DB2 graphic QMF Version 8

Sample analysis output panel

The QMF Query Information panel (RAAQMFQA) is described in this section. All of the data available from Query Analyzer are shown on this panel.

Table 94. Query Information
DB2P - QMF Query Information ------------------------------------------ 8.1.0
Command ===>
 
 
Is access through an index?................................ :
Is access through more than one index?..................... :
How many indexes use matching columns?..................... :
Is the query satisfied using only the index?............... :
Was a scan limited to certain partitions?.................. :
Is prefetching performed?.................................. :
Is data accessed/processed in parallel?.................... :
Are sorts performed?....................................... :
Are joins present in the query?............................ :
When are column functions evaluated?....................... :
Cost category-Service Units (in 100s) to run query......... :
 
Historical Data For This Query
Minimum rows...... :              Minimum CPU time......... :
Maximum rows...... :              Maximum CPU time......... :
Average rows...... :              Average CPU time......... :
 
Enter Y or press PF3 to cancel this query ===>
Enter maximum number of rows to fetch.... ===>           (0=bypass)

If you press PF1, a description of these fields and the raw access path data are available for display.

The fields on the panel are:

Is access through an index?
If an access type of I, I1, N, or MX is found in the EXPLAIN output, then this field is set to YES. This indicates that there is at least one instance of index access, although other access types may also be present.
Is access through more than one index?
If an access type of M, MX, MI, or MU is found in the EXPLAIN output, then this field is set to YES. This indicates at least one instance of a multiple index access.
How many indexes use matching columns?
If index access is found in the plan and MATCHCOLS > 0, then this number is incremented by 1. This is not the number of matching columns per index nor is it the sum of the number of matching columns, but rather it is the number of index accesses in which columns in a WHERE condition matched columns of the index. This is equal to the Matching Index scans field on the Help panel.
Is the query satisfied using only the index?
If any query in the plan has an INDEXONLY column = YES, then this field will be set to YES.
Was a scan limited to certain partitions?
If PAGE_RANGE = Y for any row in the plan table, then this field will be set to YES.
Is prefetching performed?
If PREFETCH is L or S for any row of the plan table, then this field will be set to YES.
Is data accessed/processed in parallel?
If PARALLELISM_MODE is I or C for any row in the plan table, then this field is set to YES.
Are sorts performed?
If any of the SORTN_ or SORTC_ columns of the plan table are Y, then this field is set to YES. Also, if Join Method = 3, this also indicates a sort was performed.
Are joins present in the query?
If METHOD = 1,2, or 4 for any row in the plan table, then this field is set to YES.
When are column functions evaluated?
This value can be S and/or R according to COLUMN_FN_EVAL in the plan table. SQL Parsing is not performed to determine specific function type.
Cost Category and Service Units to run query
This is the estimated processor cost for this SQL statement. This value is comprised of the Cost Category and the Service Units. Both these values come from the DSN STATEMNT TABLE and are present only if the release level of DB2 is 6.1 or higher.
Cost Category
Indicates which values DB2 used to make the estimate.
  • A indicates DB2 had enough information to make the estimate without using defaults.
  • B indicates DB2 had some condition that forced it to use defaults and, as a result, estimates may not be as accurate.
Service Units
Estimated processor cost for this SQL statement in units of 100. For example, a displayed value of 21 means approximately 2100 service units (the value is truncated).
Historical data for query
This area of the panel shows the actual historical values saved from previous runs of the named object. The object name and associated statistics will be displayed only if they are available. Because object history is tracked by QMF object name, queries run with no associated object name will not be able to access historical statistics. Given the case where the object has been tracked, the information shown is the minimum, maximum, and average counts for both rows fetched and CPU utilization (the CPU time shown is in HH:MM:SS:HS format).
Query will be canceled by Query Analyzer
If the exit routine has determined that the query should be canceled, then this message will appear at the bottom of the panel.
Enter Y or press PF3 to cancel this query
If you specify Y and press Enter or press PF3, this query will not run and you will return to the previous QMF panel. If you leave this field blank and press Enter, this query will run.
Enter maximum number of rows to fetch (0=bypass)
This value determines the number of rows that QMF will fetch before the query is canceled. A value of 0 (zero) directs Query Analyzer to bypass this row limit checking. The value entered here will not override the row limit enforced by Governor.


Go to the previous page Go to the next page

Downloads | Library | Support | Support Policy | Terms of use | Feedback
Copyright IBM Corporation 1982,2004 Copyright IBM Corporation 1982, 2004
timestamp Last updated: March, 2004