The Oracle Applications Standard Benchmark is focused on ERP applications and represents a mixed workload intended to model the most common transactions operating on the most widely used enterprise application modules. Definitions of transactions that compose the benchmark load were obtained through collaboration with functional consultants and are representative of typical customer workloads.

To ensure that the benchmarks workload characteristics remain constant across a variety of user counts the concept of atomic user group has been defined. The atomic user group is the smallest number of users that can be added at any given time while still ensuring that the workload remains unchanged, specifically in terms of the percentages of each type of user in the mix. The atomic user group consists of the sum of all the base user counts specified in the table. To run more users you simply add an integer number of atomic groups, and thereby the characteristics of the workload remain constant.

For the eBS R12 benchmark, performance comparison(s) between benchmark kits of different sizes such as small, medium and large is not encouraged. By the same token, for benchmark kit of the same size, results comparison(s) between Single Instance versus RAC is also discouraged.

Oracle Application R12 Benchmark

The EBS R12 Standard Benchmark supports four types of configurations: Small, Medium, Large and Extra-Large. The volume of data loaded into Oracle Applications schema is one of two sizes for Release 12.0.4 and only one size for Release 12.1.2. The batch workload and the maximum number of users allowed for benchmark run depend upon the type of configuration selected.

Prior to Release 12.1.2, there were 'Medium' and 'Extra-Large' kits available to cover the Small/Medium and Large/Extra-Large workloads. With Release 12.1.2, there is a single 'Extra-Large' kit available to cover all workloads. Thus the data setup is the same for all sizes; with only the number of users, or the number of lines or employees processed varying with the model size.

The modules used by the E-Business R12 benchmark are:

• Oracle Financials: Accounts Payable, Accounts Receivable, Fixed Assets, General Ledger
• Human Resource Management System
• Customer Support: Service
• Supply Chain Management (SCM): Logistics, Order Management, Purchasing, Shipping
• Self-Service Applications: Employee Self-Service, iExpenses, iProcurement, Oracle Time & Labor

A list of the transactions and their suggested completion rate is specified below in Table 3.

For a description of each transaction, please click on the transaction name in the left column.

Table 3: The R12 Atomic Group

 

Also new to R12, partners can execute individual flows from the complete atomic group. This offers more combinations of running users and may address business case situations that are of interest to particular partners.

Table 4: The R12 OLTP Flows

R12 Functional Updates from 11.5.10

There are several changes and modifications in the R12 E-Business Benchmark Kit that reflect changes from Release 11i

Four online transactions have changed from Forms based (11i) to Web based (R12). They are:

1. AR View Customer Transactions
2. AR Customer Summary (ARCS)
3. PO Create Approve Purchase Order (SS)
4. PO View Purchase Order (SS)

Due to changes in R12 application the following batch processes are now capable of being multi-threaded for better performance. In addition, the SUBMITO2C.SQL script has been modified to incorporate this change.

1. Pick release
2. Pick confirm date
3. Create Accounting

With changes to the R12.1.2 kit, the following batch processes are now capable of being multi-threaded for better performance.
1. Auto Invoice
2. Revenue Recognition

 

When running the R12.1.3 kit at the X-L user count, it may improve performance to slightly adjust the specific transaction user counts and pacing from the first transaction to the last transaction along the business flow to insure that certain queues have ample entries.

Table 5: The R12 OLTP Dependency by Flows