Using GROUP to improve Save data set performance
If you never require separate report entries for one or more OMEGAMON® XE for Db2 PE identifiers, you can use GROUP to reduce the uniqueness of the key in the Save data. This can result in fewer reduced records in the Save data set, a smaller Save data set, and improved performance.
The standard key for Save data sets contains a combination of the applicable OMEGAMON XE for Db2 PE identifiers from Db2 instrumentation records. For example, the key for an Accounting save record contains a number of OMEGAMON XE for DB2 PE identifiers. Because of the uniqueness of the standard key, there can be a very large number of reduced records in a Save data set.
If you never require separate report entries for one or more OMEGAMON XE for Db2 PE identifiers, for example, if you only produce Accounting reports by using ORDER(CONNECT-PRIMAUTH-PLANNAME), you can use groups in INCLUDE REDUCE to obtain the minimum number of records in the Save data set.
ACCOUNTING
REDUCE
SAVE
GROUP (CORRNAME (ALLCNM(*)))
GROUP (CORRNMBR (ALLCNU(*)))
GROUP (ORIGAUTH (ALLORI(*)))
GROUP (REQLOC (ALLREQ(*)))
ACCOUNTING
REDUCE
INCLUDE (CORRNAME (G(ALLCNM)),
CORRNMBR (G(ALLCNU)),
ORIGAUTH (G(ALLORI)),
REQLOC (G(ALLREQ))),
SAVE
The Save data set produced by this command stream contains reduced records for each unique value of the OMEGAMON for Db2 Performance Expert identifiers that are not grouped, for example, PRIMAUTH and PLANNAME.
During REDUCE processing, the group name is substituted for the original value of the key entry. For example, all records in the reduced data have a correlation name of ALLCNM.
Because the key now contains the group name (not the original value), you cannot use the original values of any of the grouped OMEGAMON XE for DB2 PE identifiers in REPORT or INCLUDE/EXCLUDE in the same command stream or when you restore the data.