The +SYSINORDER parameter controls the order in which SYSIN statements are processed. You can use this parameter in
BMC AMI Load
only in conjunction with
BMC AMI Unload
.
The values are defined as follows:
| (Default) BMC AMI Utilities processing determines the order in which SYSIN statements are processed. |
|---|
| BMC AMI Utilities processing follows the specified order for processing SYSIN statements. |
|---|
Always use +SYSINORDER(RANDOM) unless you have a specific job step that needs to do multiple things specifically in SYSIN order, such as a LOAD followed by a REORG. You have to perform these actions in separate steps. If you are required to perform the steps in a particular order and there is a need to have them in one step, other +parms should be reviewed and overridden or both for that step, for example ONERROR.
Using +SYSINORDER(HONOR) can cause performance degradation and limits parallelism.
Example
Consider the following unload statements in a SYSIN part of a job file:
//SYSIN DD *
LOAD DATA INTO TABLE USR001.EXPENSES
LOAD DATA INTO TABLE USR001.PAYROLL
LOAD DATA INTO TABLE USR001.GEN_LEDGER
UNLOAD FROM TABLE USR001.EXPENSES
UNLOAD FROM TABLE USR001.PAYROLL
UNLOAD FROM TABLE USR001.GEN_LEDGER
In the following example, the EXPENSES, PAYROLL, and GEN_LEDGER tables are unloaded in any order, and this order can vary from job to job. BMC AMI Utilities processing determines the best order to optimize performance.
In the following example, the tables are unloaded in the order specified: EXPENSES, PAYROLL, GEN_LEDGER:
Best practice
Use one Load statement per SYSIN and use multiple steps or multiple jobs to run the loads serially or in parallel. If for some reason you need multiple Load statements in one SYSIN specify +SYSINORDER(HONOR) if you want to process them serially like Db2 Load does.
Important
- Loading tables that have Db2 enforced referential integrity (RI) in parallel can cause contention issues.
- For the rules that apply to parameter syntax, see Parameter syntax rules.
