BestPerf: Performance & Best Practices

The Sun Fire X4270 server delivered an excellent result on the virtualization benchmark VMmark from VMware for 8 core platforms running VMware ESX 4.0 OS.

The Sun Fire X4270 server powered by two 2.93GHz Quad-Core Intel Xeon X5570 processors achieved a score of 24.18 @ 17 tiles, supporting 102 fully fledged Virtual Machine instances (17 Tiles).

With this competitive result of 24.18 @ 17 tiles for the VMmark virtualization benchmark, the Sun Fire X4270 server is within 1% of the top score of 24.35 @ 17 tiles for 8 core platforms with equivalent 1066 MHz memory.

The Sun and VMware partnership offers one of the best virtualization platforms in the industry with the performance and scalability features available on the Sun Fire X4270 server.

Under the heavy load conditions of the VMmark benchmark, the Sun Fire X4270 server delivers near linear scalability.

Consolidating multiple applications onto the Sun Fire X4270 server helps IT organizations cut cost and complexity, increase agility, and reduce data center power and cooling.

Customers can consolidate several Solaris 10 OS, Linux, and Microsoft Windows applications onto a single server using VMware Virtual Infrastructure technology

The Sun Fire X4270 server achieved the competitive result of  24.18 @17 tiles with a simple I/O configuration consisting of One single port 10Gbe network card and One 4Gb dual port FC HBA and One SAS/SATA combo HBA for supporting 8 internal SATA SSDs.

Competitive Landscape Performance

VMmark 8 Core Results (sorted by score, bigger is better)

System	CPU (GHz*)	GB	ESX ver	Spindles	RAID	Tiles	Score	Pub Date
HP BL490c G6	2 x Xeon X5570 (2.93)	96	4.0 #164009	133	0	17	24.54	09/22/09
Lenovo R525 G2	2 x Xeon X5570 (2.93)	96	4.0 #164009	55	0	17	24.35	06/30/09
Dell PowerEdge R710	2 x Xeon X5570 (2.93)	96	4.0 #164009	98	0	17	24.27	09/08/09
HP BL490 G6	2 x Xeon X5570 (2.93)	96	4.0 #158725	132	0	17	24.24	05/19/09
Fujitsu RX200 S5	2 x Xeon X5570 (2.93)	96	4.0 #164009	291	0	17	24.20	08/11/09
Sun Fire X4270	2 x Xeon X5570 (2.93)	96	4.0 #164009	235	0	17	24.18	09/28/09
HP DL380 G6	2 x Xeon X5570 (2.93)	96	4.0 #148783	120	0	17	24.15	05/19/09
Cisco B200-M1	2 x Xeon X5570 (2.93)	96	4.0 #151628	20	0	17	24.14	04/21/09
IBM BladeCenter HS22	2 x Xeon X5570 (2.93)	96	4.0 #161959	289	0	17	24.05	06/30/09
Dell PowerEdge R710	2 x Xeon X5570 (2.93)	96	4.0 #150817	170	0	17	24.00	04/21/09

Notes:
* Intel Turbo Boost up to 3.33GHz

Configuration Summary

Hardware Configuration:

1. Sun Fire X4270 Server
2 x 2.93GHz 4-Core Intel Xeon X5570 EP processors
96GB memory (12x 8GB DIMMs)
1x 32GB SATA SSD for OS
7x 32GB SATA SSD for database VMs.
1x QLE2462 4Gb dual port Fiber Channel Host Adapter
1x Intel Pro/10GbE-SR


2. Storage
9x STK2540 + 9x STK2501 RAID level 0, each with
12x 146GB SAS 15k rpm drives
1x STK2540 RAID level 0 with
12x 146GB SAS 15k rpm drives


3. Clients:
Sun Blade 6000 Chassis with 10x Sun Blade X6240
Sun Blade 6000 Chassis with 7x Sun Blade X6240
Each X6240 equipped with 2 x 2.5 GHz Quad Core AMD Opteron 2380,
32GB memory, 1x 73GB 15K rpm SAS Disk

Software Configuration:

1. VMware OS and Benchmark Software
VMware ESX 4.0 build #164009
VMmark 1.1.1


2. VMmark Virtual machines
1. Mail server
Windows 2003 32-bit Enterprise Edition
2 Virtual CPUs (vcpu)
24 GB disk
1 GB of memory
Exchange 2003

2. Java server
Windows 2003 64-bit Enterprise Edition
2 VCPUs
8 GB disk
1 GB of memory
SPECjbb2005

3. Standby server
Windows 2003 32-bit Enterprise Edition
1 VCPUs
4 GB disk
256 MB of memory
No application

4. Web server
SLES 10 64-bit
2 Virtual CPUs (vcpu)
8 GB disk
512 MB of memory
SPECweb2005

5. Database server
SLES 10 64-bit
2 Virtual CPUs (vcpu)
10 GB disk
2 GB of memory
MySQL

6. File server
SLES 10 32-bit
1 Virtual CPUs (vcpu)
8 GB disk
256 MB of memory
dbbench



3. Clients
Windows 2003 32-bit Enterprise Edition
LoadSim2003, Microsoft Outlook 2003
SPECjbb Monior
Idle VM test
SPECweb2005 client
MySQL, Sysbench
dbbench based tbench_srv.exe
BEA JRockit 5.0 JVM JDK
VMmark Harness
STAF framework and STAX execution engine.

Benchmark Description

VMmark is a benchmark developed, distributed and owned by VMware. The purpose of this benchmark is to measure performance and scalability of a pre-established mix of workloads (a Tile), which allows comparisons among similar platforms.

A Tile consists of 6 fixed workload applications, each running in its own Virtual Machine (VM) (6 VMs per Tile) such as Mail, Java, Web, Database and File Serving plus a standby server (spare Virtual Machine).

VMmark benchmark provides two key performance metrics:

1. The Number of TILES supported by a system, which is an indication of how many systems/applications can be consolidated on one platform (the higher the number of tiles supported the higher the number of consolidated systems.

2. The Score, which is an overall measure of the amount of work that is accomplished by all the Tiles in the system at a specified level of service of all the workloads during a benchmark run. The Score or Amount of Work is a composition of Actions/minute(Mail server), New Orders/minute(Java server), Access/minute(web server), Commits/minute(Database), MB/sec(file server).

Thus, among systems with the same number of tiles, the system with the higher score is the system that is capable of producing the greater amount of work. For detailed description of VMmark, tiles and score definition, please refer to http://www.vmware.com/products/vmmark/features.html.

See Also

• Sun Fire X4270 Benchmark Disclosure      VMmark Copy

• VMmark Benchmark Website

Disclosure Statement

VMware(R) VMmark(tm) is a product of VMware, Inc. VMmark utilizes SPECjbb(r)2005 and SPECweb(r)2005, which are available from the Standard Performance Evaluation Corporation (SPEC). Results from http://www.vmware.com/products/vmmark/ as of September 29, 2009.