By Brian-Oracle on May 01, 2013
Oracle produced a world record single-server SPECjEnterprise2010 benchmark result of 27,843.57 SPECjEnterprise2010 EjOPS using one of Oracle's SPARC T5-8 servers for both the application and the database tier. This result directly compares the 8-chip SPARC T5-8 server (8 SPARC T5 processors) to the 8-chip IBM Power 780 server (8 POWER7+ processor).
The 8-chip SPARC T5 processor based server is 2.6x faster than the 8-chip IBM POWER7+ processor based server.
Both Oracle and IBM used virtualization to provide 4-chips for application and 4-chips for database.
The server cost/performance for the SPARC T5 processor based server was 6.9x better than the server cost/performance of the IBM POWER7+ processor based server. The cost/performance of the SPARC T5-8 server is $10.72 compared to the IBM Power 780 at $73.83.
The total configuration cost/performance (hardware+software) for the SPARC T5 processor based server was 3.6x better than the IBM POWER7+ processor based server. The cost/performance of the SPARC T5-8 server is $56.21 compared to the IBM Power 780 at $199.42. The IBM system had 1.6x better performance per core, but this did not reduce the total software and hardware cost to the customer. As shown by this comparison, performance-per-core is a poor predictor of characteristics relevant to customers.
The total IBM hardware plus software cost was $2,174,152 versus the total Oracle hardware plus software cost of $1,565,092. At this price IBM could only provide 768 GB of memory while Oracle was able to deliver 2 TB in the SPARC T5-8 server.
The SPARC T5-8 server requires only 8 rack units, the same as the space of the IBM Power 780. In this configuration IBM has a hardware core density of 4 cores per rack unit which contrasts with the 16 cores per rack unit for the SPARC T5-8 server. This again demonstrates why performance-per-core is a poor predictor of characteristics relevant to customers.
The virtualized SPARC T5 processor based server ran the application tier servers on 4 chips using Oracle Solaris Zones and the database tier in a 4-chip Oracle Solaris Zone. The virtualized IBM POWER7+ processor based server ran the application in a 4-chip LPAR and the database in a 4-chip LPAR.
The SPARC T5-8 server ran the Oracle Solaris 11.1 operating system and used Oracle Solaris Zones to consolidate eight Oracle WebLogic application server instances and one database server instance to achieve this result. The IBM system used LPARS and AIX V7.1.
This result demonstrated less than 1 second average response times for all SPECjEnterprise2010 transactions and represents JEE 5.0 transactions generated by 227,500 users.
The application server used Oracle Fusion Middleware components including the Oracle WebLogic 12.1 application server and Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.7.0_15. The database server was configured with Oracle Database 11g Release 2.
IBM has a non-virtualized result (one server for application and one server for database). The IBM PowerLinux 7R2 achieved 13,161.07 SPECjEnterprise2010 EjOPS which means it was 2.1x slower than the SPARC T5-8 server. The total configuration cost/performance (hardware+software) for the SPARC T5 processor based server was 11% better than the IBM POWER7+ processor based server. The cost/performance of the SPARC T5-8 server is $56.21 compared to the IBM PowerLinux 7R2 at $62.26. As shown by this comparison, performance-per-core is a poor predictor of characteristics relevant to customers.
Complete benchmark results are at the SPEC website, SPECjEnterprise2010 Results.
|SPECjEnterprise2010 Performance Chart
Only Two Virtualized Results (App+DB on 1 Server) as of 5/1/2013
|Submitter||EjOPS*||Java EE Server & DB Server|
|Oracle||27,843.57||1 x SPARC T5-8
8 chips, 128 cores, 3.6 GHz SPARC T5
Oracle WebLogic 12c (12.1.1)
Oracle Database 11g (126.96.36.199)
|IBM||10,902.30||1 x IBM Power 780
8 chips, 32 cores, 4.42 GHz POWER7+
WebSphere Application Server V8.5
IBM DB2 Universal Database 10.1
* SPECjEnterprise2010 EjOPS (bigger is better)
Application and Database Server:
2 TB memory
5 x 10 GbE dual-port NIC
6 x 8 Gb dual-port HBA
Oracle WebLogic Server 12c (12.1.1)
Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.7.0_15
Oracle Database 11g (188.8.131.52)
16 GB memory
1 x 8 Gb FC HBA
4 x Sun Flash Accelerator F40 PCI-E Card
Application and Database Server:
786 GB memory
6 x 10 GbE dual-port NIC
3 x 8 Gb four-port HBA
IBM WebSphere Application Server V8.5
IBM J9 VM (build 2.6, JRE 1.7.0 IBM J9 AIX ppc-32)
IBM DB2 10.1
IBM InfoSphere Optim pureQuery Runtime v3.1.1
Benchmark DescriptionSPECjEnterprise2010 is the third generation of the SPEC organization's J2EE end-to-end industry standard benchmark application. The new SPECjEnterprise2010 benchmark has been re-designed and developed to cover the Java EE 5 specification's significantly expanded and simplified programming model, highlighting the major features used by developers in the industry today. This provides a real world workload driving the Application Server's implementation of the Java EE specification to its maximum potential and allowing maximum stressing of the underlying hardware and software systems,
- The web zone, servlets, and web services
- The EJB zone
- JPA 1.0 Persistence Model
- JMS and Message Driven Beans
- Transaction management
- Database connectivity
The primary metric of the SPECjEnterprise2010 benchmark is jEnterprise Operations Per Second (SPECjEnterprise2010 EjOPS). The primary metric for the SPECjEnterprise2010 benchmark is calculated by adding the metrics of the Dealership Management Application in the Dealer Domain and the Manufacturing Application in the Manufacturing Domain. There is NO price/performance metric in this benchmark.
Key Points and Best Practices
- Eight Oracle WebLogic server instances on the SPARC T5-8 server were hosted in 8 separate Oracle Solaris Zones to demonstrate consolidation of multiple application servers. The 8 zones were bound to 4 resource pools using 64 cores (4 cpu chips).
The database ran in a separate Oracle Solaris Zone bound to a resource pool
consisting 64 cores (4 cpu chips). The database shadow processes
were run in the FX scheduling class and bound to one of four cpu chips
- The Oracle WebLogic application servers were executed in the FX scheduling class to improve performance by reducing the frequency of context switches.
- The Oracle log writer process was run in the FX scheduling class at processor priority 60 to use the Critical Thread feature.
- SPECjEnterprise2010 Results Page
- SPARC T5-8 Result Page at SPEC
- SPARC T5-8 Server
Sun Flash Accelerator F40 PCIe Card
Oracle Database 11g Release 2
Oracle Fusion Middleware
Oracle WebLogic Suite
SPEC and the benchmark name SPECjEnterprise are registered trademarks of the Standard Performance Evaluation Corporation. Results from www.spec.org as of 5/1/2013. SPARC T5-8, 27,843.57 SPECjEnterprise2010 EjOPS; IBM Power 780, 10,902.30 SPECjEnterprise2010 EjOPS; IBM PowerLinux 7R2, 13,161.07 SPECjEnterprise2010 EjOPS. Oracle server only hardware list price is $298,494 and total hardware plus software list price is $1,565,092 from http://www.oracle.com as of 5/22/2013. IBM server only hardware list price is $804,931 and total hardware plus software cost of $2,174,152 based on public pricing from http://www.ibm.com as of 5/22/2013. IBM PowerLinux 7R2 server total hardware plus software cost of $819,451 based on public pricing from http://www.ibm.com as of 5/22/2013.