Mike Cain
DB2 for i Center of Excellence
Rochester, MN USA
[email protected]
db2fori.blogspot.com
How to Boost Query Performance
Using Indexes
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
in·dex
Something that serves to guide, point out, or
otherwise facilitate reference.
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
Indexing Technology
within DB2 for i
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
IBM Training
DB2 for i
 Two types of indexing technologies are supported
– Radix Index
– Encoded Vector Index
(OmniFind Text Index will be covered later)
 Each type of index has specific uses and advantages
 Respective indexing technologies compliment each other
 Indexes can be used for statistics and implementation
 Indexes can provide RRNs and/or data
 Indexes are scanned or probed
– Probe can only occur on contiguous, leading key columns
– Scan can occur on any key column
– Probe and scan can be used together
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Using Indexes - Probe v Scan
Index Key Columns (CUSTOMER, ORDER, ITEM)
• Probe (key positioning)
with leading, n contiguous
key columns
1
1+2
1+2+3
• Scan (test)
with any other
key columns
2
3
2+3
CUSTOMER
ORDER
ITEM
001
B507
AB-2700
001
B607
CD-2000
002
B100
XY-1005
002
B102
AZ-5000
003
B709
HH-6500
004
B043
HH-6500
...WHERE ORDER = ‘B102‘ AND CUSTOMER = 002
...WHERE ITEM = ‘HH-6500’
...WHERE ITEM = ‘CD-2000’ AND CUSTOMER = 001
…WHERE ITEM = ‘HH-6500’ OR COLOR = ‘RED’
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Radix Index
 Index “tree” structure
 Key values are compressed
– Common patterns are stored once
– Unique portion stored in “leaf” pages
– Positive impact on size and depth of the index tree
 Algorithm used to find values
– Binary search
– Modified to fit the data structure
 Maintenance
– Index data is automatically spread across all available disk units
– Tree is automatically rebalanced to maintain an efficient structure
 Temporary indexes
– Considered a temporary data structure to assist the DB engine
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Radix Index
ROOT
Database Table
001
ARKANSAS
002
MISSISSIPPI
003
MISSOURI
004
IOWA
005
ARIZONA
…
…
Test
Node
AR
IZONA
005
ADVANTAGES:
 Very fast access to a single key
value
 Also fast for small, selected range
of key values (low cardinality)
 Provides order
MISS
IOWA
004
ISSIPPI
002
OURI
003
KANSAS
001
DISADVANTAGES:
 Table rows retrieved in order of key
values (not physical order) which equates
to random I/O’s
 No way to predict which physical index
pages are next when traversing the index
for large number of key values
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Index Probe Example
EMPLOYEE Table
Given an index on table EMPLOYEE keyed on STATE...
SELECT *
FROM EMPLOYEE
WHERE STATE = ‘IOWA’
EMPLOYEE Index
STATE
Perform a probe into
the range using the
local selection
value(s)
RRN
...
IOWA (004)
IOWA (007)
IOWA (010)
IOWA (017)
KANSAS (011)
MISSISSIPPI (002)
MISSISSIPPI (013)
MISSOURI (003)
...
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
RRN
STATE
001
002
003
004
005
006
007
008
009
010
011
012
013
014
015
016
017
ARKANSAS
MISSISSIPPI
MISSOURI
IOWA
ARIZONA
MONTANA
IOWA
NEBRASKA
NEBRASKA
IOWA
KANSAS
WISCONSIN
MISSISSIPPI
WISCONSIN
WISCONSIN
ARKANSAS
IOWA
© 2012 IBM Corporation
IBM Training
Encoded Vector Index (EVI)
 Index for delivering fast data access in analytical and
reporting environments
– Advanced technology from IBM Research
– Used to produce dynamic bitmaps and RRN lists
– Fast access to statistics to improve query optimizer decision making
 Not a “tree” structure
 Can only be created through an SQL interface or System i
Navigator GUI
CREATE ENCODED VECTOR INDEX
SchemaName/IndexName ON SchemaName/TableName
(ColumnName)
INCLUDE ( SUM(ColumnName));
New in 7.1
Maintained aggregate
to support index only access
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Encoded Vector Index (EVI)
Symbol Table
Key Value
Code
Count
Include
Include
Sum()
Sum()
Arizona
1
5000
1500
2005
Arkansas
2
7300
3200
450
…
Wisconsin
49
340
575
1200
Wyoming
50
2760
210
0
optional
Vector
RRN
1
1
17
2
5
3
9
4
2
5
7
6
50
7
49
8
5
9
…
…
 Symbol table contains information for each distinct key value
- Each key value is assigned a unique code (key compression)
- Code is 1, 2, or 4 bytes depending on number of distinct key values
 Rather then a bit array for each distinct key value, use one array of codes
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Bitmap / RRN List Example
Given an EVI on table EMPLOYEE keyed on STATE...
Set bits in bitmap
or
return RRN list
...WHERE STATE = ‘ILLINOIS’
Vector
Symbol Table
Scan or
binary
search
symbol
table
for
key(s)
and
code(s)
Key
ARIZONA
ARKANSAS
CALIFORNIA
COLORADO
ILLINOIS
IOWA
KANSAS
MISSISSIPPI
...
Code
1
2
3
4
5
6
7
8
...
Scan
vector
for
code(s)
1
17
5
9
2
7
50
49
5
.
.
.
.
.
.
.
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
Bitmap
0
0
1
0
0
0
0
0
1
.
.
.
.
.
.
.
Row
1
2
3
4
5
6
7
8
9
...
© 2012 IBM Corporation
IBM Training
Index ANDing Example
State
Workdept
EVI
EVI
Intermediate
RRN list
3
5
10
15
1000 AND
1005
(Merge)
1007
3001
3050
State
SELECT *
FROM EMPLOYEE
WHERE STATE = ‘MINNESOTA'
AND WORKDEPT IN ( 'B01', C01, 'E01')
Intermediate
RRN list
Final
RRN list
3
7
10
27
1000
1010
2035
3001
4100
3
10
1000
3001
Represents
all the local
selection
Workdept
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Index ORing Example
State
Workdept
EVI
Radix
Intermediate
RRN list
4
1001
1005 OR
3051 (Merge)
State
SELECT *
FROM EMPLOYEE
WHERE STATE = ‘IOWA'
OR WORKDEPT IN ( 'B01', C01, 'E01')
Intermediate
RRN list
Final
RRN list
2
8
1004
4105
2
4
8
1001
1004
1005
3051
4105
Represents
all the local
selection
Workdept
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
EVI Symbol Table Only Example
Given an EVI on table EMPLOYEE keyed on STATE...
SELECT COUNT(*)
FROM EMPLOYEE
WHERE STATE = ‘Wisconsin’;
SELECT COUNT(DISTINCT STATE) FROM EMPLOYEE;
SELECT STATE, SUM(), SUM()
FROM EMPLOYEE
GROUP BY STATE;
Symbol Table
Key Value
Search / Scan
symbol
table
for
key(s)
and counts
Code
Count
Arizona
1
Arkansas
Include Include
Sum()
Sum()
5000
1500
2005
2
7300
3200
450
Wisconsin
49
340
575
1200
Wyoming
50
2760
210
0
…
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
DB2 for i
cardinality
The number of distinct elements in a set.
• High cardinality = large distinct number of values
• Low cardinality = small distinct number of values
selectivity
The number of elements matching the criteria.
• High selectivity = small number of rows match
• Low selectivity = large number of rows match
In general…
• A radix index is best when choosing a small set of rows and the key
cardinality is high
• An encoded vector index is best when choosing a set of rows and the key
cardinality is low
• Understanding the data and query are key
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Transactions table
Transaction
Number
Transaction
Type
100 million rows
3 transaction types
Cardinality
and
Selectivity
SELECT *
FROM TRANSACTIONS
WHERE TRANSACTION_NUMBER = 1;
SELECT *
FROM TRANSACTIONS
WHERE TRANSACTION_TYPE = ‘D’
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Creating SQL Indexes
 CREATE INDEX SQL statement
CREATE INDEX MY_IX on MY_TABLE (KEY1, KEY2)
 CREATE ENCODED VECTOR INDEX SQL statement
CREATE ENCODED VECTOR INDEX MY_EVI on MY_TABLE (KEY1)
 System i Navigator – Database graphical interface
 Primary Key, Foreign Key and Unique Key Constraints
– CREATE TABLE
– ALTER TABLE
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Creating Indexes
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Creating Indexes
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Creating Indexes
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Creating Indexes
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Creating Indexes
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Creation of Index with Derived Keys
 Creation of indexes with derived keys via SQL
CREATE INDEX ORDERPRIORITYUPPER ON T1
(UPPER(ORDERPRIORITY) AS UORDERPRIORITY ASC);
CREATE ENCODED VECTOR INDEX YEARQTR ON T1
(YEAR(ORDERDATE) AS ORDYEAR ASC,
QUARTER(ORDERDATE) AS ORDQTR ASC);
CREATE INDEX TOTALEXTENDEDPRICE ON T1
(QUANTITY * EXTENDEDPRICE AS TOTEXTPRICE ASC);
NOTE: There are some restrictions on which indexes can be used by the optimizer
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Create Sparse Indexes from SQL
 Support of WHERE clause on SQL create index
CREATE INDEX FASTDELIVER ON T1
(SHIPMODE ASC)
WHERE SHIPMODE = 'NEXTDAYAIR‘
OR SHIPMODE = 'COURIER';
NOTE: Index is NOT used by the query optimizer (CQE and SQE) prior to 7.1
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
When to a use Derived Index?
 Derived indexes may be useful for
– Case insensitive searches (i.e UPPER(COL1) = ‘ABC’
– Data extracted from a column (i.e. SUBSTR(), YEAR(), MONTH()…)
– Results of operations (i.e. COL1+COL2 , QTY * COST)
– Might be useful to allow index only access in more cases
– Reduce table scans and temporary data structures
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
IBM OmniFind Text Search Server for DB2 for i
 IBM i licensed product: 5733-OMF
– No-charge offering
– Requires IBM i 6.1 or 7.1
 Delivers common DB2 Family text search technology
– Advanced, linguistic high-speed textual searches
– Support enabled for any character-based column
– Search technology also supports Rich Text document formats
• Example: LOB columns containing PDF or Microsoft® Word
documents
• IFS documents can be indexed with extra programming
– Includes support for 26 different languages
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
OmniFind Database Requirements
 Some of the supported document format types:
– Plain text
– XML
– HTML
– Adobe PDF
– Rich Text Format (RTF)
– JustSystems Ichitaro
– Microsoft Excel
– Microsoft PowerPoint
– Microsoft Word
– Lotus® 123
– Lotus Freelance®
– Lotus WordPro
– Lotus Symphony
– OpenOffice 1.1 & 2.0
– OpenOffice Calc
– Quattro Pro
– StarOffice Calc
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Text Index vs DB2 Index
 Text indexes stored outside of DB2 in IFS
 Text indexes have delayed maintenance
– Changes logged to staging table
– Index maintenance is scheduled
 Text indexes not protected by IBM i SMAPP
 Text indexes utilize different indexing methods
– Table with VARCHAR(32000) column and 175,000 rows
• DB2 Index object size: 1.7 GB
• Text Index object size: 0.1 GB
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Interfaces
 A set of administrative stored procedures for starting and ending
the text search server.
• SYSTS_START
• SYSTS_STOP
 A set of administrative stored procedures to create, drop, and
maintain the text search index.
• SYSTS_CREATE
• SYSTS_DROP,
• SYSTS_UPDATE
 Two built in functions to query the index
• CONTAINS
• SCORE
 System i Navigator 7.1
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Built in DB2 Text Search functions
 CONTAINS
– Searches a text search index using the criteria that are
specified in a search argument and results about whether
or not a match was found
 CONTAINS(column-name, search-argument, options)
– column-name is the column over which the text search
server is built
– search-argument is the term or phrase to search for
– options is an optional parameter that can be used to modify
the query language or turn synonym matching on
– If the term or phrase is found, returns 1 (i.e. true)
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
CONTAINS example
 Find the exact string “IBM” in the COMMENT column, using a
host variable
 Assume a text search index is built over column COMMENT
EXEC SQL DECLARE C1 CURSOR FOR
SELECT CUSTKEY
FROM CUSTOMERS
WHERE CONTAINS(COMMENT, :search_arg) = 1
ORDER BY CUSTKEY;
EXEC SQL SET :search_arg = ‘”IBM”';
EXEC SQL OPEN C1;
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Built in search functions
 SCORE
– The SCORE function searches a text search index using
criteria that are specified in a search argument and returns a
relevance score that measures how well a document matches
the search argument
 SCORE(column-name, search-argument, options)
– column-name is the column over which the text search server
is built
– search-argument is the term or phrase to search for
– options is an optional parameter that can be used to modify the
query language or turn synonym matching on
– The result of SCORE is a value between 0 and 1 (i.e. decimal)
– A higher number indicates that a document contains a
match for the search argument more frequently
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Score Example
 The following statement generates a list of employees in the order of how
well their resumes match the query "programmer AND (java OR cobol)",
along with a relevance value that is normalized between 0 (zero) and 100
SELECT EMPNO,
INTEGER(SCORE(RESUME, 'programmer AND (java OR cobol)') * 100))
AS RELEVANCE
FROM
EMP_RESUME
WHERE RESUME_FORMAT = ‘ASCII'
AND
CONTAINS(RESUME, 'programmer AND (java OR cobol)') = 1
ORDER BY RELEVANCE DESC
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
XML
IBM Training
 Given a table with email messages store in xml format, within a
VARCHAR column EMAIL
 The text index can be created with:
CALL sysproc.systs_create('MYSCHEMA',
'XMLI',
’MYSCHEMA.MY_TABLE(EMAIL)',
'FORMAT XML')
 The “FORMAT XML” clause tells OmniFind that the column
should be parsed and indexed as XML data
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
XML
IBM Training
 Suppose the XML data being searched is in a format of:
<?xml version="1.0" encoding="ISO-8859-1"?>
<note>
<to>Nick</to>
<from>Tiffany</from>
<heading>OmniFind for DB2i</heading>
<body>
OmniFind is wonderful for working with XML data!
</body>
</note>
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Xpath in CONTAINS clause
 Use the CONTAINS function to search for documents that
have “Nick” in the “to” section of the xml document
SELECT EMAIL
FROM MYCHEMA.MYTABLE
WHERE CONTAINS(EMAIL, @xpath:''/note//to[.contains("Nick")]''') = 1
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
Query Optimization
(using indexes)
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
IBM Training
Data Access Methods
Cost based optimization dictates that the fastest access method
for a given table will vary based upon selectivity of the query
High
Response
Time
Table Scan
Clustered,
Skip Seq
Probe
Low
Few
Many
Number of rows searched / accessed
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Strategy for Query Optimization
Optimizing indexes will generally follow this simplified strategy:
?
Gather list of indexes for statistics and costing implementation methods
Consider the indexes based on how the index can be used
Local selection
Joining
Grouping
Ordering
Index only access
One index may be useful for statistics, and another useful for implementation
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Query Optimization Feedback
Visual
Explain
SQE Plan
Cache
Indexes
Advised
SQE Plan
Cache
Snapshots
SQL
Monitor Data
SQL request
Query
Optimization
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Indexing Advice from the Optimizer
 Both CQE and SQE provide index creation advice
– QSYS2.SYSIXADV
– System i Navigator
 CQE
–
–
–
–
–
Basic advice
Radix index only
Based on table scan and local selection columns only
Temporary index creation information also provides insight
CQE Visual Explain will try and tie pieces together to advice a better index
 SQE
–
–
–
–
–
Robust advice
Radix and EVI indexes
Based on all parts of the query
Multiple indexes can be advised for the same query
Some limitations
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Indexing Advice from the Optimizer…
Local selection, Join, Group By, Order By
Equals first, followed by one inequality (for probing)
No ORed predicates involving different columns
No derived keys
No EVI include columns
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Autonomic Index Creation
 Optimizer can have the DB Engine create a temporary index
 Both full and sparse indexes can be created
 Temporary indexes are not used for statistics (unless running 7.1 + PTFs)
 Temporary indexes are maintained
 CQE
– Temporary indexes are not reused and not shared
– Usually a bottleneck in query performance
– Can impact overall system performance
– Can increase the amount of temporary storage used
 SQE
– Temporary indexes are reused and shared across jobs and queries
– Creation is based on “watching” the query requests over time
– Creation is based on optimizer’s own index advice
– Temporary index maintenance is delayed when all associated cursors closed
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Index Evaluator via Catalog Views
select *
from qsys2.sysindexstat;
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Index Evaluator via Catalog Views
SYSINDEXSTAT
Contains one row for every SQL index. Use this
view when you want to see information for a
specific SQL index or set of SQL indexes. The
information is similar to that returned via Show
Indexes in System i Navigator.
SYSPARTITIONINDEXES
Contains one row for every index built over a table
partition or table member. Use this view when you
want to see index information for indexes built on a
specified table or set of tables. The information is
similar to that returned via Show Indexes in System
i Navigator.
SYSTABLEINDEXSTAT
Contains one row for every index that has at least
one partition or member built over a table. If the
index is over more than one partition or member,
the statistics include all those partitions and
members.
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Indexing Strategies
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Create all advised indexes?
Nothing, let the system handle it?
Monitor, analyze, and tune important tables and queries?
Get some help?
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
DB2 for i
The goals of creating indexes are:
1. Provide the optimizer the statistics needed to understand
the data, based on the query
2. Provide the optimizer implementation choices, based on
the selectivity of the query
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
The Process of Identifying Indexes
Proactive method
• Analyze the data model, application and SQL requests
• Database PK, UK, RI constraints are helpful
Reactive method
• Rely on optimizer feedback and actual implementation methods
• Rely on SQE’s ability to auto tune using temporary indexes
Understand the data being queried
• Column selectivity
• Column cardinality
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Indexing Strategy - Basic Approach
Radix Indexes
• Common local selection columns
Minimum
• Join columns
• Local selection columns + join columns
• Local selection columns + grouping columns
• Local selection columns + ordering columns
Advanced
Encoded Vector Indexes
• Local selection column for index ANDing/ORing
• Join columns (star or snowflake schema)
• Index only access
Requires
knowledge of
query
optimization
and data
lifecycle
•DISTINCT, COUNT, COUNT DISTINCT, SUM()
Note: Columns used with equal conditions are first in key list
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Indexing Strategy – React to Implementatoin
If the optimizer feedback indicates:
Full table scan
 Create an index on local selection columns
Temporary index
 Create an index on join columns
 Create an index on grouping columns
 Create an index on ordering columns
Hash table
 Create an index on join columns
 Create an index on grouping columns
“Perfect”, multiple key column radix indexes are usually best
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Analyzing the Data Model
CUSTOMERS
custkey
ORDERS
orderkey
custkey
DATES
partkey
datekey
orderdate
PARTS
shipdate
partkey
suppkey
SUPPLIERS
suppkey
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
Examples
select
from
where
*
orders
shipdate = ‘2010-10-01’;
select
from
where
and
*
orders
orderdate = ‘2010-10-01’
quantity = 0;
select
from
inner join
on
where
*
orders o
customers c
o.custkey = c.custkey
c.customer_number = 102;
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation
IBM Training
COMMON Sweden 2013 / Copyright 2013 IBM Corporation
© 2012 IBM Corporation