Over the past several weeks, we have shared good news around both the availability of Red Hat Enterprise Linux 6.4 and publication of leading performance results jointly achieved by IBM and Red Hat’s performance teams. These achievements showcase building solutions on top of industry-standard x86 hardware using the Kernel-based Virtual Machine (KVM) hypervisor, which is utilized in both Red Hat Enterprise Linux and Red Hat Enterprise Virtualization.
Today, Red Hat has news of record-setting benchmark results achieved with its leading enterprise operating platform. In this blog, we’ll provide further detail on the application performance aspects of enterprise infrastructure and explain how Red Hat Enterprise Linux -- together with recently announced IBM PowerLinux systems that feature new POWER7+ processors with increased frequency and L3 cache -- enables IBM Power customers to run their data centers using the Red Hat open source operating system with optimized availability, reliability and performance.
When it comes to gauging the performance of multi-tier, Java platform-based application environments, a leading measure is the industry-standard SPECjEnterprise2010 benchmark. This benchmark exercises all of the components of a typical solution, including the server hardware, Java Virtual Machine (JVM), database, operating system as well as storage and network subsystems.
In recent tests conducted by IBM, Red Hat Enterprise Linux 6.4 powered a pair of IBM PowerLinux 7R2 servers, one running WebSphere Application Server V8.5 in the middle tier, and another running IBM DB2® 10.1 at the database tier. Together, this solution delivered a new record-breaking performance result of 12,062.46 SPECjEnterprise2010 EjOPS on SPECjEnterprise2010 benchmark(1).
Compared to the previously published results, Red Hat Enterprise Linux and IBM PowerLinux 7R2 server deliver the best single node performance for all dual-processor systems used at the application tier of the benchmark, demonstrate better performance than all other Linux operating systems(2) in the same category and post the fastest per-core result for this test.
On this multi-tier benchmark, IBM and Red Hat demonstrate the combined capabilities of software and hardware in a solution spanning across middle and database tiers. The combination of PowerLinux servers and Red Hat Enterprise Linux offers customers a solid platform for their Java-based applications, as well as databases.
Another good example of an end-to-end enterprise application solution is the SAP Sales and Distribution (SD) Standard Application benchmark. This test was developed by SAP to demonstrate the performance of various hardware systems with their application products in a real-world ERP business environment while stressing application and database software, as well as the operating system.
IBM PowerLinux 7R2 servers running IBM DB2® 10.1 database on top of Red Hat Enterprise Linux 6.4 posted the new world-record result on the SAP Sales and Distribution (SD) Parallel Standard Application Benchmark with the SAP Enhancement Package 5 for SAP ERP 6.0 for all two-processor systems running Linux or Windows(3).
The benchmark result confirms enterprise-class performance and scalability of this solution, which exemplifies the ability to support 8,256 SAP SD-Parallel Standard Application Benchmark users. Additionally, this score demonstrates that the robustness of POWER7+-based systems combined with the economics of Red Hat Enterprise Linux provide enterprises with a solid foundation for both scale-out or scale-up deployment models, while simultaneously offering world-class reliability, security and end-to-end system optimization.
To summarize, as demonstrated by the collection of these outstanding performance results, running the latest version of Red Hat Enterprise Linux on top of IBM's System x or PowerLinux families of servers, offers advantages in terms of scalability and optimized performance, regardless of the underlying hardware architecture.
(1) SPEC, SPECjEnterprise are registered trademarks of Standard Performance Evaluation Corporation. Results from www.spec.org as of 3/4/2013.
WebSphere Application Server V8.5 on IBM PowerLinux 7R2 and DB2 10.1 on IBM PowerLinux 7R2, Red Hat Enterprise Linux 6.4: 12,062.46 SPECjEnterprise2010 EjOPS.
(2) The class of systems is defined as single-node application server with 2 processors. Best results for Linux OS from www.spec.org as of 3/4/2013 is listed below:
(a) WebSphere Application Server V8.5 on IBM System x3650 M4 and DB2 10.1 on IBM System x3650 M4, SUSE Linux Enterprise Server 11 SP2: 9,696.43 SPECjEnterprise2010 EjOPS.
(b) Oracle WebLogic Server 10 on Sun Server X3-2 and Oracle Database 11g Release 2 on Sun Server X3-2L, Oracle Linux 5 Update 7: 8,310.19 SPECjEnterprise2010 EjOPS
(3) Results as of 02/22/2013. Source: http://www.sap.com/benchmark
(a) IBM PowerLinux 7R2 on the two-tier SAP SD standard application benchmark running SAP enhancement package 5 for the SAP ERP 6.0 application; 2 processors / 16 cores / 64 threads, POWER7+ 4.22 GHz, 256 GB memory, 8,256 SD benchmark users, running Red Hat Enterprise Linux 6.4 and DB2® 10, Certification #: 2013006.
(b) IBM PowerLinux 7R2 on the two-tier SAP SD standard application benchmark running SAP enhancement package 5 for the SAP ERP 6.0 application; 2 processors / 16 cores / 64 threads, POWER7+ 4.22 GHz, 256 GB memory, 8,016 SD benchmark users, running SUSE Linux Enterprise Server 11 and DB2® 10, Certification #: 2013003.
(c) NEC Express5800 Model R120d-1M on the two-tier SAP SD standard application benchmark running SAP enhancement package 5 for the SAP ERP 6.0 application; 2 Processors / 16 Cores / 32 Threads, Intel Xeon Processor E5-2690, 2.9 GHz, 128 GB memory, 6,720 SD benchmark users, running Windows Server 2008 R2 Enterprise Edition and SQL Server 2012. Certification # 2012031.