如果查询中有属于多列索引的连接列或过滤列，则使用UPDATESTATISTICSHIGH。
Use UPDATE STATISTICS HIGH if queries have a multi-column indexed defined on join columns or filter columns.
该连接的执行过程包括扫描散列连接中的内表CUSTOMERS，并通过散列连接列的值生成一个查找表。
The join is performed by scanning the inner table of the hash join, CUSTOMERS, and generating a lookup table by hashing the join column values.
然后，它使用SHIP操作符读外表BAD_CREDIT，散列连接列的值，并检查为内表生成的查找表。
IT then reads from the outer table BAD_CREDIT using the SHIP operator, hashing the join column values, and checking in the lookup table generated for the inner table.
在检查这些表时，我们发现两个没有索引的连接列：customer .custid和order . custid。
As we examined the tables, we found that two joining columns, customer.custid and order.custid, do not have indexes.
仅当查询中有非索引连接列或过滤列时，才使用UPDATESTATISTICSMEDIUMDISTRIBUTIONS。
Use UPDATE STATISTICS MEDIUM DISTRIBUTIONS ONLY if queries have non-indexed join columns or filter columns.
随后的查询指定了clients表中的address列，从中检索地址信息，并将其作为连接的邮寄地址。
A subsequent query specifying the address column in the CLIENTS table retrieves the address information as a concatenated mailing address.
您将找到的大部分结果都是哈希索引站点——有意地连接散列值和其相应关键字的站点。
Most of the results you'll find are hash indexing sites — sites intentionally built to connect hash values and their corresponding keys.
图9展示了一个IDS连接的列视图的屏幕快照。
Figure 9 shows a screenshot of the columns view for an IDS connection.
在显示的向导中，选择将用于数据源的连接、数据对象和列。
In the wizard that is displayed, select the connection, the data objects, and the columns to be used for the data source.
您应该始终在逻辑层面上记住KNA1表上的ktokd列，它连接检查表t 077d，而不是其文本表t 077x。
You should still keep in mind that on the logical level the KTOKD column in the KNA1 table is connected to the check table T077D, not its text table T077X.
日期转换可以使用连接的Excel功能以及列引用来完成。
This conversion of the date can done using concatenate excel function along with column references.
点击id列中的连接(如图33所示)您就会看到选择应用程序中的细节信息。
Click the link in the Id column (as shown in Figure 33) and you will see the details of the selected application.
对于被合并的表的分区键内的每列，必须存在一个同等连接的谓语。
For each column in the partitioning key of the joined tables, an equijoin predicate must exist.
另外，DB 2无法在排序合并合并连接中使用随机顺序的索引列。
Also, DB2 cannot use random order index columns as part of a sort merge join.
注意，执行计划表明，每个散列连接的左边分支将返回大约800,000行，这是一个相当大的数字。
Note that the execution plan indicates that the left leg of each hash join will return about 800,000 rows, a significant number.
图8展示了如果规则使用了多个参考列，那么左外部连接就进行级联。
Figure 8 shows that the left outer join operations are cascaded if a rule USES more than one reference column.
然后，重写阶段找出source列之间无效的连接。
The rewrite phase then identifies unproductive joins between the SOURCE columns.
数据库是一堆互相连接的行和列，可以容易地进行备份、存储、查询、链接，可是，来自于文档的“数据”有着更多的信息。
You can put it in a database that can be backed up, stored, queried and linked. But the “data” comes from a document that housed much more information than you have in a database.
散列方式允许客户机连接尝试永远被路由到相同的服务器。
Hashing allows client connect attempts to always be routed to the same server.
分区键应该包括最常连接的列。
The partitioning key should include the most frequently joined columns.
利用下推连接技术的星型连接计划使得散列连接更加有效的主要原因是，在初期减少了事实表(探测表)的行数。
A star join plan utilizing push-down hash join technology makes hash joins more effective primarily due to reduction of fact table (probe table) rows early on.
它选择一个散列连接，以PART作为内部表，这个计划只需运行299秒。
It selects a hash join with PART as the inner table which runs in 299 seconds.
事实表和任何维度表之间的连接都是等值连接，这使得散列连接成为可能。
The join between the fact table and any dimension table is an equi-join, which makes a hash join possible.
确保查询中进行连接操作的列都有索引。
Ensure that columns that are joined in queries have indexes.
图4展示了具有绑定的规则，它涉及来自多个表的列，这导致了在连接操作之后进行抽样。
Figure 4 shows a rule with bindings that involve columns coming from more than one table, which causes the sampling to be done after the join operation.
星型连接计划利用下推散列连接技术。
这个查询对每个事实表和维度表之间的连接使用散列连接。
This query plan uses hash join for each join between the fact table and a dimension table.
那些连接信息也被列在表格1中。
The connection information for those is also listed in Table 1.
连接器将根据目标表和列定义在运行时生成查询s QL。
The connector generates the query SQL at runtime based on the target table and the column definitions.
事实表上的动态过滤减少了实际散列连接探测阶段使用的行数。
This dynamic filter on the fact table reduces the number of rows used in the probe phase of the actual hash joins.
应用推荐