降到最低。
Minimize the number of index pages and data rows that have to be read.
如果只需要索引中包含的值，而不需要读取这些行。
If the values contained in the index are all that is required, the rows are not read.
这些锁可以是表级锁、行级锁、索引锁等等。
These locks could be table level, row level, or lock on index, and so on.
确定了此索引后，也就知道了项的数量，也就能够确定小计行之后元素的索引了。
When you have determined this index, you then know the number of items, and you can determine the indexes of the elements occurring after the subtotal line as offsets.
每一个行是一个表或索引上的runstats命令。
JDBC适配器使用 object_key作为对客户表的索引，以检索更改的行。
The JDBC adapter uses the object_key as an index into the Customer table to retrieve the changed row.
第4行对索引进行清理，清理过程中将把所有标记为删除的文档物理删除。
The fourth line cleans up the index. IndexWriter physically deletes from the disk the documents that have been marked as deleted.
这个号码将用于作为分区映射的一个索引，以选择那一行所在的数据库分区。
That number is used as an index into the partitioning map to select the database partition for that row.
然而，如果每个不同的索引值都有许多重复的行，而所关联的表是高度群集的，那么通过索引添加的连接成本可能很低。
However, if there are many duplicate rows per distinct index value, and the associated table is highly clustered, the added costs of joining through the index can be slight.
它将查询数据库，为结果的每行中的每列建立索引，并且将每个索引条目绑定到行的主键上。
It queries your database, indexes each column in each row of the result, and ties each index entry to the row's primary key.
如果该子句中所有的列都是以用单个索引要求的顺序出现的，那么数据库服务器就使用索引来按顺序读取行，因而避免了排序。
If all the columns in the clause appear in the required sequence with a single index, the database server USES the index to read the rows in the ordered sequence, thus avoiding a sort.
基本上，它会权衡使用索引来遍历选择的行或者对表进行扫描这两种方式。
Basically, it's weighing the choice of using the index to walk through the selected rows or doing a table scan.
大多数情况下，对于通过索引定位的每一行，都必须进行单独的磁盘访问来读取该页。
In many cases, a separate disk access must be made to fetch the page for each row located through an index.
如果没有索引满足所取的行的要求顺序，或者优化器断定排序的代价低于索引扫描，那么就需要进行排序。
Sorting is required when no index satisfies the requested ordering of fetched rows, or the optimizer determines that a sort is less expensive than an index scan.
索引可以减少DBMS在执行查询时检查的行(或元组)数量，从而获得性能增益。
Performance benefits are realized when an index reduces the number of rows (or tuples) examined by the DBMS during query execution.
因此，在利用本地索引进行访问时，那些行与查询处理无关的分区仍被锁定。
Therefore, some partitions that do not contribute any rows for query processing are still locked when accessed via local index.
例如，您可以通过把可以仅从索引就确定需要的行的谓词放到查询中来将行读取降到最低。
For example, you can minimize the rows read by having predicates in the query that can determine the needed rows from just the index alone.
应用推荐