74 lines
4.3 KiB
HTML
74 lines
4.3 KiB
HTML
<?xml version="1.0" encoding="UTF-8"?>
|
|
<!DOCTYPE html
|
|
PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
|
|
<html lang="en-us" xml:lang="en-us">
|
|
<head>
|
|
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
|
|
<meta name="security" content="public" />
|
|
<meta name="Robots" content="index,follow" />
|
|
<meta http-equiv="PICS-Label" content='(PICS-1.1 "http://www.icra.org/ratingsv02.html" l gen true r (cz 1 lz 1 nz 1 oz 1 vz 1) "http://www.rsac.org/ratingsv01.html" l gen true r (n 0 s 0 v 0 l 0) "http://www.classify.org/safesurf/" l gen true r (SS~~000 1))' />
|
|
<meta name="DC.Type" content="reference" />
|
|
<meta name="DC.Title" content="Two-step grouping with DB2 Multisystem" />
|
|
<meta name="abstract" content="If the partitioning key is not included in the grouping fields, then grouping must be done using two-step grouping, because the like values of a field are not located on the same node." />
|
|
<meta name="description" content="If the partitioning key is not included in the grouping fields, then grouping must be done using two-step grouping, because the like values of a field are not located on the same node." />
|
|
<meta name="DC.subject" content="grouping, two-step, optimization, two-step grouping" />
|
|
<meta name="keywords" content="grouping, two-step, optimization, two-step grouping" />
|
|
<meta name="DC.Relation" scheme="URI" content="grouping.htm" />
|
|
<meta name="copyright" content="(C) Copyright IBM Corporation 1998, 2006" />
|
|
<meta name="DC.Rights.Owner" content="(C) Copyright IBM Corporation 1998, 2006" />
|
|
<meta name="DC.Format" content="XHTML" />
|
|
<meta name="DC.Identifier" content="twostep" />
|
|
<meta name="DC.Language" content="en-us" />
|
|
<!-- All rights reserved. Licensed Materials Property of IBM -->
|
|
<!-- US Government Users Restricted Rights -->
|
|
<!-- Use, duplication or disclosure restricted by -->
|
|
<!-- GSA ADP Schedule Contract with IBM Corp. -->
|
|
<link rel="stylesheet" type="text/css" href="./ibmdita.css" />
|
|
<link rel="stylesheet" type="text/css" href="./ic.css" />
|
|
<title>Two-step grouping with DB2 Multisystem</title>
|
|
</head>
|
|
<body id="twostep"><a name="twostep"><!-- --></a>
|
|
<!-- Java sync-link --><script language="Javascript" src="../rzahg/synch.js" type="text/javascript"></script>
|
|
<h1 class="topictitle1">Two-step grouping with DB2<sup>®</sup> Multisystem</h1>
|
|
<div><p>If the partitioning key is not included in the grouping fields,
|
|
then grouping must be done using two-step grouping, because the like values
|
|
of a field are not located on the same node.</p>
|
|
<div class="section"><p>The following code is an example of two-step grouping.
|
|
</p>
|
|
<p>SQL statement:</p>
|
|
<pre> SELECT JOB, AVG(SALARY)
|
|
FROM EMPLOYEE
|
|
GROUP BY JOB
|
|
</pre>
|
|
<p>OPNQRYF command:</p>
|
|
|
|
<pre> OPNQRYF FILE((EMPLOYEE)) FORMAT(GRPFMT2)
|
|
GRPFLD(JOB)
|
|
MAPFLD((AVGSAL '%AVG(SALARY)'))</pre>
|
|
<p>In this example, note that for the group where JOB is <tt>Clerk</tt>,
|
|
the value <tt>Clerk</tt> is on two different nodes in the EMPLOYEE distributed
|
|
file. Grouping is implemented by first running grouping in parallel on all
|
|
three nodes. This results in an initial grouping which is placed in a temporary
|
|
file at the coordinator node. The query is modified, and the grouping is run
|
|
again at the coordinator node to get the final set of grouping results.</p>
|
|
<p>Whole-file
|
|
grouping (no group by fields) is always implemented using two steps.</p>
|
|
<p>If
|
|
the query contains either a HAVING clause or the group selection expression
|
|
(GRPSLT) parameter on the OPNQRYF command, all groups from the first grouping
|
|
step are returned to the coordinator node. The HAVING clause or the GRPSLT
|
|
parameter is then processed as part of second grouping step.</p>
|
|
<p>If the
|
|
query contains a DISTINCT column (aggregate) function and two-step grouping
|
|
is required, no grouping is done in the first step. Instead, all records are
|
|
returned to the coordinator node, and all grouping is run at the coordinator
|
|
node as part of the second step.</p>
|
|
</div>
|
|
</div>
|
|
<div>
|
|
<div class="familylinks">
|
|
<div class="parentlink"><strong>Parent topic:</strong> <a href="grouping.htm" title="The implementation method for grouping in queries that use distributed files is dependent on whether the partitioning key is included in the grouping criteria.">Implementation and optimization of grouping with DB2 Multisystem</a></div>
|
|
</div>
|
|
</div>
|
|
</body>
|
|
</html> |