SPARC T5-2 Server Delivers Two World Records on a Single System on JD Edwards EnterpriseOne Benchmark
(As of Tuesday, 26 March 2013)
Oracle's SPARC T5-2 server delivered two world record results for batch-only workloads and batch with concurrent online user workloads, beating IBM's Power 770 results. JD Edwards "Day-in-the-Life" benchmark on a single SPARC T5-2 server with two processors supported 12,000 concurrent online users while concurrently executing mixed-batch workloads at 198.5 universal batch engines (UBEs) per minute. The IBM Power 770 server, with twice as many processors, also supported 12,000 concurrent online users while concurrently executing mixed-batch workloads at only 65 UBEs per minute (1). For batch-only workloads, the SPARC T5-2 server provided a batch rate of 880 UBEs per minute, while the IBM Power 770 was only capable of 643 UBEs per minute (1). The SPARC T5-2 server was configured with Oracle Solaris 10 Containers to host Oracle's JD Edwards EnterpriseOne 9.0.2, with Tools 188.8.131.52, Oracle WebLogic Server 10.3.4, Oracle HTTP Server 11g and Oracle Database 11g Release 2 to reduce server footprint.
Oracle's solution provides customers with high performance and scalability in a compact form factor with built-in virtualization and security features.
JD Edwards EnterpriseOne is an integrated applications suite of enterprise resource planning (ERP) software with 80 JD Edwards Enterprise One application modules that support a diverse set of business operations. The "Day-in-the-Life" (DIL) benchmark simulates the operation of a medium-sized manufacturing company and their typical daily interactions with an ERP system, utilizing five of the most commonly used Enterprise modules. This includes load generated from both interactive users and UBE processes, including payroll, sales order, and purchase order processing, as well as manufacturing and shipping process tracking.
The DIL test suite provides simulated interactive user load via HP LoadRunner scripts that emulate live user online transactions and track corresponding response times within acceptable limits. UBE represents a variety of non-interactive batch type workloads that involve more complex queries of the database and may run for just a few seconds or for many hours. During testing, both UBE and interactive loads may be concurrently applied to the system.
The SPARC T5-2 server executed the UBE processes concurrently at 195.5 UBEs per min while supporting 12,000 online users with sub-second average response time and executed the batch only processes at 880 UBEs per min.
The SPARC T5-2 server online + batch throughput with Short UBEs and Long Batch is 3 times higher than the IBM Power 770 server, both running 12,000 online users.
The two-chip SPARC T5-2 batch-only throughput with Short UBEs and Long Batch is 2.7 times faster per chip than the 4-chip IBM Power 770 server.
The SPARC T5-2 server was able to provide 3 times the performance for JD Edwards "Day-in-the-Life" online + batch mixed workload benchmark with the same number of users while offering 5.4 times the cost savings for the application server when compared to the IBM Power 770 application server. (2)
(2) SPARC T5-2 (SPARC T5-2 Server base package, 2xT5 16-core processors, 32x16GB-1066 DIMMS, 4x600GB 10K RPM 2.5" SAS-2 HDD,2x300GB SSDs, 4x F40 400GB Flash Accelerators, 2x Power Cables) List Price $98,190. IBM Power 770 (IBM Power 770:9917 Model MMC, 2x3.3GHz 16-core, 32x one processor activation, 2xCEC Enclosure with IBM Bezel, I/O Backplane and System Midplane,2x Service Processor, 16x 0/64GB DDR3 Memory (4x16GB) DIMMS-1066MHz Power7 CoD Memory, 24x Activation of 1 GB DDR3 Power7 Memory, 10x Activation of 100GB DDR3 Power7 Memory, 2x Disk/Media Backplane. 2x 300GB SAS 15K RPM 2.5" HDD (AIX/Linux only), 1x SATA slimline DVD-RAM drive, 4x AC Power Supply 1925W) List Price $532,143 . Source: ibm.com, collected 03/25/2013.
Copyright 2013, Oracle and/or its affiliates. All rights reserved. Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of their respective owners. Results as of 03/26/2013