SPARC T4 Servers Running Oracle Solaris 11 and Oracle RAC Deliver World Record on PeopleSoft HRMS 9.1

Oracle's SPARC T4-4 server running Oracle's PeopleSoft HRMS Self-Service 9.1 benchmark achieved world record performance with 18,000 interactive users. This was accomplished using a high availability configuration using Oracle Real Application Clusters (RAC) 11g Release 2 software for the database tier running on Oracle Solaris 11. The benchmark configuration included the SPARC T4-4 server for the application tier, a SPARC T4-2 server for the web tier and two SPARC T4-2 servers for the database tier.

  • The combination of the SPARC T4 servers running PeopleSoft HRSS 9.1 benchmark supports 4.5x the number of users an IBM pSeries 570 running PeopleSoft HRSS 8.9, with an average response time 40 percent better than IBM.

  • This result was obtained with two SPARC T4-2 servers running the database service using Oracle Real Application Clusters 11g Release 2 software in a high availability configuration.

  • The two SPARC T4-2 servers in the database tier used Oracle Solaris 11, and Oracle RAC 11g Release 2 software with database shared disk storage managed by Oracle Automatic Storage Management (ASM).

  • The average CPU utilization on one SPARC T4-4 server in the application tier handling 18,000 users is 54 percent, showing significant headroom for growth.

  • The SPARC T4 server for the application tier used Oracle Solaris Containers on Oracle Solaris 10, which provides a flexible, scalable and manageable virtualized environment.

  • The Peoplesoft HRMS Self-Service benchmark demonstrates better performance on Oracle hardware and software, engineered to work together, than Oracle software on IBM.

Performance Landscape

PeopleSoft HRMS Self-Service 9.1 Benchmark
Systems Processors Users Ave Response -
Search (sec)
Ave Response -
Save (sec)
SPARC T4-2 (web)
SPARC T4-4 (app)
2 x SPARC T4-2 (db)
2 x SPARC T4, 2.85 GHz
4 x SPARC T4, 3.0 GHz
2 x (2 x SPARC T4, 2.85 GHz)
18,000 1.048 0.742
SPARC T4-2 (web)
SPARC T4-4 (app)
SPARC T4-4 (db)
2 x SPARC T4, 2.85 GHz
4 x SPARC T4, 3.0 GHz
4 x SPARC T4, 3.0 GHz
15,000 1.01 0.63
PeopleSoft HRMS Self-Service 8.9 Benchmark
IBM Power 570 (web/app)
IBM Power 570 (db)
12 x POWER5, 1.9 GHz
4 x POWER5, 1.9 GHz
4,000 1.74 1.25
IBM p690 (web)
IBM p690 (app)
IBM p690 (db)
4 x POWER4, 1.9 GHz
12 x POWER4, 1.9 GHz
6 x 4392 MIPS/Gen1
4,000 1.35 1.01

The main differences between version 9.1 and version 8.9 of the benchmark are:

  • the database expanded from 100K employees and 20K managers to 500K employees and 100K managers,
  • the manager data was expanded,
  • a new transaction, "Employee Add Profile," was added, the percent of users executing it is less then 2%, and the transaction has a heavier footprint,
  • version 9.1 has a different benchmark metric (Average Response Search/Save time for x number of users) versus single user search/save time,
  • newer versions of the PeopleSoft application and PeopleTools software are used.

Configuration Summary

Application Server:

1 x SPARC T4-4 server
4 x SPARC T4 processors 3.0 GHz
512 GB main memory
5 x 300 GB SAS internal disks,
2 x 100 GB internal SSDs
1 x 300 GB internal SSD
Oracle Solaris 10 8/11
PeopleSoft PeopleTools 8.51.02
PeopleSoft HCM 9.1
Oracle Tuxedo, Version, 64-bit, Patch Level 031
Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.6.0_20

Web Server:

1 x SPARC T4-2 server
2 x SPARC T4 processors 2.85 GHz
256 GB main memory
2 x 300 GB SAS internal disks
1 x 100 GB internal SSD
Oracle Solaris 10 8/11
PeopleSoft PeopleTools 8.51.02
Oracle WebLogic Server 11g (10.3.3)
Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.6.0_20

Database Server:

2 x SPARC T4-2 servers, each with
2 x SPARC T4 processors 2.85 GHz
128 GB main memory
3 x 300 GB SAS internal disks
Oracle Solaris 11 11/11
Oracle Database 11g Release 2
Oracle Real Application Clusters

Database Storage:

1 x Sun Storage F5100 Flash Array (80 flash modules)
1 x COMSTAR Sun Fire X4470 M2 server
4 x Intel Xeon X7550 processors 2.0 GHz
128 GB main memory
Oracle Solaris 11 11/11
2 x COMSTAR Sun Fire X4275 servers, each with
1 x Intel Xeon E5540 processor 2.53 GHz
6 GB main memory)
12 x 2 TB SAS disks
Oracle Solaris 11 Express 2010.11


1 x 8-port 10GbE switch
1 x 24-port 1GbE switch
1 x 32-port Brocade FC switch

Benchmark Description

The purpose of the PeopleSoft HRMS Self-Service 9.1 benchmark is to measure comparative online performance of the selected processes in PeopleSoft Enterprise HCM 9.1 with Oracle Database 11g. The benchmark kit is an Oracle standard benchmark kit run by all platform vendors to measure the performance. It is an OLTP benchmark with no dependency on remote COBOL calls, there is no batch workload, and DB SQLs are moderately complex. The results are certified by Oracle and a white paper is published.

PeopleSoft defines a business transaction as a series of HTML pages that guide a user through a particular scenario. Users are defined as corporate Employees, Managers and HR administrators. The benchmark consists of 14 scenarios which emulate users performing typical HCM transactions such as viewing paychecks, promoting and hiring employees, updating employee profiles and other typical HCM application transactions.

All of these transactions are well defined in the PeopleSoft HR Self-Service 9.1 benchmark kit. This benchmark metric is the Weighted Average Response search/save time for all users.

Key Points and Best Practices

  • The combined processing power of two SPARC T4-2 servers running the highly available Oracle RAC database can provide greater throughput and Oracle RAC scalability than is available from a single server.

  • All database data files/recovery files and Oracle Clusterware files were created with Oracle Automatic Storage Management (Oracle ASM) volume manager and file system which resulted in equivalent performance of conventional volume managers, file systems, and raw devices, but with the added benefit of the ease of management provided by Oracle ASM integrated storage management solution.

  • Five Oracle PeopleSoft Domains with 200 application servers (40 per each Domain) on the SPARC T4-4 server were hosted in two separate Oracle Solaris Containers for a total of 10 Domains/400 application servers processes to demonstrate consolidation of multiple application servers, ease of administration and load balancing.

  • Each Oracle Solaris Container was bound to a separate processor set, each containing 124 virtual processors. The default set (composed of 4 virtual processors from first and third processor socket, total of 8 virtual processors) was used for network and disk interrupt handling. This was done to improve performance by reducing memory access latency by using the physical memory closest to the processors and offload I/O interrupt handling to default set virtual processors, freeing up processing resources for application server virtual processors.

See Also

Disclosure Statement

Oracle's PeopleSoft HRMS 9.1 benchmark,, results 5/1/2012.


Post a Comment:
Comments are closed for this entry.

BestPerf is the source of Oracle performance expertise. In this blog, Oracle's Strategic Applications Engineering group explores Oracle's performance results and shares best practices learned from working on Enterprise-wide Applications.

Index Pages

« August 2016