Wednesday Jul 06, 2016

SPECjEnterprise2010: Oracle Server X6-2 Top x86 Result

Two Oracle Server X6-2 systems, using the Intel Xeon E5-2699 v4 processor, produced a world record for x86 systems SPECjEnterprise2010 benchmark result of 27,803.39 SPECjEnterprise2010 EjOPS. One Oracle Server X6-2 system ran the application tier and the second Oracle Server X6-2 system ran the database tier.

  • The Oracle Server X6-2 system demonstrated 44% better performance when compared to the IBM X3650 M5 server result of 19,282.14 SPECjEnterprise2010 EjOPS.

  • The Oracle Server X6-2 system demonstrated 29% better performance when compared to the previous generation Oracle Server X5-2 system result of 21,504.30 SPECjEnterprise2010 EjOPS.

  • This result used Oracle WebLogic Server 12c, Java HotSpot 64-Bit Server 1.8.0_91, Oracle Database 12c, and Oracle Linux.

Performance Landscape

Complete benchmark results are at the SPEC website, SPECjEnterprise2010 Results. The table below shows the top two-chip x86 server results.

SPECjEnterprise2010 Performance Chart
as of 7/6/2016
Submitter EjOPS* Application Server Database Server
Oracle 27,803.39 1 x Oracle Server X6-2
2 x 2.2 GHz Intel Xeon E5-2699 v4
Oracle WebLogic 12c (12.2.1)
1 x Oracle Server X6-2
2 x 2.2 GHz Intel Xeon E5-2699 v4
Oracle Database 12c (12.1.0.2)
Oracle 21,504.30 1 x Oracle Server X5-2
2 x 2.3 GHz Intel Xeon E5-2699 v3
Oracle WebLogic 12c (12.1.3)
1 x Oracle Server X5-2
2 x 2.3 GHz Intel Xeon E5-2699 v3
Oracle Database 12c (12.1.0.2)
IBM 19,282.14 1 x IBM X3650 M5
2 x 2.6 GHz Intel Xeon E5-2697 v3
WebSphere Application Server V8.5
1 x IBM X3850 X6
4 x 2.8 GHz Intel Xeon E7-4890 v2
IBM DB2 10.5

* EjOPS – SPECjEnterprise2010 EjOPS, bigger is better.

Configuration Summary

Application Server:

1 x Oracle Server X6-2
2 x Intel Xeon Processor E5-2699 v4 (2.2 GHz)
256 GB memory
3 x 10 GbE NIC
Oracle Linux 6.7 (kernel-4.1.12-37.2.2.el6uek.x86_64)
Oracle WebLogic Server 12c (12.2.1)
Java HotSpot(TM) 64-Bit Server VM on Linux, version 1.8.0_91 (Java SE 8 Update 91)

Database Server:

1 x Oracle Server X6-2
2 x Intel Xeon Processor E5-2699 v4 (2.2 GHz)
512 GB memory
2 x 10 GbE NIC
1 x 16 Gb FC HBA
2 x Oracle Server X5-2L Storage
Oracle Linux 6.7 (kernel-4.1.12-37.2.2.el6uek.x86_64)
Oracle Database 12c Enterprise Edition Release 12.1.0.2

Benchmark Description

SPECjEnterprise2010 is the third generation of the SPEC organization's J2EE end-to-end industry standard benchmark application. The SPECjEnterprise2010 benchmark has been 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 workload consists of an end to end web based order processing domain, an RMI and Web Services driven manufacturing domain and a supply chain model utilizing document based Web Services. The application is a collection of Java classes, Java Servlets, Java Server Pages, Enterprise Java Beans, Java Persistence Entities (pojo's) and Message Driven Beans.

The SPECjEnterprise2010 benchmark heavily exercises all parts of the underlying infrastructure that make up the application environment, including hardware, JVM software, database software, JDBC drivers, and the system network.

The primary metric of the SPECjEnterprise2010 benchmark is jEnterprise Operations Per Second ("SPECjEnterprise2010 EjOPS"). This metric 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

  • Four Oracle WebLogic server instances were started using numactl binding 2 instances per chip.
  • Four Oracle database listener processes were started, 2 processes bound per processor.
  • Additional tuning information is in the report at spec.org.
  • COD (Cluster on Die) is enabled in the BIOS on the application server.

See Also

Disclosure Statement

SPEC and the benchmark name SPECjEnterprise are registered trademarks of the Standard Performance Evaluation Corporation. Oracle Server X6-2, 27,803.39 SPECjEnterprise2010 EjOPS; Oracle Server X5-2, 21,504.30 SPECjEnterprise2010 EjOPS; IBM System x3650 M5, 19,282.14 SPECjEnterprise2010 EjOPS. Results from www.spec.org as of 7/6/2016.

Wednesday Jun 29, 2016

SPECjEnterprise2010: SPARC S7-2 Secure and Unsecure Results

Oracle's SPARC S7-2 servers produced a SPECjEnterprise2010 benchmark result of 14,400.78 SPECjEnterprise2010 EjOPS using one SPARC S7-2 server for the application tier and one SPARC S7-2 server for the database server.

The SPARC S7-2 servers also obtained a result of 14,121.47 SPECjEnterprise2010 EjOPS using encrypted data. This secured result used Oracle Advanced Security Transparent Data Encryption (TDE) for the application database tablespaces with the AES-128 cipher. The network connection between the application server and the database server was also secured using Oracle's Network Data Encryption with the JDBC driver and RC4-128 cipher.

  • The SPARC S7-2 server, equipped with two SPARC S7 processors, demonstrated 43% better SPECjEnterprise2010 EjOPS/app-server-cores performance compared to the Oracle Server X6-2 system result.

  • The SPARC S7-2 server, equipped with two SPARC S7 processors, demonstrated 50% better SPECjEnterprise2010 EjOPS/app-server-cores performance compared to the Oracle Server X5-2 system result.

  • The SPARC S7-2 server, equipped with two SPARC S7 processors, demonstrated 31% better SPECjEnterprise2010 EjOPS/core performance compared to the 2-chip IBM x3650 server result.

  • The application server used Oracle Fusion Middleware components including the Oracle WebLogic 12.2 application server and Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.8.0_92. The database server was configured with Oracle Database 12c Release 1.

  • The benchmark performance using the secure SPARC S7-2 server configuration with encryption ran within 2% of the performance of the non-secure SPARC S7-2 server result.

Performance Landscape

Select single application server results. Complete benchmark results are at the SPEC website, SPECjEnterprise2010 Results.

SPECjEnterprise2010 Performance Chart
7/6/2016
Java EE Server DB Server EjOPS app
cores
EjOPS
/appcore
1 x SPARC S7-2
2 x 4.27 GHz SPARC S7
Oracle WebLogic 12c (12.2.1)
1 x SPARC S7-2
2 x 4.27 GHz SPARC S7
Oracle Database 12c (12.1.0.2)
14,400.78 16 900
1 x SPARC S7-2
2 x 4.27 GHz SPARC S7
Oracle WebLogic 12c (12.2.1)
Network Data Encryption for JDBC
1 x SPARC S7-2
2 x 4.27 GHz SPARC S7
Oracle Database 12c (12.1.0.2)
Transparent Data Encryption
14,121.47 16 882
1 x Oracle Server X6-2
2 x 2.2 GHz Intel Xeon E5-2699 v4
Oracle WebLogic 12c (12.2.1)
1 x Oracle Server X6-2
2 x 2.2 GHz Intel Xeon E5-2699 v4
Oracle Database 12c (12.1.0.2)
27,803.39 44 631
1 x Oracle Server X5-2
2 x 2.3 GHz Intel Xeon E5-2699 v3
Oracle WebLogic 12c (12.1.3)
1 x Oracle Server X5-2
2 x 2.3 GHz Intel Xeon E5-2699 v3
Oracle Database 12c (12.1.0.2)
21,504.30 36 597
1 x IBM System x3650 M5
2 x 2.6 GHz Intel Xeon E5-2697 v3
WebSphere Application Server V8.5
1 x IBM System x3850 X6
4 x 2.8 GHz Intel Xeon E7-4890 v2
IBM DB2 10.5 FP5
19,282.14 28 688
1 x IBM Power S824
4 x 3.5 GHz POWER 8
WebSphere Application Server V8.5
1 x IBM Power S824
4 x 3.5 GHz POWER 8
IBM DB2 10.5 FP3
22,543.34 24 939

EjOPS – SPECjEnterprise2010 EjOPS (bigger is better)
app cores – application server cores used
EjOPS/appcore – SPECjEnterprise2010 EjOPS divided by total application server cores (bigger is better)

Configuration Summary

Application Server:

1 x SPARC S7-2 server, with
2 x SPARC S7 processor (4.27 GHz)
512 GB memory (16 x 32 GB)
2 x 600 GB SAS-3 HDD
2 x 400 GB SAS-3 SSD
2 x Sun Dual Port 10 GBase-T Network Adapter
Oracle Solaris 11.3
Oracle WebLogic Server 12c (12.2.1)
Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.8.0_92

Database Server:

1 x SPARC S7-2 server, with
2 x SPARC S7 processor (4.27 GHz)
512 GB memory (16 x 32 GB)
2 x 600 GB SAS-3 HDD
1 x Sun Dual Port 10 GBase-T Network Adapter
2 x Sun Storage 16 Gbit Fibre Channel Universal HBA
Oracle Solaris 11.3
Oracle Database 12c (12.1.0.2)

Storage Servers:

1 x Oracle Server X6-2L (24 Slot Disk-Cage), with
2 x Intel Xeon Processor E5-2643 v4 (3.4 GHz)
256 GB memory
1 x Sun Storage 16 Gbit Fibre Channel PCI-E HBA dual port
2 x 3.2 TB NVMe SSD
2 x 600 GB SAS-3 HDD
Oracle Solaris 11.3 (11.3.8.0.4)

1 x Oracle Server X6-2L (24 Slot Disk Cage), with
2 x Intel Xeon Processor E5-2643 v4 (3.4 GHz)
256 GB memory
1 x Sun Storage 16 Gb Fibre Channel PCI-E HBA dual port
14 x 600 GB 10k RPM SAS-3 HDD
Oracle Solaris 11.3 (11.3.8.0.4)

1 x Brocade 6510 16 Gb FC switch

Benchmark Description

SPECjEnterprise2010 is the third generation of the SPEC organization's J2EE end-to-end industry standard benchmark application. The SPECjEnterprise2010 benchmark has been 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
Moreover, SPECjEnterprise2010 also heavily exercises all parts of the underlying infrastructure that make up the application environment, including hardware, JVM software, database software, JDBC drivers, and the system network.

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.

See Also

Disclosure Statement

SPEC and the benchmark name SPECjEnterprise are registered trademarks of the Standard Performance Evaluation Corporation. Results from www.spec.org as of 7/6/2016. SPARC S7-2, 14,400.78 SPECjEnterprise2010 EjOPS (unsecure); SPARC S7-2, 14,121.47 SPECjEnterprise2010 EjOPS (secure); Oracle Server X6-2, 27,803.39 SPECjEnterprise2010 EjOPS (unsecure); Oracle Server X5-2, 21,504.30 SPECjEnterprise2010 EjOPS (unsecure); IBM Power S824, 22,543.34 SPECjEnterprise2010 EjOPS (unsecure); IBM System x3650 M5, 19,282.14 SPECjEnterprise2010 EjOPS (unsecure).

Real-Time Enterprise: SPARC S7-2 Advantage Per Core Under Load Compared to 2-Chip x86 E5-2699 v3

A goal of the modern business is real-time enterprise where analytics are run simultaneously with transaction processing on the same system to provide the most effective decision making. Oracle Database 12c Enterprise Edition utilizing the Oracle In-Memory option is designed for the same database to be able to perform transactions at the highest performance and to perform analytical calculations that once took days or hours to complete orders of magnitude faster.

Oracle's SPARC S7 processor has deep innovations to take the real-time enterprise to the next level of performance. In this test both OLTP transactions and analytical queries were run in a single database instance using all of the same features of Oracle Database 12c Enterprise Edition including the Oracle In-Memory option in order to compare the advantages of the SPARC S7 processor to the Intel Xeon Processor E5-2699 v3.

The SPARC S7 processor uses the Data Analytics Accelerator (DAX). DAX is not a SIMD instruction, but rather an actual co-processor that offloads in-memory queries which frees the cores up for other processing. The DAX has direct access to the memory bus and can execute scans at near full memory bandwidth. Oracle makes the DAX API available to other applications, so this kind of acceleration is not just available to Oracle Database.

  • Oracle's SPARC S7-2 server ran the in-memory analytics RCDB based queries 2.3x faster per chip under load than a two-chip x86 Intel Xeon Processor E5-2699 v3 server on the 24 stream test. Furthermore, the SPARC S7-2 server ran the in-memory analytics RCDB based queries 5.1x faster per core under load than the same x86 server.

  • The SPARC S7-2 server and the two-chip Intel Xeon Processor E5-2699 v3 server both ran OLTP transactions and the in-memory analytics on the same database instance using Oracle Database 12c Enterprise Edition utilizing the Oracle In-Memory option.

Performance Landscape

The table below compares the SPARC S7-2 server and 2-chip x86 Intel Xeon Processor E5-2699 v3 server while running OLTP and in-memory analytics against tables in the same database instance. The same set of transactions and queries were executed on each system. All of the following results were run as part of this benchmark effort.

Real-Time Enterprise Performance Chart
24 RCDB DSS Streams, 112 OLTP users
Simultaneous Mixed
Workload
SPARC S7-2 2-Chip
x86 E5 v3
SPARC Per Chip
Advantage
SPARC Per Core
Advantage
OLTP Transactions
(Transactions per Second)
195,790 216,302 0.9x 2.0x
Analytic Queries
(Queries per Minute)
107 47 2.3x 5.1x

Configuration Summary

SPARC Server:

1 X SPARC S7-2 server
2 X SPARC S7 processor
512 GB Memory
Oracle Solaris 11.3
Oracle Database 12c Enterprise Edition Release 12.1.0.2.0

x86 Server:

1 X Oracle Server X5-2L
2 X Intel Xeon Processor E5-2699 v3
256 GB Memory
Oracle Linux 6.5 (3.8.13-16.2.1.el6uek.x86_64)
Oracle Database 12c Enterprise Edition Release 12.1.0.2.0

Benchmark Description

The Real-Time Enterprise benchmark simulates the demands of customers who simultaneously run both their OLTP database and the related historical warehouse DSS data that would be based on that OLTP data. It answers the question of how a system will perform when doing data analysis, while at the same time executing real-time on-line transactions.

The OLTP workload simulates an Order Inventory System that exercises both reads and writes with a potentially large number of users that stresses the lock management and connectivity, as well as, database access.

The number of customers, orders and users is fully parametrized. This benchmark is base on 100 GB dataset, 15 million customers, 600 million orders and up to 580 users. The workload consists of a number of transaction types including show-expenses, part-cost, supplier-phone, low-inv, high-inv, update-price, update-phone, update-cost, and new-order.

The real cardinality database (RCDB) schema was created to showcase the potential speedup one may see moving from on disk, row format data warehouse/star schema, to utilizing Oracle Database 12c's In-Memory feature for analytical queries.

The DSS workload consists of, as many as, 2,304 unique queries asking questions such as "In 2014, what was the total revenue of single item orders?" or "In August 2013, how many orders exceeded a total price of $50?" Questions like these can help a company see where to focus for further revenue growth or identify weaknesses in their offerings.

RCDB scale factor 1050 represents a 1.05 TB data warehouse. It is transformed into a star schema of 1.0 TB, and then becomes 110 GB in size when loaded in memory. It consists of 1 fact table, and 4 dimension tables with over 10.5 billion rows. There are 56 columns with most cardinalities varying between 5 and 2,000, a primary key being an example of something outside this range.

The results were obtained running a set of OLTP transactions and analytic queries simultaneously against two schema: a real time online orders system and a related historical orders schema configured as a real cardinality database (RCDB) star schema. The in-memory analytics RCDB queries are executed using the Oracle Database 12c In-Memory columnar feature.

Two reports are generated: one for the OLTP-Perf workload and one for the RCDB DSS workload. For the analytical DSS workload, queries per minute and average query elapsed times are reported. For the OLTP-Perf workload, both transactions-per-second in thousands and OLTP average response times in milliseconds are reported.

Key Points and Best Practices

  • This benchmark utilized the SPARC S7 processor's co-processor DAX for query acceleration.

  • All SPARC S7-2 server results were run with out-of-the-box tuning for Oracle Solaris.

  • All Oracle Server X5-2L system results were run with out-of-the-box tunings for Oracle Linux except for the setting in /etc/sysctl.conf to get large pages for the Oracle Database:

      vm.nr_hugepages=98304

  • To create an in memory area, the following was added to the init.ora:

      inmemory_size = 120g

  • An example of how to set a table to be in memory is below:

      ALTER TABLE CUSTOMER INMEMORY MEMCOMPRESS FOR QUERY HIGH

See Also

Disclosure Statement

Copyright 2016, 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 June 29, 2016.

The previous information is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle's products remains at the sole discretion of Oracle.

Oracle Communications ASAP – Telco Subscriber Activation: SPARC S7-2L Fastest Recorded Result

Oracle's SPARC S7-2L server delivered fastest recorded results on Oracle Communications ASAP. The SPARC S7-2L server ran Oracle Solaris 11.3 with Oracle Database 12c and Oracle Communications ASAP version 7.3, with Oracle Database 11g Release 2 Client.

  • Running Oracle Communications ASAP, the SPARC S7-2L server delivered a fastest recorded result of 3,292 ASDLs/sec (atomic network activation actions).

  • The SPARC S7-2L server running a single instance of the Oracle Communications ASAP application, with both the application and database tiers consolidated onto a single machine, easily supported the service activation volumes of 3,292 ASDLs/sec which is representative of a typical mobile operator with more than 100 million subscribers.

Performance Landscape

All of the following results were run as part of this benchmark effort.

ASAP 7.2 Test Results – 16 NEP
System ASDLs/sec CPU Usage
SPARC S7-2L 3,292 63%

Configuration Summary

Hardware Configuration:

SPARC S7-2L server
2 x SPARC S7 processors (4.27 GHz)
512 GB memory
Flash Storage

Software Configuration:

Oracle Communications ASAP 7.3 Version B122
Oracle Solaris 11.3
Oracle Database 12c Release 12.1.0.2.0
Oracle WebLogic Server 12.1.3.0.0
Oracle JDK 7 update 95

Benchmark Description

Oracle Communications ASAP provides a convergent service activation platform that automatically activates customer services in a heterogeneous network and IT environment. It supports the activation of consumer and business services in fixed and mobile domains against network and IT applications.

ASAP enables rapid service design and network technology introduction by means of its metadata-driven architecture, design-time configuration environment, and catalog of pre-built activation cartridges to reduce deployment time, cost, and risk. The application has been deployed for mobile (3G, 4G and M2M) services and fixed multi-play (broadband, voice, video, and IT) services in telecommunications, cable and satellite environments as well as for business voice, data, and IT cloud services.

It may be deployed in a fully integrated manner as part of the Oracle Communications Service Fulfillment solution or directly integrated with third- party upstream systems. Market-proven for high-volume performance and scalability, Oracle Communications ASAP is deployed by more than 75 service providers worldwide and activates services for approximately 250 million subscribers globally.

The throughput of ASAP is measured in atomic actions per second (or ASDLs/sec). An atomic action is a single command or operation that can be executed on a network element. Atomic actions are grouped together to form a common service action, where each service action typically relates to an orderable item, such as "GSM voice" or "voice mail" or "GSM data". One or more service actions are invoked by an order management system via an activation work order request.

The workload resembles a typical mobile order to activate a GSM subscriber. A single service action to add a subscriber consists of seven atomic actions where each atomic action executes a command on a network element. Each network element was serviced by a dedicated Network Element Processor (NEP). The ASAP benchmark can vary the number of NEPs, which correlate to the complexity of a Telco operator's environment.

See Also

  • SPARC S7-2L Server
    oracle.com     OTN     Blog
  • Oracle Communications ASAP
    oracle.com
  • Oracle Solaris
    oracle.com     OTN     Blog
  • Oracle Database
    oracle.com     OTN     Blog
  • Oracle Fusion Middleware
    oracle.com     OTN     Blog
  • Disclosure Statement

    Copyright 2016, 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 June 29, 2016.

    Tuesday May 31, 2016

    SAP Two-Tier Standard Sales and Distribution SD Benchmark: SPARC M7-8 World Record 8 Processors

    Oracle's SPARC M7-8 server produced a world record result for 8-processors on the SAP two-tier Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement Package 5 for SAP ERP 6.0 (8 chips / 256 cores / 2048 threads).

    • The SPARC M7-8 server achieved 130,000 SAP SD benchmark users running the two-tier SAP Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement Package 5 for SAP ERP 6.0.

    • The SPARC M7-8 server is 1.5x faster per core than x86-based HPE Integrity Superdome X running the two-tier SAP Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement Package 5 for SAP ERP 6.0.

    • The SPARC M7-8 server result was run with Oracle Solaris 11 and used Oracle Database 12c.

    • Previously the SPARC T7-2 server set the 2-chip server world record achieving 30,800 SAP SD benchmark users running the two-tier SAP Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement Package 5 for SAP ERP 6.0.

    Performance Landscape

    SAP-SD 2-tier performance table in decreasing performance order with SAP ERP 6.0 Enhancement Package 5 for SAP ERP 6.0 results (current version of the benchmark as of May 2012).

    SAP SD Two-Tier Benchmark
    System
    Processor
    OS
    Database
    Users Resp Time
    (sec)
    Users/
    core
    Cert#
    SPARC M7-8
    8 x SPARC M7 (8x 32core)
    Oracle Solaris 11
    Oracle Database 12c
    130,000 0.93 508 2016020
    HPE Integrity Superdome X
    16 x Intel E7-8890 v3 (16x 18core)
    Windows Server 2012 R2
    Datacenter Edition
    SQL Server 2014
    100,000 0.99 347 2016002

    Number of cores presented are per chip, to get system totals, multiple by the number of chips.

    Complete benchmark results may be found at the SAP benchmark website http://www.sap.com/benchmark.

    Configuration Summary and Results

    Database/Application Server:

    1 x SPARC M7-8 server with
    8 x SPARC M7 processors (4.13 GHz, total of 8 processors / 256 cores / 2048 threads)
    4 TB memory
    Oracle Solaris 11.3
    Oracle Database 12c

    Database Storage:
    7 x Sun Server X3-2L each with
    2 x Intel Xeon Processors E5-2609 (2.4 GHz)
    16 GB memory
    4 x Sun Flash Accelerator F40 PCIe Card
    12 x 3 TB SAS disks
    Oracle Solaris 11

    REDO log Storage:
    1 x Pillar FS-1 Flash Storage System, with
    2 x FS1-2 Controller (Netra X3-2)
    2 x FS1-2 Pilot (X4-2)
    4 x DE2-24P Disk enclosure
    96 x 300 GB 10000 RPM SAS Disk Drive Assembly

    Certified Results (published by SAP)

    Number of SAP SD benchmark users: 130,000
    Average dialog response time: 0.93 seconds
    Throughput:
      Fully processed order line items per hour: 14,269,670
      Dialog steps per hour: 42,809,000
      SAPS: 713,480
    Average database request time (dialog/update): 0.018 sec / 0.039 sec
    SAP Certification: 2016020

    Benchmark Description

    The SAP Standard Application SD (Sales and Distribution) Benchmark is an ERP business test that is indicative of full business workloads of complete order processing and invoice processing, and demonstrates the ability to run both the application and database software on a single system. The SAP Standard Application SD Benchmark represents the critical tasks performed in real-world ERP business environments.

    SAP is one of the premier world-wide ERP application providers, and maintains a suite of benchmark tests to demonstrate the performance of competitive systems on the various SAP products.

    See Also

    Disclosure Statement

    Two-tier SAP Sales and Distribution (SD) standard application benchmarks, SAP Enhancement Package 5 for SAP ERP 6.0 as of 5/14/16:

    SPARC M7-8 (8 processors, 256 cores, 2048 threads) 130,000 SAP SD users, 8 x 4.13 GHz SPARC M7, 4 TB memory, Oracle Database 12c, Oracle Solaris 11, Cert# 2016020
    SPARC T7-2 (2 processors, 64 cores, 512 threads) 30,800 SAP SD users, 2 x 4.13 GHz SPARC M7, 1 TB memory, Oracle Database 12c, Oracle Solaris 11, Cert# 2015050
    HPE Integrity Superdome X (16 processors, 288 cores, 576 threads) 100,000 SAP SD users, 16 x 2.5 GHz Intel Xeon Processor E7-8890 v3 4096 GB memory, SQL Server 2014, Windows Server 2012 R2 Datacenter Edition, Cert# 2016002

    SAP, R/3, reg TM of SAP AG in Germany and other countries. More info www.sap.com/benchmark

    Tuesday Mar 22, 2016

    Siebel PSPP: SPARC T7-2 World Record Result, Beats IBM

    Oracle set a new world record for the Siebel Platform Sizing and Performance Program (PSPP) benchmark using Oracle's SPARC T7-2 server for the application server with Oracle's Siebel CRM 8.1.1.4 Industry Applications and Oracle Database 12c running on Oracle Solaris.

    • The SPARC T7-2 server running the application tier achieved 55,000 users with sub-second response time and with throughput of 457,909 business transactions per hour on the Siebel PSPP benchmark.

    • The SPARC T7-2 server in the application tier delivered 3.3 times more users on a per chip basis compared to published IBM POWER8 based server results.

    • For the new Oracle results, eight cores of a SPARC T7-2 server were used for the database tier running at 32% utilization (as measured by mpstat). The IBM result used 6 cores at about 75% utilization for the database/gateway tier.

    • The SPARC T7-2 server in the application tier delivered nearly the same number of users on a per core basis compared to published IBM POWER8 based server results.

    • The SPARC T7-2 server in the application tier delivered nearly 2.8 times more users on a per chip basis compared to earlier published SPARC T5-2 server results.

    • The SPARC T7-2 server in the application tier delivered nearly 1.4 times more users on a per core basis compared to earlier published SPARC T5-2 server results.

    • The SPARC T7-2 server used Oracle Solaris Zones which provide flexible, scalable and manageable virtualization to scale the application within and across multiple nodes.

    The Siebel 8.1.1.4 PSPP workload includes Siebel Call Center and Order Management System.

    Performance Landscape

    Application Server TPH Users Users/
    Chip
    Users/
    Core
    Response Times
    Call
    Center
    Order
    Mgmt
    1 x SPARC T7-2
    (2 x SPARC M7 @4.13 GHz)
    457,909 55,000 27,500 859 0.045 sec 0.257 sec
    3 x IBM S824 (each 2 x 8 active core
    LPARs, POWER8 @4.1 GHz)
    418,976 50,000 8,333 1041 0.031 sec 0.175 sec
    2 x SPARC T5-2
    (each with 2 x SPARC T5 @3.6 GHz)
    333,339 40,000 10,000 625 0.110 sec 0.608 sec

    TPH – Business transactions throughput per hour

    Configuration Summary

    Application Server:

    1 x SPARC T7-2 server with
    2 x SPARC M7 processors, 4.13 GHz
    1 TB memory
    6 x 300 GB SAS internal disks
    Oracle Solaris 11.3
    Siebel CRM 8.1.1.4 SIA

    Web/Database/Gateway Server:

    1 x SPARC T7-2 server with
    2 x SPARC M7 processors, 4.13 GHz (20 active cores: 8 cores for DB, 12 for Web/Gateway)
    512 TB memory
    6 x 300 GB SAS internal disks
    2 x 1.6 TB NVMe SSD
    Oracle Solaris 11.3
    Siebel CRM 8.1.1.4 SIA
    iPlanet Web Server 7
    Oracle Database 12c (12.1.0.2)

    Benchmark Description

    Siebel PSPP benchmark includes Call Center and Order Management:

    • Siebel Financial Services Call Center – Provides the most complete solution for sales and service, allowing customer service and telesales representatives to provide superior customer support, improve customer loyalty, and increase revenues through cross-selling and up-selling.

      High-level description of the use cases tested: Incoming Call Creates Opportunity, Quote and Order and Incoming Call Creates Service Request. Three complex business transactions are executed simultaneously for specific number of concurrent users. The ratios of these 3 scenarios were 30%, 40%, 30% respectively, which together were totaling 70% of all transactions simulated in this benchmark. Between each user operation and the next one, the think time averaged approximately 10, 13, and 35 seconds respectively.

    • Siebel Order Management – Oracle's Siebel Order Management allows employees such as salespeople and call center agents to create and manage quotes and orders through their entire life cycle. Siebel Order Management can be tightly integrated with back-office applications allowing users to perform tasks such as checking credit, confirming availability, and monitoring the fulfillment process.

      High-level description of the use cases tested: Order & Order Items Creation and Order Updates. Two complex Order Management transactions were executed simultaneously for specific number of concurrent users concurrently with aforementioned three Call Center scenarios above. The ratio of these 2 scenarios was 50% each, which together were totaling 30% of all transactions simulated in this benchmark. Between each user operation and the next one, the think time averaged approximately 20 and 67 seconds respectively.

    See Also

    Disclosure Statement

    Copyright 2016, 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 March 22, 2016.

    Thursday Nov 19, 2015

    SPECvirt_2013: SPARC T7-2 World Record Performance for Two- and Four-Chip Systems

    Oracle's SPARC T7-2 server delivered a world record SPECvirt_sc2013 result for systems with two to four chips.

    • The SPARC T7-2 server produced a result of 3198 @ 179 VMs SPECvirt_sc2013.

    • The two-chip SPARC T7-2 server beat the best four-chip x86 Intel E7-8890 v3 server (HP ProLiant DL580 Gen9), demonstrating that the SPARC M7 processor is 2.1 times faster than the Intel Xeon Processor E7-8890 v3 (chip-to-chip comparison).

    • The two-chip SPARC T7-2 server beat the best two-chip x86 Intel E5-2699 v3 server results by nearly 2 times (Huawei FusionServer RH2288H V3, HP ProLiant DL360 Gen9).

    • The two-chip SPARC T7-2 server delivered nearly 2.2 times the performance of the four-chip IBM Power System S824 server solution which used 3.5 GHz POWER8 six core chips.

    • The SPARC T7-2 server running Oracle Solaris 11.3 operating system, utilizes embedded virtualization products as the Oracle Solaris 11 zones, which in turn provide a low overhead, flexible, scalable and manageable virtualization environment.

    • The SPARC T7-2 server result used Oracle VM Server for SPARC 3.3 and Oracle Solaris Zones providing a flexible, scalable and manageable virtualization environment.

    Performance Landscape

    Complete benchmark results are at the SPEC website, SPECvirt_sc2013 Results. The following table highlights the leading two-, and four-chip results for the benchmark, bigger is better.

    SPECvirt_sc2013
    Leading Two to Four-Chip Results
    System
    Processor
    Chips Result @ VMs Virtualization Software
    SPARC T7-2
    SPARC M7 (4.13 GHz, 32core)
    2 3198 @ 179 Oracle VM Server for SPARC 3.3
    Oracle Solaris Zones
    HP ProLiant DL580 Gen9
    Intel E7-8890 v3 (2.5 GHz, 18core)
    4 3020 @ 168 Red Hat Enterprise Linux 7.1 KVM
    Lenovo System x3850 X6
    Intel E7-8890 v3 (2.5 GHz, 18core)
    4 2655 @ 147 Red Hat Enterprise Linux 6.6 KVM
    Huawei FusionServer RH2288H V3
    Intel E5-2699 v3 (2.3 GHz, 18core)
    2 1616 @ 95 Huawei FusionSphere V1R5C10
    HP ProLiant DL360 Gen9
    Intel E5-2699 v3 (2.3 GHz, 18core)
    2 1614 @ 95 Red Hat Enterprise Linux 7.1 KVM
    IBM Power S824
    POWER8 (3.5 GHz, 6core)
    4 1370 @ 79 PowerVM Enterprise Edition 2.2.3

    Configuration Summary

    System Under Test Highlights:

    Hardware:
    1 x SPARC T7-2 server, with
    2 x 4.13 GHz SPARC M7
    1 TB memory
    2 Sun Dual Port 10GBase-T Adapter
    2 Sun Storage Dual 16 Gb Fibre Channel PCIe Universal HBA

    Software:
    Oracle Solaris 11.3
    Oracle VM Server for SPARC 3.3 (LDom)
    Oracle Solaris Zones
    Oracle iPlanet Web Server 7.0.20
    Oracle PHP 5.3.29
    Dovecot v2.2.18
    Oracle WebLogic Server Standard Edition Release 10.3.6
    Oracle Database 12c Enterprise Edition (12.1.0.2.0)
    Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.7.0_85-b15

    Storage:
    3 x Oracle Server X5-2L, with
    2 x Intel Xeon Processor E5-2630 v3 8-core 2.4 GHz
    32 GB memory
    4 x Oracle Flash Accelerator F160 PCIe Card
    Oracle Solaris 11.3

    1 x Oracle Server X5-2L, with
    2 x Intel Xeon Processor E5-2630 v3 8-core 2.4 GHz
    32 GB memory
    4 x Oracle Flash Accelerator F160 PCIe Card
    4x 400 GB SSDs
    Oracle Solaris 11.3

    Benchmark Description

    SPECvirt_sc2013 is SPEC's updated benchmark addressing performance evaluation of datacenter servers used in virtualized server consolidation. SPECvirt_sc2013 measures the end-to-end performance of all system components including the hardware, virtualization platform, and the virtualized guest operating system and application software. It utilizes several SPEC workloads representing applications that are common targets of virtualization and server consolidation. The workloads were made to match a typical server consolidation scenario of CPU resource requirements, memory, disk I/O, and network utilization for each workload. These workloads are modified versions of SPECweb2005, SPECjAppServer2004, SPECmail2008, and SPEC CPU2006. The client-side SPECvirt_sc2013 harness controls the workloads. Scaling is achieved by running additional sets of virtual machines, called "tiles", until overall throughput reaches a peak.

    Key Points and Best Practices

    • The SPARC T7-2 server running the Oracle Solaris 11.3, utilizes embedded virtualization products as the Oracle VM Server for SPARC and Oracle Solaris Zones, which provide a low overhead, flexible, scalable and manageable virtualization environment.

    • In order to provide a high level of data integrity and availability, all the benchmark data sets are stored on mirrored (RAID1) storage

    • Using Oracle VM Server for SPARC to bind the SPARC M7 processor with its local memory optimized the memory use in this virtual environment.

    • This improved result used a fractional tile to fully saturate the system.

    See Also

    Disclosure Statement

    SPEC and the benchmark name SPECvirt_sc are registered trademarks of the Standard Performance Evaluation Corporation. Results from www.spec.org as of 11/19/2015. SPARC T7-2, SPECvirt_sc2013 3198@179 VMs; HP ProLiant DL580 Gen9, SPECvirt_sc2013 3020@168 VMs; Lenovo x3850 X6; SPECvirt_sc2013 2655@147 VMs; Huawei FusionServer RH2288H V3, SPECvirt_sc2013 1616@95 VMs; HP ProLiant DL360 Gen9, SPECvirt_sc2013 1614@95 VMs; IBM Power S824, SPECvirt_sc2013 1370@79 VMs.

    Monday Oct 26, 2015

    Real-Time Enterprise: SPARC T7-1 Faster Than x86 E5 v3

    A goal of the modern business is real-time enterprise where analytics are run simultaneously with transaction processing on the same system to provide the most effective decision making. Oracle Database 12c Enterprise Edition utilizing the In-Memory option is designed to have the same database able to perform transactions at the highest performance and to transform analytical calculations that once took days or hours to complete orders of magnitude faster.

    Oracle's SPARC M7 processor has deep innovations to take the real-time enterprise to the next level of performance. In this test both OLTP transactions and analytical queries were run in a single database instance using all of the same features of Oracle Database 12c Enterprise Edition utilizing the In-Memory option in order to compare the advantages of the SPARC M7 processor compared to a generic x86 processor. On both systems the OLTP and analytical queries both took about half of the processing load of the server.

    In this test Oracle's SPARC T7-1 server is compared to a two-chip x86 E5 v3 based server. On analytical queries the SPARC M7 processor is 8.2x faster than the x86 E5 v3 processor. Simultaneously on OLTP transactions the SPARC M7 processor is 2.9x faster than the x86 E5 v3 processor. In addition, the SPARC T7-1 server had better OLTP transactional response time than the x86 E5 v3 server.

    The SPARC M7 processor does this by using the Data Accelerator co-processor (DAX). DAX is not a SIMD instruction set, but rather an actual co-processor that offloads in-memory queries which frees the cores up for other processing. The DAX has direct access to the memory bus and can execute scans at near full memory bandwidth. Oracle makes the DAX API available to other applications, so this kind of acceleration is not just to the Oracle database, it is open.

    The results below were obtained running a set of OLTP transactions and analytic queries simultaneously against two schema: a real time online orders system and a related historical orders schema configured as a real cardinality database (RCDB) star schema. The in-memory analytics RCDB queries are executed using the Oracle Database 12c In-Memory columnar feature.

    • The SPARC T7-1 server and the x86 E5 v3 server both ran OLTP transactions and the in-memory analytics on the same database instance using Oracle Database 12c Enterprise Edition utilizing the In-Memory option.

    • The SPARC T7-1 server ran the in-memory analytics RCDB based queries 8.2x faster per chip than a two-chip x86 E5 v3 server on the 48 stream test.

    • The SPARC T7-1 server delivers 2.9x higher OLTP transaction throughput results per chip than a two-chip x86 E5 v3 server on the 48 stream test.

    Performance Landscape

    The table below compares the SPARC T7-1 server and 2-chip x86 E5 v3 server while running OLTP and in-memory analytics against tables in the same database instance. The same set of transactions and queries were executed on each system.

    Real-Time Enterprise Performance Chart
    48 RCDB DSS Streams, 224 OLTP users
    System OLTP Transactions Analytic Queries
    Trans Per
    Second
    Per Chip
    Advantage
    Average
    Response Time
    Queries Per
    Minute
    Per Chip
    Advantage
    SPARC T7-1
    1 x SPARC M7 (32core)
    338 K 2.9x 11 (msec) 267 8.2x
    x86 E5 v3 server
    2 x Intel E5-2699 v3 (2x 18core)
    236 K 1.0 12 (msec) 65 1.0

    The number of cores listed is per chip.
    The Per Chip Advantage it computed by normalizing to a single chip's performance

    Configuration Summary

    SPARC Server:

    1 X SPARC T7-1 server
    1 X SPARC M7 processor
    256 GB Memory
    Oracle Solaris 11.3
    Oracle Database 12c Enterprise Edition Release 12.1.0.2.10

    x86 Server:

    1 X Oracle Server X5-2L
    2 X Intel Xeon Processor E5-2699 v3
    256 GB Memory
    Oracle Linux 6 Update 5 (3.8.13-16.2.1.el6uek.x86_64)
    Oracle Database 12c Enterprise Edition Release 12.1.0.2.10

    Benchmark Description

    The Real-Time Enterprise benchmark simulates the demands of customers who want to simultaneously run both their OLTP database and the related historical warehouse DSS data that would be based on that OLTP data. It answers the question of how a system will perform when doing data analysis while at the same time executing real-time on-line transactions.

    The OLTP workload simulates an Order Inventory System that exercises both reads and writes with a potentially large number of users that stresses the lock management and connectivity, as well as, database access.

    The number of customers, orders and users is fully parametrized. This benchmark is base on 100 GB dataset, 15 million customers, 600 million orders and up to 580 users. The workload consists of a number of transaction types including show-expenses, part-cost, supplier-phone, low-inv, high-inv, update-price, update-phone, update-cost, and new-order.

    The real cardinality database (RCDB) schema was created to showcase the potential speedup one may see moving from on disk, row format data warehouse/Star Schema, to utilizing Oracle Database 12c's In-Memory feature for analytical queries.

    The workload consists of as many as 2,304 unique queries asking questions such as "In 2014, what was the total revenue of single item orders", or "In August 2013, how many orders exceeded a total price of $50". Questions like these can help a company see where to focus for further revenue growth or identify weaknesses in their offerings.

    RCDB scale factor 1050 represents a 1.05 TB data warehouse. It is transformed into a star schema of 1.0 TB, and then becomes 110 GB in size when loaded in memory. It consists of 1 fact table, and 4 dimension tables with over 10.5 billion rows. There are 56 columns with most cardinalities varying between 5 and 2,000, a primary key being an example of something outside this range.

    Two reports are generated: one for the OLTP-Perf workload and one for the RCDB DSS workload. For the analytical DSS workload, queries per minute and average query elapsed times are reported. For the OLTP-Perf workload, both transactions-per-seconds in thousands and OLTP average response times in milliseconds are reported.

    Key Points and Best Practices

    • This benchmark utilized the SPARC M7 processor's co-processor DAX for query acceleration.

    • All SPARC T7-1 server results were run with out-of-the-box tuning for Oracle Solaris.

    • All Oracle Server X5-2L system results were run with out of the box tunings for Oracle Linux except for the setting in /etc/sysctl.conf to get large pages for the Oracle Database:

      • vm.nr_hugepages=98304

    • To create an in memory area, the following was added to the init.ora:

        inmemory_size = 120g

    • An example of how to set a table to be in memory is below:

        ALTER TABLE CUSTOMER INMEMORY MEMCOMPRESS FOR QUERY HIGH

    See Also

    Disclosure Statement

    Copyright 2015, 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 25 October 2015.

    In-Memory Database: SPARC T7-1 Faster Than x86 E5 v3

    Fast analytics on large databases are critical to transforming key business processes. Oracle's SPARC M7 processors are specifically designed to accelerate in-memory analytics using Oracle Database 12c Enterprise Edition utilizing the In-Memory option. The SPARC M7 processor outperforms an x86 E5 v3 chip by up to 10.8x on analytics queries. In order to test real world deep analysis on the SPARC M7 processor a scenario with over 2,300 analytical queries was run against a real cardinality database (RCDB) star schema. This benchmark was audited by Enterprise Strategy Group (ESG). ESG is an IT research, analyst, strategy, and validation firm focused on the global IT community.

    The SPARC M7 processor does this by using Data Accelerator co-processor (DAX). DAX is not a SIMD instruction but rather an actual co-processor that offloads in-memory queries which frees the cores up for other processing. The DAX has direct access to the memory bus and can execute scans at near full memory bandwidth. Oracle makes the DAX API available to other applications, so this kind of acceleration not just for the Oracle database, it is open.

    • The SPARC M7 processor delivers up to a 10.8x Query Per Minute speedup per chip over the Intel Xeon Processor E5-2699 v3 when executing analytical queries using the In-Memory option of Oracle Database 12c.

    • Oracle's SPARC T7-1 server delivers up to a 5.4x Query Per Minute speedup over the 2-chip x86 E5 v3 server when executing analytical queries using the In-Memory option of Oracle Database 12c.

    • The SPARC T7-1 server delivers over 143 GB/sec of memory bandwidth which is up to 7x more than the 2-chip x86 E5 v3 server when the Oracle Database 12c is executing the same analytical queries against the RCDB.

    • The SPARC T7-1 server scanned over 48 billion rows per second through the database.

    • The SPARC T7-1 server compresses the on-disk RCDB star schema by around 6x when using the Memcompress For Query High setting (more information following below) and by nearly 10x compared to a standard data warehouse row format version of the same database.

    Performance Landscape

    The table below compares the SPARC T7-1 server and 2-chip x86 E5 v3 server. The x86 E5 v3 server single chip compares are from actual measurements against a single chip configuration.

    The number of cores is per chip, multiply by number of chips to get system total.

    RCDB Performance Chart
    2,304 Queries
    System Elapsed
    Seconds
    Queries Per
    Minute
    System
    Adv
    Chip
    Adv
    DB Memory
    Bandwidth
    SPARC T7-1
    1 x SPARC M7 (32core)
    381 363 5.4x 10.8x 143 GB/sec
    x86 E5 v3 server
    2 x Intel E5-2699 v3 (2x 18core)
    2059 67 1.0x 2.0x 20 GB/sec
    x86 E5 v3 server
    1 x Intel E5-2699 v3 (18core)
    4096 34 0.5x 1.0x 10 GB/sec

    Fused Decompress + Scan

    The In-Memory feature of Oracle Database 12c puts tables in columnar format. There are different levels of compression that can be applied. One of these is Oracle Zip (OZIP) which is used with the "MEMCOMPRESS FOR QUERY HIGH" setting. Typically when compression is applied to data, in order to operate on it, the data must be:

      (1) Decompressed
      (2) Written back to memory in uncompressed form
      (3) Scanned and the results returned.

    When OZIP is applied to the data inside of an In-Memory Columnar Unit (or IMCU, an N sized chunk of rows), the DAX is able to take this data in its compressed format and operate (scan) directly upon it, returning results in a single step. This not only saves on compute power by not having the CPU do the decompression step, but also on memory bandwidth as the uncompressed data is not put back into memory. Only the results are returned. To illustrate this, a microbenchmark was used which measured the amount of rows that could be scanned per second.

    SAE hpk-uperf

    Compression

    This performance test was run on a Scale Factor 1750 database, which represents a 1.75 TB row format data warehouse. The database is then transformed into a star schema which ends up around 1.1 TB in size. The star schema is then loaded in memory with a setting of "MEMCOMPRESS FOR QUERY HIGH", which focuses on performance with somewhat more aggressive compression. This memory area is a separate part of the System Global Area (SGA) which is defined by the database initialization parameter "inmemory_size". See below for an example. Here is a breakdown of each table in memory with compression ratios.

    Column Name Original Size
    (Bytes)
    In Memory
    Size (Bytes)
    Compression
    Ratio
    LINEORDER 1,103,524,528,128 178,586,451,968 6.2x
    DATE 11,534,336 1,179,648 9.8x
    PART 11,534,336 1,179,648 9.8x
    SUPPLIER 11,534,336 1,179,648 9.8x
    CUSTOMER 11,534,336 1,179,648 9.8x

    Configuration Summary

    SPARC Server:

    1 X SPARC T7-1 server
    1 X SPARC M7 processor
    512 GB memory
    Oracle Solaris 11.3
    Oracle Database 12c Enterprise Edition Release 12.1.0.2.13

    x86 Server:

    1 X Oracle Server X5-2L
    2 X Intel Xeon Processor E5-2699 v3
    512 GB memory
    Oracle Linux 6 Update 5 (3.8.13-16.2.1.el6uek.x86_64)
    Oracle Database 12c Enterprise Edition Release 12.1.0.2.13

    Benchmark Description

    The real cardinality database (RCDB) benchmark was created to showcase the potential speedup one may see moving from on disk, row format data warehouse/Star Schema, to utilizing Oracle Database 12c's In-Memory feature for analytical queries.

    The workload consists of 2,304 unique queries asking questions such as "In 2014, what was the total revenue of single item orders", or "In August 2013, how many orders exceeded a total price of $50". Questions like these can help a company see where to focus for further revenue growth or identify weaknesses in their offerings.

    RCDB scale factor 1750 represents a 1.75 TB data warehouse. It is transformed into a star schema of 1.1 TB, and then becomes 179 GB in size when loaded in memory. It consists of 1 fact table, and 4 dimension tables with over 10.5 billion rows. There are 56 columns with most cardinalities varying between 5 and 2,000, a primary key being an example of something outside this range.

    One problem with many industry standard generated databases is that as they have grown in size the cardinalities for the generated columns have become exceedingly unrealistic. For instance one industry standard benchmark uses a schema where at scale factor 1 TB it calls for the number of parts to be SF * 800,000. A 1 TB database that calls for 800 million unique parts is not very realistic. Therefore RCDB attempts to take some of these unrealistic cardinalities and size them to be more representative of at least a section of customer data. Obviously one cannot encompass every database in one schema, this is just an example.

    We carefully scaled each system so that the optimal number of users was run on each system under test so that we did not create artificial bottlenecks. Each user ran an equal number of queries and the same queries were run on each system, allowing for a fair comparison of the results.

    Key Points and Best Practices

    • This benchmark utilized the SPARC M7 processor's co-processor DAX for query acceleration.

    • All SPARC T7-1 server results were run with out of the box tuning for Oracle Solaris.

    • All Oracle Server X5-2L system results were run with out of the box tunings for Oracle Linux except for the setting in /etc/sysctl.conf to get large pages for the Oracle Database:

      • vm.nr_hugepages=64520

    • To create an in memory area, the following was added to the init.ora:

        inmemory_size = 200g

    • An example of how to set a table to be in memory is below:

        ALTER TABLE CUSTOMER INMEMORY MEMCOMPRESS FOR QUERY HIGH

    See Also

    Disclosure Statement

    Copyright 2015, 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 10/25/2015.

    SPECjEnterprise2010: SPARC T7-1 World Record with Single Application Server Using 1 to 4 Chips

    Oracle's SPARC T7-1 servers have set a world record for the SPECjEnterprise2010 benchmark for solutions using a single application server with one to four chips. The result of 25,818.85 SPECjEnterprise2010 EjOPS used two SPARC T7-1 servers, one server for the application tier and the other server for the database tier.

    • The SPARC T7-1 servers obtained a result of 25,093.06 SPECjEnterprise2010 EjOPS using encrypted data. This secured result used Oracle Advanced Security Transparent Data Encryption (TDE) for the application database tablespaces with the AES-256-CFB cipher. The network connection between the application server and the database server was also encrypted using the secure JDBC.

    • The SPARC T7-1 server solution delivered 34% more performance compared to the two-chip IBM x3650 M5 server result of 19,282.14 SPECjEnterprise2010 EjOPS.

    • The SPARC T7-1 server solution delivered 14% more performance compared to the four-chip IBM Power System S824 server result of 22,543.34 SPECjEnterprise2010 EjOPS.

    • The SPARC T7-1 server based results demonstrated 20% more performance compared to the Oracle Server X5-2 system result of 21,504.30 SPECjEnterprise2010 EjOPS. Oracle holds the top x86 two-chip application server SPECjEnterprise2010 result.

    • 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 1.8.0_60. The database server was configured with Oracle Database 12c Release 1.

    • For the secure result, the application data was encrypted in the Oracle database using the Oracle Advanced Security Transparent Data Encryption (TDE) feature. Hardware accelerated cryptography support in the SPARC M7 processor for the AES-256-CFB cipher was used to provide data security.

    • The benchmark performance using the secure SPARC T7-1 server configuration with encryption was less than 3% when compared to the peak result.

    • This result demonstrated less than 1 second average response times for all SPECjEnterprise2010 transactions and represents Jave EE 5.0 transactions generated by over 210,000 users.

    Performance Landscape

    Select single application server results. Complete benchmark results are at the SPEC website, SPECjEnterprise2010 Results.

    SPECjEnterprise2010 Performance Chart
    10/25/2015
    Submitter EjOPS* Java EE Server DB Server Notes
    Oracle 25,818.85 1 x SPARC T7-1
    1 x 4.13 GHz SPARC M7
    Oracle WebLogic 12c (12.1.3)
    1 x SPARC T7-1
    1 x 4.13 GHz SPARC M7
    Oracle Database 12c (12.1.0.2)
    -
    Oracle 25,093.06 1 x SPARC T7-1
    1 x 4.13 GHz SPARC M7
    Oracle WebLogic 12c (12.1.3)
    Network Data Encryption for JDBC
    1 x SPARC T7-1
    1 x 4.13 GHz SPARC M7
    Oracle Database 12c (12.1.0.2)
    Transparent Data Encryption
    Secure
    IBM 22,543.34 1 x IBM Power S824
    4 x 3.5 GHz POWER 8
    WebSphere Application Server V8.5
    1 x IBM Power S824
    4 x 3.5 GHz POWER 8
    IBM DB2 10.5 FP3
    -
    Oracle 21,504.30 1 x Oracle Server X5-2
    2 x 2.3 GHz Intel Xeon E5-2699 v3
    Oracle WebLogic 12c (12.1.3)
    1 x Oracle Server X5-2
    2 x 2.3 GHz Intel Xeon E5-2699 v3
    Oracle Database 12c (12.1.0.2)
    COD
    IBM 19,282.14 1 x System x3650 M5
    2 x 2.6 GHz Intel Xeon E5-2697 v3
    WebSphere Application Server V8.5
    1 x System x3850 X6
    4 x 2.8 GHz Intel Xeon E7-4890 v2
    IBM DB2 10.5 FP5
    -

    * SPECjEnterprise2010 EjOPS (bigger is better)

    The Cluster on Die (COD) mode is a BIOS setting that effectively splits the chip in half, making the operating system think it has twice as many chips as it does (in this case, four, 9 core chips). Intel has stated that COD is appropriate only for highly NUMA optimized workloads. Dell has shown that there is a 3.7x slower bandwidth to the other half of the chip split by COD.

    Configuration Summary

    Application Server:

    1 x SPARC T7-1 server, with
    1 x SPARC M7 processor (4.13 GHz)
    256 GB memory (16 x 16 GB)
    2 x 600 GB SAS HDD
    2 x 400 GB SAS SSD
    3 x Sun Dual Port 10 GbE PCIe 2.0 Networking card with Intel 82599 10 GbE Controller
    Oracle Solaris 11.3 (11.3.0.0.30)
    Oracle WebLogic Server 12c (12.1.3)
    Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.8.0_60

    Database Server:

    1 x SPARC T7-1 server, with
    1 x SPARC M7 processor (4.13 GHz)
    512 GB memory (16 x 32 GB)
    2 x 600 GB SAS HDD
    1 x Sun Dual Port 10 GbE PCIe 2.0 Networking card with Intel 82599 10 GbE Controller
    1 x Sun Storage 16 Gb Fibre Channel Universal HBA
    Oracle Solaris 11.3 (11.3.0.0.30)
    Oracle Database 12c (12.1.0.2)

    Storage Servers:

    1 x Oracle Server X5-2L (8-Drive), with
    2 x Intel Xeon Processor E5-2699 v3 (2.3 GHz)
    32 GB memory
    1 x Sun Storage 16 Gb Fibre Channel Universal HBA
    4 x 1.6 TB NVMe SSD
    2 x 600 GB SAS HDD
    Oracle Solaris 11.3 (11.3.0.0.30)
    1 x Oracle Server X5-2L (24-Drive), with
    2 x Intel Xeon Processor E5-2699 v3 (2.3 GHz)
    32 GB memory
    1 x Sun Storage 16 Gb Fibre Channel Universal HBA
    14 x 600 GB SAS HDD
    Oracle Solaris 11.3 (11.3.0.0.30)

    1 x Brocade 6510 16 Gb FC switch

    Benchmark Description

    SPECjEnterprise2010 is the third generation of the SPEC organization's J2EE end-to-end industry standard benchmark application. The SPECjEnterprise2010 benchmark has been 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
    Moreover, SPECjEnterprise2010 also heavily exercises all parts of the underlying infrastructure that make up the application environment, including hardware, JVM software, database software, JDBC drivers, and the system network.

    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

    • Four Oracle WebLogic server instances on the SPARC T7-1 server were hosted in 4 separate Oracle Solaris Zones.
    • 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 RT scheduling class.

    See Also

    Disclosure Statement

    SPEC and the benchmark name SPECjEnterprise are registered trademarks of the Standard Performance Evaluation Corporation. Results from www.spec.org as of 10/25/2015. SPARC T7-1, 25,818.85 SPECjEnterprise2010 EjOPS (unsecure); SPARC T7-1, 25,093.06 SPECjEnterprise2010 EjOPS (secure); Oracle Server X5-2, 21,504.30 SPECjEnterprise2010 EjOPS (unsecure); IBM Power S824, 22,543.34 SPECjEnterprise2010 EjOPS (unsecure); IBM x3650 M5, 19,282.14 SPECjEnterprise2010 EjOPS (unsecure);

    SPECvirt_sc2013: SPARC T7-2 World Record for 2 and 4 Chip Systems

    Oracle has had a new result accepted by SPEC as of November 19, 2015. This new result may be found here.

    Oracle's SPARC T7-2 server delivered a world record SPECvirt_sc2013 result for systems with two to four chips.

    • The SPARC T7-2 server produced a result of 3026 @ 168 VMs SPECvirt_sc2013.

    • The two-chip SPARC T7-2 server beat the best two-chip x86 Intel E5-2699 v3 server results by nearly 1.9 times (Huawei FusionServer RH2288H V3, HP ProLiant DL360 Gen9).

    • The two-chip SPARC T7-2 server delivered nearly 2.2 times the performance of the four-chip IBM Power System S824 server solution which used 3.5 GHz POWER8 six core chips.

    • The SPARC T7-2 server running Oracle Solaris 11.3 operating system, utilizes embedded virtualization products as the Oracle Solaris 11 zones, which in turn provide a low overhead, flexible, scalable and manageable virtualization environment.

    • The SPARC T7-2 server result used Oracle VM Server for SPARC 3.3 and Oracle Solaris Zones providing a flexible, scalable and manageable virtualization environment.

    Performance Landscape

    Complete benchmark results are at the SPEC website, SPECvirt_sc2013 Results. The following table highlights the leading two-, and four-chip results for the benchmark, bigger is better.

    SPECvirt_sc2013
    Leading Two to Four-Chip Results
    System
    Processor
    Chips Result @ VMs Virtualization Software
    SPARC T7-2
    SPARC M7 (4.13 GHz, 32core)
    2 3026 @ 168 Oracle VM Server for SPARC 3.3
    Oracle Solaris Zones
    HP DL580 Gen9
    Intel E7-8890 v3 (2.5 GHz, 18core)
    4 3020 @ 168 Red Hat Enterprise Linux 7.1 KVM
    Lenovo System x3850 X6
    Intel E7-8890 v3 (2.5 GHz, 18core)
    4 2655 @ 147 Red Hat Enterprise Linux 6.6 KVM
    Huawei FusionServer RH2288H V3
    Intel E5-2699 v3 (2.3 GHz, 18core)
    2 1616 @ 95 Huawei FusionSphere V1R5C10
    HP DL360 Gen9
    Intel E5-2699 v3 (2.3 GHz, 18core)
    2 1614 @ 95 Red Hat Enterprise Linux 7.1 KVM
    IBM Power S824
    POWER8 (3.5 GHz, 6core)
    4 1370 @ 79 PowerVM Enterprise Edition 2.2.3

    Configuration Summary

    System Under Test Highlights:

    Hardware:
    1 x SPARC T7-2 server, with
    2 x 4.13 GHz SPARC M7
    1 TB memory
    2 Sun Dual Port 10GBase-T Adapter
    2 Sun Storage Dual 16 Gb Fibre Channel PCIe Universal HBA

    Software:
    Oracle Solaris 11.3
    Oracle VM Server for SPARC 3.3 (LDom)
    Oracle Solaris Zones
    Oracle iPlanet Web Server 7.0.20
    Oracle PHP 5.3.29
    Dovecot v2.2.18
    Oracle WebLogic Server Standard Edition Release 10.3.6
    Oracle Database 12c Enterprise Edition (12.1.0.2.0)
    Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.7.0_85-b15

    Storage:
    3 x Oracle Server X5-2L, with
    2 x Intel Xeon Processor E5-2630 v3 8-core 2.4 GHz
    32 GB memory
    4 x Oracle Flash Accelerator F160 PCIe Card
    Oracle Solaris 11.3

    1 x Oracle Server X5-2L, with
    2 x Intel Xeon Processor E5-2630 v3 8-core 2.4 GHz
    32 GB memory
    4 x Oracle Flash Accelerator F160 PCIe Card
    4x 400 GB SSDs
    Oracle Solaris 11.3

    Benchmark Description

    SPECvirt_sc2013 is SPEC's updated benchmark addressing performance evaluation of datacenter servers used in virtualized server consolidation. SPECvirt_sc2013 measures the end-to-end performance of all system components including the hardware, virtualization platform, and the virtualized guest operating system and application software. It utilizes several SPEC workloads representing applications that are common targets of virtualization and server consolidation. The workloads were made to match a typical server consolidation scenario of CPU resource requirements, memory, disk I/O, and network utilization for each workload. These workloads are modified versions of SPECweb2005, SPECjAppServer2004, SPECmail2008, and SPEC CPU2006. The client-side SPECvirt_sc2013 harness controls the workloads. Scaling is achieved by running additional sets of virtual machines, called "tiles", until overall throughput reaches a peak.

    Key Points and Best Practices

    • The SPARC T7-2 server running the Oracle Solaris 11.3, utilizes embedded virtualization products as the Oracle VM Server for SPARC and Oracle Solaris Zones, which provide a low overhead, flexible, scalable and manageable virtualization environment.

    • In order to provide a high level of data integrity and availability, all the benchmark data sets are stored on mirrored (RAID1) storage

    • Using Oracle VM Server for SPARC to bind the SPARC M7 processor with its local memory optimized system memory use in this virtual environment.

    See Also

    Disclosure Statement

    SPEC and the benchmark name SPECvirt_sc are registered trademarks of the Standard Performance Evaluation Corporation. Results from www.spec.org as of 10/25/2015. SPARC T7-2, SPECvirt_sc2013 3026@168 VMs; HP DL580 Gen9, SPECvirt_sc2013 3020@168 VMs; Lenovo x3850 X6; SPECvirt_sc2013 2655@147 VMs; Huawei FusionServer RH2288H V3, SPECvirt_sc2013 1616@95 VMs; HP ProLiant DL360 Gen9, SPECvirt_sc2013 1614@95 VMs; IBM Power S824, SPECvirt_sc2013 1370@79 VMs.

    Oracle FLEXCUBE Universal Banking: SPARC T7-1 World Record

    Oracle's SPARC T7-1 servers running Oracle FLEXCUBE Universal Banking Release 12 along with Oracle Database 12c Enterprise Edition with Oracle Real Application Clusters on Oracle Solaris 11 produced record results for two processor solutions.

    • Two SPARC T7-1 servers each running Oracle FLEXCUBE Universal Banking Release 12 (v 12.0.1) and Oracle Real Application Clusters 12c database on Oracle Solaris 11 achieved record End of Year batch processing of 25 million accounts with 200 branches in 4 hrs 34 minutes (total of two processors).

    • A single SPARC T7-1 server running Oracle FLEXCUBE Universal Banking Release 12 processing 100 branches was able to complete the workload in similar time as the two node 200 branches End of Year workload, demonstrating good scaling of the application.

    • The customer representative workload for all 25 million accounts included saving accounts, current accounts, loans and TD accounts were created on the basis 25 million Customer IDs with 200 branches.

    • Oracle's SPARC M7 and T7 Servers running Oracle Solaris 11 with built-in Silicon Secured Memory with Oracle Database 12c can benefit global retail and corporate financial institutions who are running Oracle FLEXCUBE Universal Banking Release 12. The uniquely co-engineered Oracle software and hardware unlock unique agile capabilities demanded by modern business environments.

    • The SPARC T7-1 system and Oracle Solaris are able to provide a combination of uniquely essential characteristics that resonate with core values for a modern financial services institution.

    • The SPARC M7 processor based systems are capable of delivering higher performance and lower total cost of ownership (TCO) than older SPARC infrastructure, without introducing the unseen tax and risk of migrating applications away from older SPARC systems.

    Performance Landscape

    Oracle FLEXCUBE Universal Banking Release 12
    End of Year Batch Processing
    System Branches Time in Minutes
    2 x SPARC T7-1 200 274 (min)
    1 x SPARC T7-1 100 268 (min)

    Configuration Summary

    Systems Under Test:

    2 x SPARC T7-1 each with
    1 x SPARC M7 processor, 4.13 GHz
    256 GB memory
    Oracle Solaris 11.3 (11.3.0.27.0)
    Oracle Database 12c (RAC/ASM 12.1.0.2 BP7)
    Oracle FLEXCUBE Universal Banking Release 12

    Storage Configuration:

    Oracle ZFS Storage ZS4-4 appliance

    Benchmark Description

    The Oracle FLEXCUBE Universal Banking Release 12 benchmark models an actual customer bank with End of Cycle transaction batch jobs which typically execute during non-banking hours. This benchmark includes accrual for savings and term deposit accounts, interest capitalization for saving accounts, interest pay out for term deposit accounts and consumer load processing.

    This benchmark helps banks refine their infrastructure requirements for the volumes and scale of operations for business expansion. The end of cycle can be year, month or day, with year having the most processing followed by month and then day.

    See Also

    Disclosure Statement

    Copyright 2015, 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 25 October 2015.

    Oracle Communications ASAP – Telco Subscriber Activation: SPARC T7-2 World Record

    Oracle's SPARC T7-2 server delivered world record results on Oracle Communications ASAP. The SPARC T7-2 server ran Oracle Solaris 11 with Oracle Database 11g Release 2, Oracle WebLogic Server 12c and Oracle Communications ASAP version 7.2.

    • Running Oracle Communications ASAP, the SPARC T7-2 server delivered a world record result of 3,018 ASDLs/sec (atomic network activation actions).

    • Oracle's SPARC M7 processor delivered over 2.5 times the throughput per ASDL cost compared to the previous generation SPARC T5 processor.

    • The SPARC T7-2 server running a single instance of the Oracle Communications ASAP application, with both the application and database tiers consolidated onto a single machine, easily supported the service activation volumes of 3,018 ASDLs/sec which is representative of a typical mobile operator with more than 100 million subscribers.

    • Oracle Communications ASAP v7.2.0.4 delivered 35% higher throughput on the SPARC T7-2 server when compared to the SPARC T5-4 server.

    Performance Landscape

    All of the following results were run as part of this benchmark effort.

    ASAP 7.2.0.4 Test Results – 16 NEP
    Both tests used 1 cpu for App tier and 1 cpu for DB tier
    System ASDLs/sec CPU Usage CPU Cost per ASDL Cost Improvement Ratio
    SPARC T7-2 3,018.56 11.4% 1.10 2.6
    SPARC T5-4 2,238.97 29.6% 2.15 1.0

    CPU Cost per ASDL – computing cost per ASDL (smaller is better)
    Cost Improvement Ratio – improvement per cpu of SPARC T7-2 to SPARC T5-4

    Configuration Summary

    Hardware Configuration:

    SPARC T7-2 server
    2 x SPARC M7 processors (4.13 GHz)
    512 GB memory

    SPARC T5-4 server
    4 x SPARC T5 processors (3.6 GHz)
    512 GB memory

    Storage Configuration:

    Pillar Axiom

    Software Configuration:

    Oracle Communications ASAP 7.2.0.4.1
    Oracle Solaris 11.2
    Oracle Database 12c Release 12.1.0.1.0
    Oracle WebLogic Server 10.3.6.0
    Oracle JDK 7 update 75

    Benchmark Description

    Oracle Communications ASAP provides a convergent service activation platform that automatically activates customer services in a heterogeneous network and IT environment. It supports the activation of consumer and business services in fixed and mobile domains against network and IT applications.

    ASAP enables rapid service design and network technology introduction by means of its metadata-driven architecture, design-time configuration environment, and catalog of pre-built activation cartridges to reduce deployment time, cost, and risk. The application has been deployed for mobile (3G, 4G and M2M) services and fixed multi-play (broadband, voice, video, and IT) services in telecommunications, cable and satellite environments as well as for business voice, data, and IT cloud services.

    It may be deployed in a fully integrated manner as part of the Oracle Communications Service Fulfillment solution or directly integrated with third- party upstream systems. Market-proven for high-volume performance and scalability, Oracle Communications ASAP is deployed by more than 75 service providers worldwide and activates services for approximately 250 million subscribers globally.

    The throughput of ASAP is measured in atomic actions per second (or ASDLs/sec). An atomic action is a single command or operation that can be executed on a network element. Atomic actions are grouped together to form a common service action, where each service action typically relates to an orderable item, such as "GSM voice" or "voice mail" or "GSM data". One or more service actions are invoked by an order management system via an activation work order request.

    The workload resembles a typical mobile order to activate a GSM subscriber. A single service action to add a subscriber consists of seven atomic actions where each atomic action executes a command on a network element. Each network element was serviced by a dedicated Network Element Processor (NEP). The ASAP benchmark can vary the number of NEPs, which correlate to the complexity of a Telco operator's environment.

    See Also

    Disclosure Statement

    Copyright 2015, 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 25 October 2015.

    Oracle E-Business Suite Applications R12.1.3 (OLTP X-Large): SPARC M7-8 World Record

    Oracle's SPARC M7-8 server, using a four-chip Oracle VM Server for SPARC (LDom) virtualized server, produced a world record 20,000 users running the Oracle E-Business OLTP X-Large benchmark. The benchmark runs five Oracle E-Business online workloads concurrently: Customer Service, iProcurement, Order Management, Human Resources Self-Service, and Financials.

    • The virtualized four-chip LDom on the SPARC M7-8 was able to handle more users than the previous best result which used eight processors of Oracle's SPARC M6-32 server.

    • The SPARC M7-8 server using Oracle VM Server for SPARC provides enterprise applications high availability, where each application is executed on its own environment, insulated and independent of the others.

    Performance Landscape

    Oracle E-Business (3-tier) OLTP X-Large Benchmark
    System Chips Total Online Users Weighted Average
    Response Time (sec)
    90th Percentile
    Response Time (s)
    SPARC M7-8 4 20,000 0.70 1.13
    SPARC M6-32 8 18,500 0.61 1.16

    Break down of the total number of users by component.

    Users per Component
    Component SPARC M7-8 SPARC M6-32
    Total Online Users 20,000 users 18,500 users
    HR Self-Service
    Order-to-Cash
    iProcurement
    Customer Service
    Financial
    5000 users
    2500 users
    2700 users
    7000 users
    2800 users
    4000 users
    2300 users
    2400 users
    7000 users
    2800 users

    Configuration Summary

    System Under Test:

    SPARC M7-8 server
    8 x SPARC M7 processors (4.13 GHz)
    4 TB memory
    2 x 600 GB SAS-2 HDD
    using a Logical Domain with
    4 x SPARC M7 processors (4.13 GHz)
    2 TB memory
    2 x Sun Storage Dual 16Gb Fibre Channel PCIe Universal HBA
    2 x Sun Dual Port 10GBase-T Adapter
    Oracle Solaris 11.3
    Oracle E-Business Suite 12.1.3
    Oracle Database 11g Release 2

    Storage Configuration:

    4 x Oracle ZFS Storage ZS3-2 appliances each with
    2 x Read Flash Accelerator SSD
    1 x Storage Drive Enclosure DE2-24P containing:
    20 x 900 GB 10K RPM SAS-2 HDD
    4 x Write Flash Accelerator SSD
    1 x Sun Storage Dual 8Gb FC PCIe HBA
    Used for Database files, Zones OS, EBS Mid-Tier Apps software stack
    and db-tier Oracle Server
    2 x Sun Server X4-2L server with
    2 x Intel Xeon Processor E5-2650 v2
    128 GB memory
    1 x Sun Storage 6Gb SAS PCIe RAID HBA
    4 x 400 GB SSD
    14 x 600 GB HDD
    Used for Redo log files, db backup storage.

    Benchmark Description

    The Oracle E-Business OLTP X-Large benchmark simulates thousands of online users executing transactions typical of an internal Enterprise Resource Processing, simultaneously executing five application modules: Customer Service, Human Resources Self Service, iProcurement, Order Management and Financial.

    Each database tier uses a database instance of about 600 GB in size, supporting thousands of application users, accessing hundreds of objects (tables, indexes, SQL stored procedures, etc.).

    Key Points and Best Practices

    This test demonstrates virtualization technologies running concurrently various Oracle multi-tier business critical applications and databases on four SPARC M7 processors contained in a single SPARC M7-8 server supporting thousands of users executing a high volume of complex transactions with constrained (<1 sec) weighted average response time.

    The Oracle E-Business LDom is further configured using Oracle Solaris Zones.

    This result of 20,000 users was achieved by load balancing the Oracle E-Business Suite Applications 12.1.3 five online workloads across two Oracle Solaris processor sets and redirecting all network interrupts to a dedicated third processor set.

    Each applications processor set (set-1 and set-2) was running concurrently two Oracle E-Business Suite Application servers and two database servers instances, each within its own Oracle Solaris Zone (4 x Zones per set).

    Each application server network interface (to a client) was configured to map with the locality group associated to the CPUs processing the related workload, to guarantee memory locality of network structures and application servers hardware resources.

    All external storage was connected with at least two paths to the host multipath-capable fibre channel controller ports and Oracle Solaris I/O multipathing feature was enabled.

    See Also

    Disclosure Statement

    Oracle E-Business Suite R12 extra-large multiple-online module benchmark, SPARC M7-8, SPARC M7, 4.13 GHz, 4 chips, 128 cores, 1024 threads, 2 TB memory, 20,000 online users, average response time 0.70 sec, 90th percentile response time 1.13 sec, Oracle Solaris 11.3, Oracle Solaris Zones, Oracle VM Server for SPARC, Oracle E-Business Suite 12.1.3, Oracle Database 11g Release 2, Results as of 10/25/2015.

    PeopleSoft Enterprise Financials 9.2: SPARC T7-2 World Record

    Oracle's SPARC T7-2 server achieved world record performance being the first to publish on Oracle's PeopleSoft Enterprise Financials 9.2 benchmark. This result was obtained using one Oracle VM Server for SPARC (LDom) virtualized system configured with a single SPARC M7 processor.

    • The single processor LDom on the SPARC T7-2 server achieved world record performance executing 200 million Journal Lines in 18.60 minutes.

    • The single processor LDom on the SPARC T7-2 server was able to process General Ledger Journal Edit and Post batch jobs at 10,752,688 journal lines/min which reflects a large customer environment that utilizes a back-end database of nearly 1.0 TB performing highly competitive journal processing for Ledger.

    Performance Landscape

    Results are presented for PeopleSoft Financials Benchmark 9.2. Results obtained with PeopleSoft Financials Benchmark 9.2 are not comparable to the the previous version of the benchmark, PeopleSoft Financials Benchmark 9.1, due to significant change in data model and supports only batch.

    PeopleSoft Financials Benchmark, Version 9.2
    Solution Under Test Batch Journal lines/min
    SPARC T7-2 (using 1 x SPARC M7, 4.13 GHz) 18.60 min 10,752,688

    Configuration Summary

    System:

    SPARC T7-2 server with
    2 x SPARC M7 processors
    1 TB memory
    4 x Oracle Flash Accelerator F160 PCIe Card (DB Redo, DB undo and DB Data)
    4 x 600 GB internal disks
    Oracle Solaris 11.3
    Oracle Database 11g (11.2.0.4)
    PeopleSoft Financials (9.20.348)
    PeopleSoft PeopleTools (8.53.09)
    Java HotSpot 64-Bit Server VM (build 1.7.0_45-b18)
    Oracle Tuxedo, Version 11.1.1.3.0, 64-bit
    Oracle WebLogic Server 11g (10.3.6)

    LDom Under Test:

    Oracle VM Server for SPARC (LDom) virtualized server (APP & DB Tier)
    1 x SPARC M7 processor
    512 GB memory

    Benchmark Description

    The PeopleSoft Enterprise Financials 9.2 benchmark emulates a large enterprise that processes and validates a large number of financial journal transactions before posting the journal entry to the ledger. The validation process certifies that the journal entries are accurate, ensuring that ChartFields values are valid, debits and credits equal out, and inter/intra-units are balanced. Once validated, the entries are processed, ensuring that each journal line posts to the correct target ledger, and then changes the journal status to posted. In this benchmark, the Journal Edit & Post is set up to edit and post both Inter-Unit and Regular multi-currency journals. The benchmark processes 200 million journal lines using AppEngine for edits and Cobol for post processes.

    Key Points and Best Practices

    • The PeopleSoft Enterprise Financials 9.2 Batch benchmark ran on a one chip LDom consisting of 32 cores, each core had 8 threads. All total there were 256 virtual processors.

    • The LDom contained two Oracle Solaris Zones: a database tier zone and an application tier zone. The application tier zone consisted of 1 core with 8 virtual processors. The database tier zone consisted of 244 virtual processors from 31 cores. The remaining four virtual processors were dedicated to network and disk interrupt handling.

    • Inside of the database tier zone, the database log writer ran under 4 virtual processors and eight virtual processors were dedicated to four database writers.

    • There were 160 PeopleSoft Application instance processes running 320 streams of PeopleSoft Financial workload in the Oracle Solaris Fixed Priority FX class.

    See Also

    Disclosure Statement

    Copyright 2015, 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 25 October 2015.

    SAP Two-Tier Standard Sales and Distribution SD Benchmark: SPARC T7-2 World Record 2 Processors

    Oracle's SPARC T7-2 server produces a world record result for 2-processors on the SAP two-tier Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement Package 5 for SAP ERP 6.0 (2 chips / 64 cores / 512 threads).

    • The SPARC T7-2 server achieved 30,800 SAP SD benchmark users running the two-tier SAP Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement Package 5 for SAP ERP 6.0.

    • The SPARC T7-2 server achieved 1.9 times more users than the Dell PowerEdge R730 server result.

    • The SPARC T7-2 server achieved 1.5 times more users than the IBM Power System S824 server result.

    • The SPARC T7-2 server achieved 1.9 times more users than the HP ProLiant DL380 Gen9 server result.

    • The SPARC T7-2 server result was run with Oracle Solaris 11 and used Oracle Database 12c.

    Performance Landscape

    SAP-SD 2-tier performance table in decreasing performance order for leading two-processor systems and four-processor IBM Power System S824 server, with SAP ERP 6.0 Enhancement Package 5 for SAP ERP 6.0 results (current version of the benchmark as of May, 2012).

    SAP SD Two-Tier Benchmark
    System
    Processor
    OS
    Database
    Users Resp Time
    (sec)
    Version Cert#
    SPARC T7-2
    2 x SPARC M7 (2x 32core)
    Oracle Solaris 11
    Oracle Database 12c
    30,800 0.96 EHP5 2015050
    IBM Power S824
    4 x POWER8 (4x 6core)
    AIX 7
    DB2 10.5
    21,212 0.98 EHP5 2014016
    Dell PowerEdge R730
    2 x Intel E5-2699 v3 (2x 18core)
    Red Hat Enterprise Linux 7
    SAP ASE 16
    16,500 0.99 EHP5 2014033
    HP ProLiant DL380 Gen9
    2 x Intel E5-2699 v3 (2x 18core)
    Red Hat Enterprise Linux 6.5
    SAP ASE 16
    16,101 0.99 EHP5 2014032

    Version – Version of SAP, EHP5 refers to SAP ERP 6.0 Enhancement Package 5 for SAP ERP 6.0

    Number of cores presented are per chip, to get system totals, multiple by the number of chips.

    Complete benchmark results may be found at the SAP benchmark website http://www.sap.com/benchmark.

    Configuration Summary and Results

    Database/Application Server:

    1 x SPARC T7-2 server with
    2 x SPARC M7 processors (4.13 GHz, total of 2 processors / 64 cores / 512 threads)
    1 TB memory
    Oracle Solaris 11.3
    Oracle Database 12c

    Database Storage:
    3 x Sun Server X3-2L each with
    2 x Intel Xeon Processors E5-2609 (2.4 GHz)
    16 GB memory
    4 x Sun Flash Accelerator F40 PCIe Card
    12 x 3 TB SAS disks
    Oracle Solaris 11

    REDO log Storage:
    1 x Pillar FS-1 Flash Storage System, with
    2 x FS1-2 Controller (Netra X3-2)
    2 x FS1-2 Pilot (X4-2)
    4 x DE2-24P Disk enclosure
    96 x 300 GB 10000 RPM SAS Disk Drive Assembly

    Certified Results (published by SAP)

    Number of SAP SD benchmark users: 30,800
    Average dialog response time: 0.96 seconds
    Throughput:
      Fully processed order line items per hour: 3,372,000
      Dialog steps per hour: 10,116,000
      SAPS: 168,600
    Average database request time (dialog/update): 0.022 sec / 0.047 sec
    SAP Certification: 2015050

    Benchmark Description

    The SAP Standard Application SD (Sales and Distribution) Benchmark is an ERP business test that is indicative of full business workloads of complete order processing and invoice processing, and demonstrates the ability to run both the application and database software on a single system. The SAP Standard Application SD Benchmark represents the critical tasks performed in real-world ERP business environments.

    SAP is one of the premier world-wide ERP application providers, and maintains a suite of benchmark tests to demonstrate the performance of competitive systems on the various SAP products.

    See Also

    Disclosure Statement

    Two-tier SAP Sales and Distribution (SD) standard application benchmarks, SAP Enhancement Package 5 for SAP ERP 6.0 as of 10/23/15:

    SPARC T7-2 (2 processors, 64 cores, 512 threads) 30,800 SAP SD users, 2 x 4.13 GHz SPARC M7, 1 TB memory, Oracle Database 12c, Oracle Solaris 11, Cert# 2015050.
    IBM Power System S824 (4 processors, 24 cores, 192 threads) 21,212 SAP SD users, 4 x 3.52 GHz POWER8, 512 GB memory, DB2 10.5, AIX 7, Cert#2014016
    Dell PowerEdge R730 (2 processors, 36 cores, 72 threads) 16,500 SAP SD users, 2 x 2.3 GHz Intel Xeon Processor E5-2699 v3 256 GB memory, SAP ASE 16, RHEL 7, Cert#2014033
    HP ProLiant DL380 Gen9 (2 processors, 36 cores, 72 threads) 16,101 SAP SD users, 2 x 2.3 GHz Intel Xeon Processor E5-2699 v3 256 GB memory, SAP ASE 16, RHEL 6.5, Cert#2014032

    SAP, R/3, reg TM of SAP AG in Germany and other countries. More info www.sap.com/benchmark

    Friday Apr 03, 2015

    Oracle Server X5-2 Produces World Record 2-Chip Single Application Server SPECjEnterprise2010 Result

    Two Oracle Server X5-2 systems, using the Intel Xeon E5-2699 v3 processor, produced a World Record x86 two-chip single application server SPECjEnterprise2010 benchmark result of 21,504.30 SPECjEnterprise2010 EjOPS. One Oracle Server X5-2 ran the application tier and the second Oracle Server X5-2 was used for the database tier.

    • The Oracle Server X5-2 system demonstrated 11% better performance when compared to the IBM X3650 M5 server result of 19,282.14 SPECjEnterprise2010 EjOPS.

    • The Oracle Server X5-2 system demonstrated 1.9x better performance when compared to the previous generation Sun Server X4-2 server result of 11,259.88 SPECjEnterprise2010 EjOPS.

    • This result used Oracle WebLogic Server 12c, Java HotSpot(TM) 64-Bit Server 1.8.0_40 Oracle Database 12c, and Oracle Linux.

    Performance Landscape

    Complete benchmark results are at the SPEC website, SPECjEnterprise2010 Results. The table below shows the top single application server, two-chip x86 results.

    SPECjEnterprise2010 Performance Chart
    as of 4/1/2015
    Submitter EjOPS* Application Server Database Server
    Oracle 21,504.30 1x Oracle Server X5-2
    2x 2.3 GHz Intel Xeon E5-2699 v3
    Oracle WebLogic 12c (12.1.3)
    1x Oracle Server X5-2
    2x 2.3 GHz Intel Xeon E5-2699 v3
    Oracle Database 12c (12.1.0.2)
    IBM 19,282.14 1x IBM X3650 M5
    2x 2.6 GHz Intel Xeon E5-2697 v3
    WebSphere Application Server V8.5
    1x IBM X3850 X6
    4x 2.8 GHz Intel Xeon E7-4890 v2
    IBM DB2 10.5
    Oracle 11,259.88 1x Sun Server X4-2
    2x 2.7 GHz Intel Xeon E5-2697 v2
    Oracle WebLogic 12c (12.1.2)
    1x Sun Server X4-2L
    2x 2.7 GHz Intel Xeon E5-2697 v2
    Oracle Database 12c (12.1.0.1)

    * SPECjEnterprise2010 EjOPS, bigger is better.

    Configuration Summary

    Application Server:

    1 x Oracle Server X5-2
    2 x 2.3 GHz Intel Xeon E5-2699 v3 processors
    256 GB memory
    3 x 10 GbE NIC
    Oracle Linux 6 Update 5 (kernel-2.6.39-400.243.1.el6uek.x86_64)
    Oracle WebLogic Server 12c (12.1.3)
    Java HotSpot(TM) 64-Bit Server VM on Linux, version 1.8.0_40 (Java SE 8 Update 40)
    BIOS SW 1.2

    Database Server:

    1 x Oracle Server X5-2
    2 x 2.3 GHz Intel Xeon E5-2699 v3 processors
    512 GB memory
    2 x 10 GbE NIC
    1 x 16 Gb FC HBA
    2 x Oracle Server X5-2L Storage
    Oracle Linux 6 Update 5 (kernel-3.8.13-16.2.1.el6uek.x86_64)
    Oracle Database 12c Enterprise Edition Release 12.1.0.2

    Benchmark Description

    SPECjEnterprise2010 is the third generation of the SPEC organization's J2EE end-to-end industry standard benchmark application. The SPECjEnterprise2010 benchmark has been 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 workload consists of an end to end web based order processing domain, an RMI and Web Services driven manufacturing domain and a supply chain model utilizing document based Web Services. The application is a collection of Java classes, Java Servlets, Java Server Pages, Enterprise Java Beans, Java Persistence Entities (pojo's) and Message Driven Beans.

    The SPECjEnterprise2010 benchmark heavily exercises all parts of the underlying infrastructure that make up the application environment, including hardware, JVM software, database software, JDBC drivers, and the system network.

    The primary metric of the SPECjEnterprise2010 benchmark is jEnterprise Operations Per Second ("SPECjEnterprise2010 EjOPS"). This metric 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

    • Four Oracle WebLogic server instances were started using numactl binding 2 instances per chip.
    • Four Oracle database listener processes were started, 2 processes bound per processor.
    • Additional tuning information is in the report at http://spec.org.
    • COD (Cluster on Die) is enabled in the BIOS on the application server.

    See Also

    Disclosure Statement

    SPEC and the benchmark name SPECjEnterprise are registered trademarks of the Standard Performance Evaluation Corporation. Oracle Server X5-2, 21,504.30 SPECjEnterprise2010 EjOPS; IBM System X3650 M5, 19,282.14 SPECjEnterprise2010 EjOPS. Sun Server X4-2, 11,259.88 SPECjEnterprise2010 EjOPS; Results from www.spec.org as of 4/1/2015.

    Thursday Mar 27, 2014

    SPARC M6-32 Produces SAP SD Two-Tier Benchmark World Record for 32-Processor Systems

    Oracle's SPARC M6-32 server produced a world record result for 32-processors on the SAP two-tier Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement Package 5 for SAP ERP 6.0 (32 chips / 384 cores / 3072 threads).

    • SPARC M6-32 server achieved 140,000 SAP SD benchmark users with a low average dialog response time of 0.58 seconds running the SAP two-tier Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement package 5 for SAP ERP 6.0.

    • The SPARC M6-32 delivered 2.5 times more users than the IBM Power 780 result using SAP Enhancement Package 5 for SAP ERP 6.0. The IBM result also had 1.7 times worse average dialog response time compared to the SPARC M6-32 server result.

    • The SPARC M6-32 delivered 3.0 times more users than the Fujitsu PRIMEQUEST 2800E (with Intel Xeon E7-8890 v2 processors) result. The Fujitsu result also had 1.7 times worse average dialog response time compared to the SPARC M6-32 server result.

    • The SPARC M6-32 server solution was run with Oracle Solaris 11 and used Oracle Database 11g.

    Performance Landscape

    SAP-SD 2-Tier Performance Table (in decreasing performance order). With SAP ERP 6.0 Enhancement Package 4 for SAP ERP 6.0 (Old version of the benchmark, obsolete at the end of April, 2012), and SAP ERP 6.0 Enhancement Package 5 for SAP ERP 6.0 results (current version of the benchmark as of May, 2012).

    System
    Processor
    Ch / Co / Th — Memory
    OS
    Database
    Users Resp Time
    (sec)
    Version Cert#
    Fujitsu SPARC M10-4S
    SPARC64 X @3.0 GHz
    40 / 640 / 1280 — 10 TB
    Solaris 11
    Oracle 11g
    153,000 0.87 EHP5 2013014
    SPARC M6-32 Server
    SPARC M6 @3.6 GHz
    32 / 384 / 3072 — 16 TB
    Solaris 11
    Oracle 11g
    140,000 0.58 EHP5 2014008
    IBM Power 795
    POWER7 @4 GHz
    32 / 256 / 1024 — 4 TB
    AIX 7.1
    DB2 9.7
    126,063 0.98 EHP4 2010046
    IBM Power 780
    POWER7+ @3.72 GHz
    12 / 96 / 834 — 1536 GB
    AIX 7.1
    DB2 10
    57,024 0.98 EHP5 2012033
    Fujitsu PRIMEQUEST 2800E
    Intel Xeon E7-8890 v2 @2.8 GHz
    8 / 120 / 240 — 1024 GB
    Windows Server 2012 SE
    SQL Server 2012
    47,500 0.97 EHP5 2014003
    IBM Power 760
    POWER7+ @3.41 GHz
    8 / 48 / 192 — 1024 GB
    AIX 7.1
    DB2 10
    25,488 0.99 EHP5 2013004

    Version – Version of SAP, EHP5 refers to SAP ERP 6.0 Enhancement Package 5 for SAP ERP 6.0 and EHP4 refers to SAP ERP 6.0 Enhancement Package 4 for SAP ERP 6.0

    Ch / Co / Th – Total chips, coreas and threads

    Complete benchmark results may be found at the SAP benchmark website http://www.sap.com/benchmark.

    Configuration Summary and Results

    Hardware Configuration:

    1 x SPARC M6-32 server with
    32 x 3.6 GHz SPARC M6 processors (total of 32 processors / 384 cores / 3072 threads)
    16 TB memory
    6 x Sun Server X3-2L each with
    2 x Intel Xeon E5-2609 2.4 GHz Processors
    16 GB Memory
    4 x Flash Accelerator F40
    12 x 3 TB SAS disks
    2 x Sun Server X3-2L each with
    2 x Intel Xeon E5-2609 2.4 GHz Processors
    16 GB Memory
    1 x 8-Port 6Gbps SAS-2 RAID PCI Express HBA
    12 x 3 TB SAS disks

    Software Configuration:

    Oracle Solaris 11
    SAP Enhancement Package 5 for SAP ERP 6.0
    Oracle Database 11g Release 2

    Certified Results (published by SAP)

    Number of SAP SD benchmark users:
    140,000
    Average dialog response time:
    0.58 seconds
    Throughput:

      Fully processed order line items per hour:
    15,878,670
      Dialog steps per hour:
    47,636,000
      SAPS:
    793,930
    Average database request time (dialog/update):
    0.020 sec / 0.041 sec
    SAP Certification:
    2014008

    Benchmark Description

    The SAP Standard Application SD (Sales and Distribution) Benchmark is an ERP business test that is indicative of full business workloads of complete order processing and invoice processing, and demonstrates the ability to run both the application and database software on a single system. The SAP Standard Application SD Benchmark represents the critical tasks performed in real-world ERP business environments.

    SAP is one of the premier world-wide ERP application providers, and maintains a suite of benchmark tests to demonstrate the performance of competitive systems on the various SAP products.

    See Also

    Disclosure Statement

    Two-tier SAP Sales and Distribution (SD) standard application benchmarks, SAP Enhancement Package 5 for SAP ERP 6.0 as of 3/26/14:

    SPARC M6-32 (32 processors, 384 cores, 3072 threads) 140,000 SAP SD users, 32 x 3.6 GHz SPARC M6, 16 TB memory, Oracle Database 11g, Oracle Solaris 11, Cert# 2014008. Fujitsu SPARC M10-4S (40 processors, 640 cores, 1280 threads) 153,000 SAP SD users, 40 x 3.0 GHz SPARC65 X, 10 TB memory, Oracle Database 11g, Oracle Solaris 11, Cert# 2013014. IBM Power 780 (12 processors, 96 cores, 384 threads) 57,024 SAP SD users, 12 x 3.72 GHz IBM POWER7+, 1536 GB memory, DB210, AIX7.1, Cert#2012033. Fujitsu PRIMEQUEST 2800E (8 processors, 120 cores, 240 threads) 47,500 SAP SD users, 8 x 2.8 GHz Intel Xeon Processor E7-8890 v2, 1024 GB memory, SQL Server 2012, Windows Server 2012 Standard Edition, Cert# 2014003. IBM Power 760 (8 processors, 48 cores, 192 threads) 25,488 SAP SD users, 8 x 3.41 GHz IBM POWER7+, 1024 GB memory, DB2 10, AIX 7.1, Cert#2013004.

    Two-tier SAP Sales and Distribution (SD) standard application benchmarks, SAP Enhancement Package 4 for SAP ERP 6.0 as of 3/26/14:

    IBM Power 795 (32 processors, 256 cores, 1024 threads) 126,063 SAP SD users, 32 x 4 GHz IBM POWER7, 4 TB memory, DB2 9.7, AIX7.1, Cert#2010046.

    SAP, R/3, reg TM of SAP AG in Germany and other countries. More info www.sap.com/benchmark

    Wednesday Mar 05, 2014

    SPARC T5-2 Delivers World Record 2-Socket SPECvirt_sc2010 Benchmark

    Oracle's SPARC T5-2 server delivered a world record two-chip SPECvirt_sc2010 result of 4270 @ 264 VMs, establishing performance superiority in virtualized environments of the SPARC T5 processors with Oracle Solaris 11, which includes as standard virtualization products Oracle VM for SPARC and Oracle Solaris Zones.

    • The SPARC T5-2 server has 2.3x better performance than an HP BL620c G7 blade server (with two Westmere EX processors) which used VMware ESX 4.1 U1 virtualization software (best SPECvirt_sc2010 result on two-chip servers using VMware software).

    • The SPARC T5-2 server has 1.6x better performance than an IBM Flex System x240 server (with two Sandy Bridge processors) which used Kernel-based Virtual Machines (KVM).

    • This is the first SPECvirt_sc2010 result using Oracle production level software: Oracle Solaris 11.1, Oracle WebLogic Server 10.3.6, Oracle Database 11g Enterprise Edition, Oracle iPlanet Web Server 7 and Oracle Java Development Kit 7 (JDK). The only exception for the Dovecot mail server.

    Performance Landscape

    Complete benchmark results are at the SPEC website, SPECvirt_sc2010 Results. The following table highlights the leading two-chip results for the benchmark, bigger is better.

    SPECvirt_sc2010
    Leading Two-Chip Results
    System Processor Result @ VMs Virtualization Software
    SPARC T5-2 2 x SPARC T5, 3.6 GHz 4270 @ 264 Oracle VM Server for SPARC 3.0
    Oracle Solaris Zones
    IBM Flex System x240 2 x Intel E5-2690, 2.9 GHz 2741 @ 168 Red Hat Enterprise Linux 6.4 KVM
    HP Proliant BL6200c G7 2 x Intel E7-2870, 2.4 GHz 1878 @ 120 VMware ESX 4.1 U1

    Configuration Summary

    System Under Test Highlights:

    1 x SPARC T5-2 server, with
    2 x 3.6 GHz SPARC T5 processors
    1 TB memory
    Oracle Solaris 11.1
    Oracle VM Server for SPARC 3.0
    Oracle iPlanet Web Server 7.0.15
    Oracle PHP 5.3.14
    Dovecot 2.1.17
    Oracle WebLogic Server 11g (10.3.6)
    Oracle Database 11g (11.2.0.3)
    Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.7.0_51

    Benchmark Description

    The SPECvirt_sc2010 benchmark is SPEC's first benchmark addressing performance of virtualized systems. It measures the end-to-end performance of all system components that make up a virtualized environment.

    The benchmark utilizes several previous SPEC benchmarks which represent common tasks which are commonly used in virtualized environments. The workloads included are derived from SPECweb2005, SPECjAppServer2004 and SPECmail2008. Scaling of the benchmark is achieved by running additional sets of virtual machines until overall throughput reaches a peak. The benchmark includes a quality of service criteria that must be met for a successful run.

    Key Points and Best Practices

    • The SPARC T5 server running the Oracle Solaris 11.1, utilizes embedded virtualization products as the Oracle VM for SPARC and Oracle Solaris Zones, which provide a low overhead, flexible, scalable and manageable virtualization environment.

    • In order to provide a high level of data integrity and availability, all the benchmark data sets are stored on mirrored (RAID1) storage.

    See Also

    Disclosure Statement

    SPEC and the benchmark name SPECvirt_sc are registered trademarks of the Standard Performance Evaluation Corporation. Results from www.spec.org as of 3/5/2014. SPARC T5-2, SPECvirt_sc2010 4270 @ 264 VMs; IBM Flex System x240, SPECvirt_sc2010 2741 @ 168 VMs; HP Proliant BL620c G7, SPECvirt_sc2010 1878 @ 120 VMs.

    Thursday Jan 23, 2014

    SPARC T5-2 Delivers World Record 2-Socket Application Server for SPECjEnterprise2010 Benchmark

    Oracle's SPARC T5-2 servers have set the world record for the SPECjEnterprise2010 benchmark using two-socket application servers with a result of 17,033.54 SPECjEnterprise2010 EjOPS. The result used two SPARC T5-2 servers, one server for the application tier and the other server for the database tier.

    • The SPARC T5-2 server delivered 29% more performance compared to the 2-socket IBM PowerLinux server result of 13,161.07 SPECjEnterprise2010 EjOPS.

    • The two SPARC T5-2 servers have 1.2x better price performance than the two IBM PowerLinux 7R2 POWER7+ processor-based servers (based on hardware plus software configuration costs for both tiers). The price performance of the SPARC T5-2 server is $35.99 compared to the IBM PowerLinux 7R2 at $44.75.

    • The SPARC T5-2 server demonstrated 1.5x more performance compared to Oracle's x86-based 2-socket Sun Server X4-2 system (Ivy Bridge) result of 11,259.88 SPECjEnterprise2010 EjOPS. Oracle holds the top x86 2-socket application server SPECjEnterprise2010 result.

    • This SPARC T5-2 server result represents the best performance per socket for a single system in the application tier of 8,516.77 SPECjEnterprise2010 EjOPS per socket.

    • 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_45. The database server was configured with Oracle Database 12c Release 1.

    • This result demonstrated less than 1 second average response times for all SPECjEnterprise2010 transactions and represents Jave EE 5.0 transactions generated by 139,000 users.

    Performance Landscape

    Select 2-socket single application server results. Complete benchmark results are at the SPEC website, SPECjEnterprise2010 Results.

    SPECjEnterprise2010 Performance Chart
    1/22/2014
    Submitter EjOPS* Java EE Server DB Server
    Oracle 17,033.54 1 x SPARC T5-2
    2 x 3.6 GHz SPARC T5
    Oracle WebLogic 12c (12.1.2)
    1 x SPARC T5-2
    2 x 3.6 GHz SPARC T5
    Oracle Database 12c (12.1.0.1)
    IBM 13,161.07 1x IBM PowerLinux 7R2
    2 x 4.2 GHz POWER 7+
    WebSphere Application Server V8.5
    1x IBM PowerLinux 7R2
    2 x 4.2 GHz POWER 7+
    IBM DB2 10.1 FP2
    Oracle 11,259.88 1x Sun Server X4-2
    2 x 2.7 GHz Intel Xeon E5-2697 v2
    Oracle WebLogic 12c (12.1.2)
    1x Sun Server X4-2L
    2 x 2.7 GHz Intel Xeon E5-2697 v2
    Oracle Database 12c (12.1.0.1)

    * SPECjEnterprise2010 EjOPS (bigger is better)

    Configuration Summary

    Application Server:

    1 x SPARC T5-2 server, with
    2 x 3.6 GHz SPARC T5 processors
    512 GB memory
    2 x 10 GbE dual-port NIC
    Oracle Solaris 11.1 (11.1.13.6.0)
    Oracle WebLogic Server 12c (12.1.2)
    Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.7.0_45

    Database Server:

    1 x SPARC T5-2 server, with
    2 x 3.6 GHz SPARC T5 processors
    512 GB memory
    1 x 10 GbE dual-port NIC
    2 x 8 Gb FC HBA
    Oracle Solaris 11.1 (11.1.13.6.0)
    Oracle Database 12c (12.1.0.1)

    Storage Servers:

    2 x Sun Server X4-2L (24-Drive), with
    2 x 2.6 GHz Intel Xeon
    64 GB memory
    1 x 8 Gb FC HBA
    4 x Sun Flash Accelerator F80 PCI-E Cards
    Oracle Solaris 11.1

    Benchmark Description

    SPECjEnterprise2010 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
    Moreover, SPECjEnterprise2010 also heavily exercises all parts of the underlying infrastructure that make up the application environment, including hardware, JVM software, database software, JDBC drivers, and the system network.

    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

    • Two Oracle WebLogic server instances on the SPARC T5-2 server were hosted in 2 separate Oracle Solaris Zones.
    • 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 RT scheduling class.

    See Also

    Disclosure Statement

    SPEC and the benchmark name SPECjEnterprise are registered trademarks of the Standard Performance Evaluation Corporation. Results from www.spec.org as of 1/22/2014. SPARC T5-2, 17,033.54 SPECjEnterprise2010 EjOPS; IBM PowerLinux 7R2, 13,161.07 SPECjEnterprise2010 EjOPS; Sun Server X4-2, 11,259.88 SPECjEnterprise2010 EjOPS.

    The SPARC T5-2 configuration cost is the total application and database server hardware plus software. List price is $613,052 from http://www.oracle.com as of 1/22/2014. The IBM PowerLinux 7R2 configuration total hardware plus software list price is $588,970 based on public pricing from http://www.ibm.com as of 1/22/2014. Pricing does not include database storage hardware for IBM or Oracle.

    Thursday Sep 26, 2013

    SPARC T5-8 Delivers World Record Single Server SPECjEnterprise2010 Benchmark, Utilizes Virtualized Environment

    Oracle produced a world record single-server SPECjEnterprise2010 benchmark result of 36,571.36 SPECjEnterprise2010 EjOPS using one of Oracle's SPARC T5-8 servers for both the application and the database tier. Oracle VM Server for SPARC was used to virtualize the system to achieve this result.

    • The 8-chip SPARC T5 processor based server is 3.3x faster than the 8-chip IBM Power 780 server (POWER7+ processor based).

    • The SPARC T5-8 has 4.4x better price performance than the IBM Power 780, a POWER7+ processor based server (based on hardware plus software configuration costs). The price performance of the SPARC T5-8 server is $40.68 compared to the IBM Power 780 at $177.41. The IBM Power 780, POWER7+ based system has 1.2x 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 SPARC T5-8 virtualized price performance was also less than the low-end IBM PowerLinux 7R2 at $62.26.

    • The SPARC T5-8 server ran the Oracle Solaris 11.1 operating system and used Oracle VM Server for SPARC to consolidate ten Oracle WebLogic application server instances and one database server instance to achieve this result.

    • This result demonstrated sub-second average response times for all SPECjEnterprise2010 transactions and represents JEE 5.0 transactions generated by 299,000 users.

    • 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 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_25. The database server was configured with Oracle Database 12c Release 1.

    • The SPARC T5-8 server is 2.8x faster than a non-virtualized IBM POWER7+ based server result (one server for application and one server for database), the IBM PowerLinux 7R2 achieved 13,161.07 SPECjEnterprise2010 EjOPS.

    Performance Landscape

    SPECjEnterprise2010 Performance Chart
    Only Three Virtualized Results (App+DB on 1 Server) as of 9/23/2013
    Submitter EjOPS* Chips per Server Java EE Server & DB Server
    App DB
    Oracle 36,571.36 5 3 1 x SPARC T5-8
    8 chips, 128 cores, 3.6 GHz SPARC T5
    Oracle WebLogic 12c (12.1.2)
    Oracle Database 12c (12.1.0.1)
    Oracle 27,843.57 4 4 1 x SPARC T5-8
    8 chips, 128 cores, 3.6 GHz SPARC T5
    Oracle WebLogic 12c (12.1.1)
    Oracle Database 11g (11.2.0.3)
    IBM 10,902.30 4 4 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)

    Complete benchmark results are at the SPEC website, SPECjEnterprise2010 Results.

    Configuration Summary

    Oracle Summary

    Application and Database Server:

    1 x SPARC T5-8 server, with
    8 x 3.6 GHz SPARC T5 processors
    2 TB memory
    9 x 10 GbE dual-port NIC
    6 x 8 Gb dual-port HBA
    Oracle Solaris 11.1 SRU 10.5
    Oracle VM Server for SPARC
    Oracle WebLogic Server 12c (12.1.2)
    Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.7.0_25
    Oracle Database 12c (12.1.0.1)

    Storage Servers:

    6 x Sun Server X3-2L (12-Drive), with
    2 x 2.4 GHz Intel Xeon
    16 GB memory
    1 x 8 Gb FC HBA
    4 x Sun Flash Accelerator F40 PCI-E Card
    Oracle Solaris 11.1

    2 x Sun Storage 2540-M2 Array
    12 x 600 GB 15K RPM SAS HDD

    Switch Hardware:

    1 x Sun Network 10 GbE 72-port Top of Rack (ToR) Switch

    IBM Summary

    Application and Database Server:

    1 x IBM Power 780 server, with
    8 x 4.42 GHz POWER7+ processors
    786 GB memory
    6 x 10 GbE dual-port NIC
    3 x 8 Gb four-port HBA
    IBM AIX V7.1 TL2
    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

    Storage:

    2 x DS5324 Disk System with
    48 x 146 GB 15K E-DDM Disks

    1 x v7000 Disk Controller with
    16 x 400 GB SSD Disks

    Benchmark Description

    SPECjEnterprise2010 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
    Moreover, SPECjEnterprise2010 also heavily exercises all parts of the underlying infrastructure that make up the application environment, including hardware, JVM software, database software, JDBC drivers, and the system network.

    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

    • Ten Oracle WebLogic server instances on the SPARC T5-8 server were hosted in 10 separate Oracle Solaris Zones within a separate guest domain on 80 cores (5 cpu chips).
    • The database ran in a separate guest domain consisting of 47 cores (3 cpu chips). One core was reserved for the primary domain.
    • 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.

    See Also

    Disclosure Statement

    SPEC and the benchmark name SPECjEnterprise are registered trademarks of the Standard Performance Evaluation Corporation. Results from www.spec.org as of 9/23/2013. SPARC T5-8, 36,571.36 SPECjEnterprise2010 EjOPS (using Oracle VM for SPARC and 5+3 split); SPARC T5-8, 27,843.57 SPECjEnterprise2010 EjOPS (using Oracle Zones and 4+4 split); IBM Power 780, 10,902.30 SPECjEnterprise2010 EjOPS; IBM PowerLinux 7R2, 13,161.07 SPECjEnterprise2010 EjOPS. SPARC T5-8 server total hardware plus software list price is $1,487,792 from http://www.oracle.com as of 9/20/2013. IBM Power 780 server total hardware plus software cost of $1,934,162 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 whywebsphere.com/2013/04/29/weblogic-12c-on-oracle-sparc-t5-8-delivers-half-the-transactions-per-core-at-double-the-cost-of-the-websphere-on-ibm-power7/ retrieved 9/20/2013.

    Wednesday Sep 25, 2013

    Sun Server X4-2 Delivers Single App Server, 2-Chip x86 World Record SPECjEnterprise2010

    Oracle's Sun Server X4-2 and Sun Server X4-2L servers, using the Intel Xeon E5-2697 v2 processor, produced a world record x86 two-chip single application server SPECjEnterprise2010 benchmark result of 11,259.88 SPECjEnterprise2010 EjOPS. The Sun Server X4-2 ran the application tier and the Sun Server X4-2L was used for the database tier.

    • The 2-socket Sun Server X4-2 demonstrated 16% better performance when compared to the 2-socket IBM X3650 M4 server result of 9,696.43 SPECjEnterprise2010 EjOPS.

    • This result used Oracle WebLogic Server 12c, Java HotSpot(TM) 64-Bit Server 1.7.0_40, Oracle Database 12c, and Oracle Linux.

    Performance Landscape

    Complete benchmark results are at the SPEC website, SPECjEnterprise2010 Results. The table below shows the top single application server, two-chip x86 results.

    SPECjEnterprise2010 Performance Chart
    as of 9/22/2013
    Submitter EjOPS* Application Server Database Server
    Oracle 11,259.88 1x Sun Server X4-2
    2x 2.7 GHz Intel Xeon E5-2697 v2
    Oracle WebLogic 12c (12.1.2)
    1x Sun Server X4-2L
    2x 2.7 GHz Intel Xeon E5-2697 v2
    Oracle Database 12c (12.1.0.1)
    IBM 9,696.43 1x IBM X3650 M4
    2x 2.9 GHz Intel Xeon E5-2690
    WebSphere Application Server V8.5
    1x IBM X3650 M4
    2x 2.9 GHz Intel Xeon E5-2690
    IBM DB2 10.1
    Oracle 8,310.19 1x Sun Server X3-2
    2x 2.9 GHz Intel Xeon E5-2690
    Oracle WebLogic 11g (10.3.6)
    1x Sun Server X3-2L
    2x 2.9 GHz Intel Xeon E5-2690
    Oracle Database 11g (11.2.0.3)

    * SPECjEnterprise2010 EjOPS, bigger is better.

    Configuration Summary

    Application Server:

    1 x Sun Server X4-2
    2 x 2.7 GHz Intel Xeon processor E5-2697 v2
    256 GB memory
    4 x 10 GbE NIC
    Oracle Linux 5 Update 9 (kernel-2.6.39-400.124.1.el5uek)
    Oracle WebLogic Server 12c (12.1.2)
    Java HotSpot(TM) 64-Bit Server VM on Linux, version 1.7.0_40 (Java SE 7 Update 40)

    Database Server:

    1 x Sun Server X4-2L
    2 x 2.7 GHz Intel Xeon E5-2697 v2
    256 GB memory
    1 x 10 GbE NIC
    2 x FC HBA
    3 x Sun StorageTek 2540 M2
    Oracle Linux 5 Update 9 (kernel-2.6.39-400.124.1.el5uek)
    Oracle Database 12c Enterprise Edition Release 12.1.0.1

    Benchmark Description

    SPECjEnterprise2010 is the third generation of the SPEC organization's J2EE end-to-end industry standard benchmark application. The SPECjEnterprise2010 benchmark has been 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 workload consists of an end to end web based order processing domain, an RMI and Web Services driven manufacturing domain and a supply chain model utilizing document based Web Services. The application is a collection of Java classes, Java Servlets, Java Server Pages, Enterprise Java Beans, Java Persistence Entities (pojo's) and Message Driven Beans.

    The SPECjEnterprise2010 benchmark heavily exercises all parts of the underlying infrastructure that make up the application environment, including hardware, JVM software, database software, JDBC drivers, and the system network.

    The primary metric of the SPECjEnterprise2010 benchmark is jEnterprise Operations Per Second ("SPECjEnterprise2010 EjOPS"). This metric 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

    • Four Oracle WebLogic server instances were started using numactl binding 2 instances per chip.
    • Two Oracle database listener processes were started and each was bound to a separate chip.
    • Additional tuning information is in the report at http://spec.org.

    See Also

    Disclosure Statement

    SPEC and the benchmark name SPECjEnterprise are registered trademarks of the Standard Performance Evaluation Corporation. Sun Server X4-2, 11,259.88 SPECjEnterprise2010 EjOPS; Sun Server X3-2, 8,310.19 SPECjEnterprise2010 EjOPS; IBM System X3650 M4, 9,696.43 SPECjEnterprise2010 EjOPS. Results from www.spec.org as of 9/22/2013.

    Tuesday Sep 10, 2013

    Oracle ZFS Storage ZS3-4 Produces Best 2-Node Performance on SPECsfs2008 NFSv3

    The Oracle ZFS Storage ZS3-4 storage system delivered world record two-node performance on the SPECsfs2008 NFSv3 benchmark, beating results published on NetApp's dual-controller and four-node high-end FAS6240 storage systems.

    • The Oracle ZFS Storage ZS3-4 storage system delivered a world record two-node result of 450,702 SPECsfs2008_nfs.v3 Ops/sec with an Overall Response Time (ORT) of 0.70 msec on the SPECsfs2008 NFSv3 benchmark.

    • The Oracle ZFS Storage ZS3-4 storage system delivered 2.4x higher throughput than the dual-controller NetApp FAS6240 and 4.5x higher throughput than the dual-controller NetApp FAS3270 on the SPECsfs2008_nfs.v3 benchmark at less than half the list price of either result.

    • The Oracle ZFS Storage ZS3-4 storage system had 42 percent higher throughput than the four-node NetApp FAS6240 on the SPECsfs2008 NFSv3 benchmark.

    • The Oracle ZFS Storage ZS3-4 storage aystem has 54 percent better Overall Response Time than the 4-node NetApp FAS6240 on the SPECsfs2008 NFSv3 benchmark.

    Performance Landscape

    Two node results for SPECsfs2008_nfs.v3 presented (in decreasing SPECsfs2008_nfs.v3 Ops/sec order) along with other select results.

    Sponsor System Nodes Disks Throughput
    (Ops/sec)
    Overall Response
    Time (msec)
    Oracle ZS3-4 2 464 450,702 0.70
    IBM SONAS 1.2 2 1975 403,326 3.23
    NetApp FAS6240 4 288 260,388 1.53
    NetApp FAS6240 2 288 190,675 1.17
    EMC VG8 312 135,521 1.92
    Oracle 7320 2 136 134,140 1.51
    EMC NS-G8 100 110,621 2.32
    NetApp FAS3270 2 360 101,183 1.66

    Throughput SPECsfs2008_nfs.v3 Ops/sec — the Performance Metric
    Overall Response Time — the corresponding Response Time Metric
    Nodes — Nodes and Controllers are being used interchangeably

    Complete SPECsfs2008 benchmark results may be found at http://www.spec.org/sfs2008/results/sfs2008.html.

    Configuration Summary

    Storage Configuration:

    Oracle ZFS Storage ZS3-4 storage system in clustered configuration
    2 x Oracle ZFS Storage ZS3-4 controllers, each with
    8 x 2.4 GHz Intel Xeon E7-4870 processors
    2 TB memory
    2 x 10GbE NICs
    20 x Sun Disk shelves
    18 x shelves with 24 x 300 GB 15K RPM SAS-2 drives
    2 x shelves with 20 x 300 GB 15K RPM SAS-2 drives and 8 x 73 GB SAS-2 flash-enabled write-cache

    Benchmark Description

    SPECsfs2008 is the latest version of the Standard Performance Evaluation Corporation (SPEC) benchmark suite measuring file server throughput and response time, providing a standardized method for comparing performance across different vendor platforms. SPECsfs2008 results summarize the server's capabilities with respect to the number of operations that can be handled per second, as well as the overall latency of the operations. The suite is a follow-on to the SFS97_R1 benchmark, adding a CIFS workload, an updated NFSv3 workload, support for additional client platforms, and a new test harness and reporting/submission framework.

    See Also

    Disclosure Statement

    SPEC and SPECsfs are registered trademarks of Standard Performance Evaluation Corporation (SPEC). Results as of September 10, 2013, for more information see www.spec.org. Oracle ZFS Storage ZS3-4 Appliance 450,702 SPECsfs2008_nfs.v3 Ops/sec, 0.70 msec ORT, NetApp Data ONTAP 8.1 Cluster-Mode (4-node FAS6240) 260,388 SPECsfs2008_nfs.v3 Ops/Sec, 1.53 msec ORT, NetApp FAS6240 190,675 SPECsfs2008_nfs.v3 Ops/Sec, 1.17 msec ORT. NetApp FAS3270 101,183 SPECsfs2008_nfs.v3 Ops/Sec, 1.66 msec ORT.

    Nodes refer to the item in the SPECsfs2008 disclosed Configuration Bill of Materials that have the Processing Elements that perform the NFS Processing Function. These are the first item listed in each of disclosed Configuration Bill of Materials except for EMC where it is both the first and third items listed, and HP, where it is the second item listed as Blade Servers. The number of nodes is from the QTY disclosed in the Configuration Bill of Materials as described above. Configuration Bill of Materials list price for Oracle result of US$ 423,644. Configuration Bill of Materials list price for NetApp FAS3270 result of US$ 1,215,290. Configuration Bill of Materials list price for NetApp FAS6240 result of US$ 1,028,118. Oracle pricing from https://shop.oracle.com/pls/ostore/f?p=dstore:home:0, traverse to "Storage and Tape" and then to "NAS Storage". NetApp's pricing from http://www.netapp.com/us/media/na-list-usd-netapp-custom-state-new-discounts.html.

    Oracle ZFS Storage ZS3-2 Beats Comparable NetApp on SPECsfs2008 NFSv3

    Oracle ZFS Storage ZS3-2 storage system delivered outstanding performance on the SPECsfs2008 NFSv3 benchmark, beating results published on NetApp's fastest midrange platform, the NetApp FAS3270, the NetApp FAS6240 and the EMC Gateway NS-G8 Server Failover Cluster.

    • The Oracle ZFS Storage ZS3-2 storage system delivered 210,535 SPECsfs2008_nfs.v3 Ops/sec with an Overall Response Time (ORT) of 1.12 msec on the SPECsfs2008 NFSv3 benchmark.

    • The Oracle ZFS Storage ZS3-2 storage system delivered 10% higher throughput than the NetApp FAS6240 on the SPECsfs2008 NFSv3 benchmark.

    • The Oracle ZFS Storage ZS3-2 storage system has 52% higher throughput than the NetApp FAS3270 on the SPECsfs2008 NFSv3 benchmark.

    • The Oracle ZFS Storage ZS3-2 storage system has 5% better Overall Response Time than the NetApp FAS6240 on the SPECsfs2008 NFSv3 benchmark.

    • The Oracle ZFS Storage ZS3-2 storage system has 33% better Overall Response Time than the NetApp FAS3270 on the SPECsfs2008 NFSv3 benchmark.

    Performance Landscape

    Results for SPECsfs2008 NFSv3 (in decreasing SPECsfs2008_nfs.v3 Ops/sec order) for competitive systems.

    Sponsor System Throughput
    (Ops/sec)
    Overall Response
    Time (msec)
    Oracle ZS3-2 210,535 1.12
    NetApp FAS6240 190,675 1.17
    EMC VG8 135,521 1.92
    EMC NS-G8 110,621 2.32
    NetApp FAS3270 101,183 1.66
    NetApp FAS3250 100,922 1.76

    Throughput SPECsfs2008_nfs.v3 Ops/sec = the Performance Metric
    Overall Response Time = the corresponding Response Time Metric

    Complete SPECsfs2008 benchmark results may be found at http://www.spec.org/sfs2008/results/sfs2008.html.

    Configuration Summary

    Storage Configuration:

    Oracle ZFS Storage ZS3-2 storage system in clustered configuration
    2 x Oracle ZFS Storage ZS3-2 controllers, each with
    4 x 2.1 GHz Intel Xeon E5-2658 processors
    512 GB memory
    8 x Sun Disk shelves
    3 x shelves with 24 x 900 GB 10K RPM SAS-2 drives
    3 x shelves with 20 x 900 GB 10K RPM SAS-2 drives
    2 x shelves with 20 x 900 GB 10K RPM SAS-2 drives and 4 x 73 GB SAS-2 flash-enabled write-cache

    Benchmark Description

    SPECsfs2008 is the latest version of the Standard Performance Evaluation Corporation (SPEC) benchmark suite measuring file server throughput and response time, providing a standardized method for comparing performance across different vendor platforms. SPECsfs2008 results summarize the server's capabilities with respect to the number of operations that can be handled per second, as well as the overall latency of the operations. The suite is a follow-on to the SFS97_R1 benchmark, adding a CIFS workload, an updated NFSv3 workload, support for additional client platforms, and a new test harness and reporting/submission framework.

     

    See Also

    Disclosure Statement

    SPEC and SPECsfs are registered trademarks of Standard Performance Evaluation Corporation (SPEC). Results as of September 10, 2013, for more information see www.spec.org. Oracle ZFS Storage ZS3-2 Appliance 210,535 SPECsfs2008_nfs.v3 Ops/sec, 1.12 msec ORT, NetApp FAS6240 190,675 SPECsfs2008_nfs.v3 Ops/Sec, 1.17 msec ORT, EMC Celerra VG8 Server Failover Cluster, 2 Data Movers (1 stdby) / Symmetrix VMAX 135,521 SPECsfs2008_nfs.v3 Ops/Sec, 1.92 msec ORT, EMC Celerra Gateway NS-G8 Server Failover Cluster, 3 Datamovers (1 stdby) / Symmetrix V-Max 110,621 SPECsfs2008_nfs.v3 Ops/Sec, 2.32 msec ORT. NetApp FAS3270 101,183 SPECsfs2008_nfs.v3 Ops/Sec, 1.66 msec ORT. NetApp FAS3250 100,922 SPECsfs2008_nfs.v3 Ops/Sec, 1.76 msec ORT.

    Monday Jul 01, 2013

    Quick Note about Blog Posting from John

    Please have a look at John's Blog if you are interested in SPEC performance and how people use it in different ways to come to different conclusions.

    Wednesday May 01, 2013

    SPARC T5-8 Delivers Best Single System SPECjEnterprise2010 Benchmark, Beats IBM

    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.

    Performance Landscape

    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 (11.2.0.3)
    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)

    Configuration Summary

    Oracle Summary

    Application and Database Server:

    1 x SPARC T5-8 server, with
    8 x 3.6 GHz SPARC T5 processors
    2 TB memory
    5 x 10 GbE dual-port NIC
    6 x 8 Gb dual-port HBA
    Oracle Solaris 11.1 SRU 4.5
    Oracle WebLogic Server 12c (12.1.1)
    Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.7.0_15
    Oracle Database 11g (11.2.0.3)

    Storage Servers:

    6 x Sun Server X3-2L (12-Drive), with
    2 x 2.4 GHz Intel Xeon
    16 GB memory
    1 x 8 Gb FC HBA
    4 x Sun Flash Accelerator F40 PCI-E Card
    Oracle Solaris 11.1

    2 x Sun Storage 2540-M2 Array
    12 x 600 GB 15K RPM SAS HDD

    Switch Hardware:

    1 x Sun Network 10 GbE 72-port Top of Rack (ToR) Switch

    IBM Summary

    Application and Database Server:

    1 x IBM Power 780 server, with
    8 x 4.42 GHz POWER7+ processors
    786 GB memory
    6 x 10 GbE dual-port NIC
    3 x 8 Gb four-port HBA
    IBM AIX V7.1 TL2
    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

    Storage:

    2 x DS5324 Disk System with
    48 x 146GB 15K E-DDM Disks

    1 x v7000 Disk Controller with
    16 x 400GB SSD Disks

    Benchmark Description

    SPECjEnterprise2010 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
    Moreover, SPECjEnterprise2010 also heavily exercises all parts of the underlying infrastructure that make up the application environment, including hardware, JVM software, database software, JDBC drivers, and the system network.

    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 using the plgrp command.
    • 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.

    See Also

    Disclosure Statement

    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.

    Tuesday Mar 26, 2013

    SPARC T5-8 Delivers SPECjEnterprise2010 Benchmark World Record Performance

    Oracle produced a world record SPECjEnterprise2010 benchmark result of 57,422.17 SPECjEnterprise2010 EjOPS using Oracle's SPARC T5-8 server in the application tier and another SPARC T5-8 server for the database tier.

    • The SPARC T5-8 server demonstrated 3.4x better performance compared to an 8-socket IBM Power 780 server result of 16,646.34 SPECjEnterprise2010 EjOPS. The SPARC T5-8 is 3.7x less expensive for the application server hardware list cost compared to the IBM configuration.

    • The SPARC T5 processor demonstrated 1.7x better performance per core compared to the POWER7 processor used in the IBM Power 780 SPECjEnterprise2010 result.

    • The SPARC T5-8 server demonstrated 2.2x better performance compared to the Cisco UCS B440 M2 Blade Server result of 26,118.67 SPECjEnterprise2010 EjOPS.

    • The SPARC T5-8 servers used in the application and database tiers ran the Oracle Solaris 11.1 operating system.

    • The SPARC T5-8 server for the application tier used Oracle Solaris Zones to consolidate sixteen Oracle WebLogic Server instances to achieve this result.

    • This result demonstrated less than 1 second response time for all SPECjEnterprise2010 transactions, while demonstrating a sustained load of Java EE 5 transactions equivalent to 468,000 users.

    • The SPARC T5-8 application server used Oracle Fusion Middleware components including the Oracle WebLogic 12.1 application server and Oracle JDK 7 Update 15. The SPARC T5-8 database server was configured with Oracle Database 11g Release 2.

    • This result used six Sun Server X3-2L systems each configured with 4 x 400 GB Sun Flash Accelerator F40 PCIe Card devices as storage servers for the database files.

    • This result represents the best performance/socket for a single system in the application tier of 7,177.77 SPECjEnterprise2010 EjOPS per socket.

    • A single SPARC T5-8 server in the application tier producing 57,422.17 SPECjEnterprise2010 EjOPS can replace a total of 4x SPARC T4-4 servers that obtained 40,104.86 SPECjEnterprise2010 EjOPS. A single SPARC T5-8 server in the application tier producing 57,422.17 SPECjEnterprise2010 EjOPS can replace 6x SPARC T3-4 servers where each SPARC T3-4 server obtained 9,456.28 SPECjEnterprise2010 EjOPS.

    • Oracle Fusion Middleware provides a family of complete, integrated, hot pluggable and best-of-breed products known for enabling enterprise customers to create and run agile and intelligent business applications. Oracle WebLogic Server's on-going, record-setting Java application server performance demonstrates why so many customers rely on Oracle Fusion Middleware as their foundation for innovation.

    Performance Landscape

    Complete benchmark results are at the SPEC website, SPECjEnterprise2010 Results.

    SPECjEnterprise2010 Performance Chart
    as of 3/26/2013
    Submitter EjOPS* Java EE Server DB Server
    Oracle 57,422.17 1 x SPARC T5-8
    8 chips, 128 cores, 3.6 GHz SPARC T5
    Oracle WebLogic 12c (12.1.1)
    1 x SPARC T5-8
    8 chips, 128 cores, 3.6 GHz SPARC T5
    Oracle Database 11g (11.2.0.3)
    Oracle 40,104.86 4 x SPARC T4-4
    4 chips, 32 cores, 3.0 GHz SPARC T4
    Oracle WebLogic 11g (10.3.5)
    2 x SPARC T4-4
    4 chips, 32 cores, 3.0 GHz SPARC T4
    Oracle Database 11g (11.2.0.2)
    Oracle 27,150.05 1x Sun Server X2-8
    8x 2.4 GHz Intel Xeon E7-8870
    Oracle WebLogic 12c
    1x Sun Server X2-4
    4x 2.4 GHz Intel Xeon E7-4870
    Oracle Database 11g (11.2.0.2)
    Cisco 26,118.67 2 x Cisco UCS B440 M2
    4 chips, 40 cores, 2.4 GHz Xeon E7-4870
    Oracle WebLogic 11g (10.3.5)
    1 x Cisco UCS C460 M2
    4 chips, 40 cores, 2.4 GHz Xeon E7-4870
    Oracle Database 11g (11.2.0.2)
    IBM 16,646.34 1 x IBM Power 780
    8 chips, 64 cores, 3.86 GHz POWER7
    WebSphere Application Server V7.0
    1 x IBM Power 750 Express
    4 chips, 32 cores, 3.55 GHz POWER7
    IBM DB2 Universal Database 9.7

    * SPECjEnterprise2010 EjOPS (bigger is better)

    Configuration Summary

    Application Server:

    1 x SPARC T5-8 server, with
    8 x 3.6 GHz SPARC T5 processors
    2 TB memory
    8 x 10 GbE dual-port NIC
    Oracle Solaris 11.1 SRU 4.5
    Oracle WebLogic Server 12c (12.1.1)
    Oracle JDK 7 Update 15

    Database Server:

    1 x SPARC T5-8 server, with
    8 x 3.6 GHz SPARC T5 processors
    2 TB memory
    5 x 10 GbE dual-port NIC
    6 x 8 Gb FC dual-port HBA
    Oracle Solaris 11.1 SRU 4.5
    Oracle Database 11g Enterprise Edition Release 11.2.0.3

    Storage Servers:

    6 x Sun Server X3-2L (12-Drive), with
    2 x 2.4 GHz Intel Xeon
    16 GB memory
    1 x 8 Gb FC HBA
    4 x Sun Flash Accelerator F40 PCI-E Card
    Oracle Solaris 11.1

    2 x Sun Storage 2540-M2 Array
    12 x 600 GB 15K RPM SAS HDD

    Switch Hardware:

    1 x Sun Network 10 GbE 72-port Top of Rack (ToR) Switch

    Benchmark Description

    SPECjEnterprise2010 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
    Moreover, SPECjEnterprise2010 also heavily exercises all parts of the underlying infrastructure that make up the application environment, including hardware, JVM software, database software, JDBC drivers, and the system network.

    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

    • Sixteen Oracle WebLogic server instances on the SPARC T5-8 server were hosted in 16 separate Oracle Solaris Zones to demonstrate consolidation of multiple application servers.
    • Each Oracle Solaris Zone was bound to a separate processor set, each contained total 58 hardware strands. This was done to improve performance by using the physical memory closest to the processors to reduce memory access latency. The default set was used for network and disk interrupt handling.
    • The Oracle WebLogic application servers were executed in the FX scheduling class to improve performance by reducing the frequency of context switches.
    • The Oracle database processes were run in 8 processor sets using psrset(1M) and executed in the FX scheduling class. This improved performance by reducing memory access latency and reducing context switches.
    • The Oracle log writer process was run in a separate processor set containing a single core and run in the RT scheduling class. This insured that the log writer had the most efficient use of CPU resources.

    See Also

    Disclosure Statement

    SPEC and the benchmark name SPECjEnterprise are registered trademarks of the Standard Performance Evaluation Corporation. Results from www.spec.org as of 3/26/2013. SPARC T5-8, 57,422.17 SPECjEnterprise2010 EjOPS; SPARC T4-4, 40,104.86 SPECjEnterprise2010 EjOPS; Sun Server X2-8, 27,150.05 SPECjEnterprise2010 EjOPS; Cisco UCS B440 M2, 26,118.67 SPECjEnterprise2010 EjOPS; IBM Power 780, 16,646.34 SPECjEnterprise2010 EjOPS. SPARC T3-4 9456.28 SPECjEnterprise2010 EjOPS.

    SPARC T5-8 (SPARC T5-8 Server base package, 8xSPARC T5 16-core processors, 128x16GB-1066 DIMMS, 2x600GB 10K RPM 2.5. SAS-2 HDD, 4x Power Cables) List Price $268,742. IBM Power 780 (IBM Power 780:9179 Model MHB, 8x3.8GHz 16-core, 64x one processor activation, 4xCEC Enclosure with IBM Bezel, I/O Backplane and System Midplane,16x 0/32GB DDR3 Memory (4x8GB) DIMMS-1066MHz Power7 CoD Memory, 12x Activation of 1 GB DDR3 Power7 Memory, 5x Activation of 100GB DDR3 Power7 Memory, 1x Disk/Media Backplane. 2x 146.8GB SAS 15K RPM 2.5. HDD (AIX/Linux only), 4x AC Power Supply 1725W) List Price $992,023. Source: Oracle.com and IBM.com, collected 03/18/2013.

    SPARC T5-8 Realizes SAP SD Two-Tier Benchmark World Record for 8 Chip Systems

    Oracle's SPARC T5-8 server produced a world record result for systems with 8 processors on the two-tier SAP Sales and Distribution (SD) Standard Application Benchmark.

    • The SPARC T5-8 server achieved 40,000 users with running the two-tier SAP Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement package 5 for SAP ERP 6.0.

    • The SPARC T5-8 server is 57% faster than the IBM Power 760 8-chip running SAP Enhancement Package 5 for SAP ERP 6.0.

    • The SPARC T5-8 server delivers 5% more SAP users per chip than the IBM Power 780 12-chip running SAP Enhancement Package 5 for SAP ERP 6.0.

    • The SPARC T5-8 server solution was run with Oracle Solaris 11 and used Oracle Database 11g.

    Performance Landscape

    SAP-SD 2-Tier Performance Table (in decreasing performance order). SAP ERP 6.0 Enhancement Pack 5 for SAP ERP 6.0 results (New version of the benchmark as of May 2012).

    System OS
    Database
    Users SAPS SAP
    ERP/ECC
    Release
    Date
    SPARC T5-8 Server
    8x SPARC T5 @3.6 GHz, 2 TB
    Solaris 11
    Oracle 11g
    40,000 220,950 EHP5 for SAP
    ERP 6.0
    25-Mar-13
    IBM Power 760
    8xPOWER7+ @3.41 GHz, 1024 GB
    AIX 7.1
    DB2 10
    25,488 139,220 EHP5 for SAP
    ERP 6.0
    5-Feb-13

    SAP ERP 6.0 Enhancement Pack 4 for SAP ERP 6.0 Results
    (Old version of the benchmark, obsolete at the end of April, 2012)

    System OS
    Database
    Users SAPS SAP
    ERP/ECC
    Release
    Date
    IBM Power 795
    32xPOWER7 @4 GHz, 4 TB
    AIX 7.1
    DB2 9.7
    126,063 688,630 EHP4 for SAP
    ERP 6.0
    15-Nov-10
    SPARC Enterprise Server M9000
    64xSPARC64 VII @2.88 GHz, 1152 GB
    Solaris 10
    Oracle 10g
    32,000 175,600 EHP4 for SAP
    ERP 6.0
    18-Nov-09

    Complete benchmark results may be found at the SAP benchmark website http://www.sap.com/benchmark.

    Configuration Summary and Results

    Hardware Configuration:

    1 x SPARC T5-8 server with
    8 x 3.6 GHz SPARC T5 processors (total of 8 processors / 128 cores / 1024 threads)
    2 TB memory
    1 x Sun ZFS Storage 7420 appliance with
    72 x 600 GB 15K RPM 3.5" SAS-2 disk
    32 x 32 GB memory
    1 x Sun Fire X4270 M2 server configured as a COMSTAR device with
    10 x 2 TB 7.2K 3.5" SAS disk
    18 x 8 GB memory

    Software Configuration:

    Oracle Solaris 11
    SAP enhancement package 5 for SAP ERP 6.0
    Oracle Database 11g Release 2

    Certified Results (published by SAP)

    Performance:
    40,000 benchmark users
    SAP Certification:
    2013008

    Benchmark Description

    The SAP Standard Application SD (Sales and Distribution) Benchmark is a two-tier ERP business test that is indicative of full business workloads of complete order processing and invoice processing, and demonstrates the ability to run both the application and database software on a single system. The SAP Standard Application SD Benchmark represents the critical tasks performed in real-world ERP business environments.

    SAP is one of the premier world-wide ERP application providers, and maintains a suite of benchmark tests to demonstrate the performance of competitive systems on the various SAP products.

    See Also

    Disclosure Statement

    Two-tier SAP Sales and Distribution (SD) Standard Application benchmarks SAP Enhancement package 5 for SAP ERP 6.0 as of 3/26/13:

    SPARC T5-8 (8 processors, 128 cores, 1024 threads) 40,000 SAP SD users, 8 x 3.6 GHz SPARC T5, 2 TB memory, Oracle Database 11g, Oracle Solaris 11, Cert# 2013008. IBM Power 760 (8 processors, 48 cores, 192 threads) 25,488 SAP SD users, 8 x 3.41 GHz IBM POWER7+, 1024 GB memory, DB2 10, AIX 7.1, Cert#2013004.

    Two-tier SAP Sales and Distribution (SD) Standard Application benchmarks SAP Enhancement package 4 for SAP ERP 6.0 as of 4/30/12:

    IBM Power 795 (32 processors, 256 cores, 1024 threads) 126,063 SAP SD users, 32 x 4 GHz IBM POWER7, 4 TB memory, DB2 9.7, AIX7.1, Cert#2010046. SPARC Enterprise Server M9000 (64 processors, 256 cores, 512 threads) 32,000 SAP SD users, 64 x 2.88 GHz SPARC64 VII, 1152 GB memory, Oracle Database 10g, Oracle Solaris 10, Cert# 2009046.

    SAP, R/3, reg TM of SAP AG in Germany and other countries. More info www.sap.com/benchmark

    SPARC M5-32 Produces SAP SD Two-Tier Benchmark World Record for SAP Enhancement Package 5 for SAP ERP 6.0

    Oracle's SPARC M5-32 server produced a world record result on the two-tier SAP Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement package 5 for SAP ERP 6.0.

    • The SPARC M5-32 server achieved 85,050 users running the two-tier SAP Sales and Distribution (SD) Standard Application Benchmark using SAP Enhancement package 5 for SAP ERP 6.0.

    • The SPARC M5-32 solution was run with Oracle Solaris 11 and used the Oracle Database 11g.

    Performance Landscape

    SAP-SD 2-Tier Performance Table (in decreasing performance order). SAP ERP 6.0 Enhancement Pack 5 for SAP ERP 6.0 results (new version of the benchmark as of May, 2012).

    System OS
    Database
    Users SAPS SAP
    ERP/ECC
    Release
    Date
    SPARC M5-32 Server
    32x SPARC M5 @3.6 GHz, 4 TB
    Solaris 11
    Oracle 11g
    85,050 472,600 EHP5 for SAP
    ERP 6.0
    25-Mar-13
    IBM Power 780
    12xPOWER7+ @3.72 GHz, 1536 GB
    AIX 7.1
    DB2 10
    57,024 311,720 EHP5 for SAP
    ERP 6.0
    3-Oct-12
    IBM Power 760
    8xPOWER7+ @3.41 GHz, 1024 GB
    AIX 7.1
    DB2 10
    25,488 139,220 EHP5 for SAP
    ERP 6.0
    5-Feb-13

    SAP ERP 6.0 Enhancement Pack 4 for SAP ERP 6.0 Results
    (Old version of the benchmark, obsolete at the end of April, 2012)

    System OS
    Database
    Users SAPS SAP
    ERP/ECC
    Release
    Date
    IBM Power 795
    32xPOWER7 @4 GHz, 4 TB
    AIX 7.1
    DB2 9.7
    126,063 688,630 EHP4 for SAP
    ERP 6.0
    15-Nov-10
    SPARC Enterprise Server M9000
    64xSPARC64 VII @2.88 GHz, 1152 GB
    Solaris 10
    Oracle 10g
    32,000 175,600 EHP4 for SAP
    ERP 6.0
    18-Nov-09

    Complete benchmark results may be found at the SAP benchmark website http://www.sap.com/benchmark.

    Configuration Summary and Results

    Hardware Configuration:

    1 x SPARC M5-32 server with
    32 x 3.6 GHz SPARC M5 processors (total of 32 processors / 192 cores / 1536 threads)
    4 TB memory
    1 x Sun Storage 2540-M2 (12 x 300 GB 5K RPM 3.5" SAS-2 disk & 2 GB cache)
    Flash Storage

    Software Configuration:

    Oracle Solaris 11
    SAP enhancement package 5 for SAP ERP 6.0
    Oracle Database 11g Release 2

    Certified Results (published by SAP)

    Performance: 85,050 benchmark users
    SAP Certification: 2013009

    Benchmark Description

    The SAP Standard Application SD (Sales and Distribution) Benchmark is a two-tier ERP business test that is indicative of full business workloads of complete order processing and invoice processing, and demonstrates the ability to run both the application and database software on a single system. The SAP Standard Application SD Benchmark represents the critical tasks performed in real-world ERP business environments.

    SAP is one of the premier world-wide ERP application providers, and maintains a suite of benchmark tests to demonstrate the performance of competitive systems on the various SAP products.

    See Also

    Disclosure Statement

    Two-tier SAP Sales and Distribution (SD) standard application benchmarks, SAP Enhancement package 5 for SAP ERP 6.0 as of 3/26/13:

    SPARC M5-32 (32 processors, 192 cores, 1536 threads) 85,050 SAP SD users, 32 x 3.6 GHz SPARC M5, 4 TB memory, Oracle Database 11g, Oracle Solaris 11, Cert# 2013009. IBM Power 780 (12 processors, 96 cores, 384 threads) 57,024 SAP SD users, 12 x 3.72 GHz IBM POWER7+, 1536 GB memory, DB210, AIX7.1, Cert#2012033. IBM Power 760 (8 processors, 48 cores, 192 threads) 25,488 SAP SD users, 8 x 3.41 GHz IBM POWER7+, 1024 GB memory, DB2 10, AIX 7.1, Cert#2013004.

    Two-tier SAP Sales and Distribution (SD) standard application benchmarks, SAP Enhancement package 4 for SAP ERP 6.0 as of 3/26/13:

    IBM Power 795 (32 processors, 256 cores, 1024 threads) 126,063 SAP SD users, 32 x 4 GHz IBM POWER7, 4 TB memory, DB2 9.7, AIX7.1, Cert#2010046. SPARC Enterprise Server M9000 (64 processors, 256 cores, 512 threads) 32,000 SAP SD users, 64 x 2.88 GHz SPARC64 VII, 1152 GB memory, Oracle Database 10g, Oracle Solaris 10, Cert# 2009046.

    SAP, R/3, reg TM of SAP AG in Germany and other countries. More info www.sap.com/benchmark

    SPARC T5-2 Achieves JD Edwards EnterpriseOne Benchmark World Records

    Oracle produced World Record batch throughput for single system results on Oracle's JD Edwards EnterpriseOne Day-in-the-Life benchmark using Oracle's SPARC T5-2 server running Oracle Solaris Containers and consolidating JD Edwards EnterpriseOne, Oracle WebLogic servers and the Oracle Database 11g Release 2. There are two workloads tested: online plus batch workload and batch-only workload.

    Online plus batch workload:

    • The SPARC T5-2 server delivered a result of 12,000 online users at 180 msec average response time while concurrently executing a mix of JD Edwards EnterpriseOne long and short batch processes at 198.5 UBEs/min (Universal Batch Engines per minute).

    • The SPARC T5-2 server online plus batch throughput is 2.7x higher than the IBM Power 770 server, both running 12,000 online users.

    • The SPARC T5-2 server online plus batch throughput is 6x higher per chip than the IBM Power 770 server. The SPARC T5-2 server has 2 chips and the IBM Power 770 has 4 chips, both ran 12,000 online users.

    • The SPARC T5-2 server online plus batch throughput is 3x higher per core than the IBM Power 770 server. Both servers have 32 cores and ran 12,000 online users.

    Batch-only workload:

    • The SPARC T5-2 server delivered throughput of 880 UBEs/min while executing the batch-only workload (Long and Short batch processes).

    • The SPARC T5-2 server batch-only throughput is 2.7x faster per chip than the IBM Power 770 server. The SPARC T5-2 server has 2 chips and the IBM Power 770 has 4 chips.

    • The SPARC T5-2 server batch-only throughput is 1.4x higher per core than the IBM Power 770 server. Both servers have 32 cores.

    • The SPARC T5-2 server batch-only throughput is 61% faster than the Cisco multiple system solution.

    • The SPARC T5-2 server batch-only throughput is 5x faster per chip than the Cisco UCS B200/B250 M2 servers. The SPARC T5-2 server has 2 chips and the Cisco 3 server solution has 6 chips.

    • The SPARC T5-2 server batch-only throughput is 18x higher per core than the Cisco UCS B200/B250 M2 servers. The SPARC T5-2 server has 32 cores while the Cisco solution utilized 36 cores.

    Both workloads:

    • The SPARC T5-2 server offers a 5.4x cost savings for the application server when compared to the IBM Power 770 application server.

    • The SPARC T5-2 server running Oracle Solaris Containers and consolidating JD Edwards EnterpriseOne, Oracle WebLogic servers and the Oracle Database 11g Release 2 utilized a maximum 65% of the available CPU power, leaving headroom for additional processing.

    • The database server in a shared-server configuration allows for optimized CPU resource utilization and significant memory savings on the SPARC T5-2 server without sacrificing performance.

    Performance Landscape

    JD Edwards EnterpriseOne Day in the Life (DIL) Benchmark
    Consolidated Online with Batch Workload
    System Rack
    Units (U)
    Batch
    Rate
    (UBEs/min)
    Online
    Users
    Users/
    U
    UBEs/
    Core
    UBEs/
    Chip
    Version
    SPARC T5-2 (2 x SPARC T5, 3.6 GHz) 3 198.5 12000 4000 6.2 99 9.0.2
    IBM Power 770 (4 x POWER7, 3.3 GHz) 8 65 12000 1500 2.0 16 9.0.2

    Batch Rate (UBEs/min) — Batch transaction rate in UBEs per minute.

    JD Edwards EnterpriseOne Batch Only Benchmark
    System Rack
    Units (U)
    Batch
    Rate
    (UBEs/min)
    UBEs/
    U
    UBEs/
    Core
    UBEs/
    Chip
    Version
    SPARC T5-2 (2 x SPARC T5, 3.6 GHz) 3 880 267 25 440 9.0.2
    IBM Power 770 (4 x POWER7, 3.3 GHz) 8 643 81 20 161 9.0.2
    2 x Cisco B200 M2 (2 x X5690, 3.46 GHz)
    1 x Cisco B250 M2 (2 x X5680, 3.33 GHz)
    3 546 182 15 91 9.0.2

    Configuration Summary

    Hardware Configuration:

    1 x SPARC T5-2 server with
    2 x SPARC T5 processors, 3.6 GHz
    512 GB memory
    4 x 300 GB 10K RPM SAS internal disk
    2 x 300 GB internal SSD
    4 x Sun Flash Accelerator F40 PCIe Card (4 x 93 GB)

    Software Configuration:

    Oracle Solaris 10 1/13
    Oracle Solaris Containers
    JD Edwards EnterpriseOne 9.0.2
    JD Edwards EnterpriseOne Tools (8.98.4.2)
    Oracle WebLogic Server 11g (10.3.4)
    Oracle HTTP Server 11g
    Oracle Database 11g Release 2 (11.2.0.3)

    Benchmark Description

    JD Edwards EnterpriseOne is an integrated applications suite of Enterprise Resource Planning (ERP) software. Oracle offers 70 JD Edwards EnterpriseOne application modules to support a diverse set of business operations.

    Oracle's Day in the Life (DIL) kit is a suite of scripts that exercises most common transactions of JD Edwards EnterpriseOne applications, including business processes such as payroll, sales order, purchase order, work order, and manufacturing processes, such as ship confirmation. These are labeled by industry acronyms such as SCM, CRM, HCM, SRM and FMS. The kit's scripts execute transactions typical of a mid-sized manufacturing company.

    • The workload consists of online transactions and the UBE – Universal Business Engine workload of 61 short and 4 long UBEs.

    • LoadRunner runs the DIL workload, collects the user’s transactions response times and reports the key metric of Combined Weighted Average Transaction Response time.

    • The UBE processes workload runs from the JD Enterprise Application server.

      • Oracle's UBE processes come as three flavors:
        • Short UBEs < 1 minute engage in Business Report and Summary Analysis,
        • Mid UBEs > 1 minute create a large report of Account, Balance, and Full Address,
        • Long UBEs > 2 minutes simulate Payroll, Sales Order, night only jobs.
      • The UBE workload generates large numbers of PDF files reports and log files.
      • The UBE Queues are categorized as the QBATCHD, a single threaded queue for large and medium UBEs, and the QPROCESS queue for short UBEs run concurrently.

    Oracle's UBE process performance metric is Number of Maximum Concurrent UBE processes at transaction rate, UBEs/minute.

    Key Points and Best Practices

    Four Oracle Solaris processors sets were used with Oracle Solaris Containers assigned to the processor sets as follows:

    • one JD Edwards EnterpriseOne Application server, two Oracle WebLogic Servers 11g Release 1 each coupled with an Oracle Web Tier HTTP server instances (online workload), each in an Oracle Solaris Container (three total),

    • one JD Edwards EnterpriseOne Application server (for batch only workload) in an Oracle Solaris Container,

    • Oracle Database 11g Release 2.0.3 database in an Oracle Solaris Container,

    • the Oracle database log writer.

    Other items of note:

    • Each Oracle WebLogic vertical cluster, with twelve managed instances, was configured in a dedicated webserver container in order to load balance users' requests and to provide the infrastructure to support high number of users with ease of deployment and high availability.

    • The database redo logs were configured on the raw disk partitions.

    • The mixed batch workload of 44 short UBEs and 8 long UBEs was executed concurrently with the 12,000 online application users, producing a sustained rate of 198.5 UBE/min.

    See Also

    Disclosure Statement

    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

    SPARC T5-2 (SPARC T5-2 Server base package, 2xSPARC T5 16-core processors, 32x16GB-1066 DIMMS, 4x600GB 10K RPM 2.5. SAS-2 HDD,2x300GB SSDs, 4x Sun Flash Accelerator F40 PCIe Cards, 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 03/18/2013.

    About

    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
    Search

    Archives
    « July 2016
    SunMonTueWedThuFriSat
         
    1
    2
    3
    4
    5
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    27
    28
    29
    30
    31
          
    Today