SPECsfs2008_cifs Result

Fujitsu : PRIMERGY TX300 S5 (with 3x6 disks RAID 0)
SPECsfs2008_cifs = 16488 Ops/Sec (Overall Response Time = 4.68 msec)


Performance

Throughput
(ops/sec)
Response
(msec)
1679 1.0
3357 1.5
5033 2.3
6712 3.3
8403 4.2
10064 5.0
11805 6.8
13512 9.3
15222 9.2
16488 10.5
Performance Graph


Product and Test Information

Tested By Fujitsu
Product Name PRIMERGY TX300 S5 (with 3x6 disks RAID 0)
Hardware Available April 2009
Software Available March 2009
Date Tested April 2009
SFS License Number 19
Licensee Locations Paderborn, Germany

The PRIMERGY TX300 S5 is a 2-way tower server. In the measurement it was used for providing network file service utilizing the SAS RAID storage housed in the server's chassis.

Configuration Bill of Materials

Item No Qty Type Vendor Model/Name Description
1 1 Server Fujitsu PRIMERGY TX300 S5 Tower Server with two Intel Xeon X5570 processors, two onboard Gigabit network interfaces, two 8port SAS RAID controllers (each with 512 MB cache and BBU) and 20 directly connected SAS disks.

Server Software

OS Name and Version Microsoft Windows Server 2003 R2 Enterprise x64 Edition SP2 (Build 3790)
Other Software N/A
Filesystem Software N/A

Server Tuning

Name Value Description
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\FileSystem\NtfsDisable8dot3NameCreation dword:00000001 NTFS avoids the short-name attribute creation (8.3 DOS naming convention) for files
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\FileSystem\NtfsMemoryUsage dword:00000002 increase memory usage of NTFS
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\Interfaces\(INTERFACE NUMBER)\TcpAckFrequency dword:0000000d If only acking data and not any control packets, ack once every 13 packets, instead of the default of two.

Server Tuning Notes

None

Disks and Filesystems

Description Number of Disks Usable Size
2 arrays of 6 disks each (SAS, 146 GB, 10000 rpm) connnected with the first 8port SAS RAID controller; each array configured as a RAID 0 bound as a LUN on which a 817 GB NTFS file system was built. 12 1.6 TB
array of 6 disks (SAS, 146 GB, 10000 rpm) connnected with the second 8port SAS RAID controller, configured as a RAID 0 bound as a LUN on which a 817 GB NTFS file system was built. 6 817.0 GB
2 disks (SAS, 36 GB, 15000 rpm) connected with the second 8port SAS RAID controller, configured as RAID 1 used for the OS. 2 33.5 GB
Total 20 2.4 TB
Number of Filesystems 3
Total Exported Capacity 2.4 TB
Filesystem Type NTFS
Filesystem Creation Options default
Filesystem Config Three filesystems are created on the LUNs on the 8port SAS RAID controllers. The filesystems are 1, 2 and 3.
Fileset Size 1968.3 GB

Network Configuration

Item No Network Type Number of Ports Used Notes
1 Gigabit Ethernet 2 used Gigabit Ethernet ports on the PRIMERGY TX300 S5

Network Configuration Notes

All network interfaces were connected to a Netgear JGS516 switch.

Benchmark Network

Each of the two PRIMERGY TX300 S5 server onboard network interfaces were connected to a switch. Each of the LG single on-board network interface was connected to the switch.

Processing Elements

Item No Qty Type Description Processing Function
1 2 CPU Intel Xeon X5570 (2.93 GHz, 1 chip, 4 cores per chip, 1/4 MB (I+D) L2 cache per core, 8 MB (I+D) L3 cache per chip) TCP/IP, CIFS protocol, NTFS

Processing Element Notes

None

Memory

Description Size in GB Number of Instances Total GB Nonvolatile
NAS server main memory 96 1 96 V
Storage Processor main memory 0.5 2 1 NV
Grand Total Memory Gigabytes     97  

Memory Notes

The RAID controllers' write caches are backed up with battery backup units able to retain the write caches for 72 hours in the event of a power failure.

Stable Storage

Stable write and commit operations are not acknowledged until after the RAID storage system has acknowledged that the related data stored at least in non-volatile RAM. When the RAID controllers' battery backup units are off-line, the system turns off the write caches and writes directly to disk before acknowledging any write operations.

System Under Test Configuration Notes

See above for the tuning params.

Other System Notes

none

Test Environment Bill of Materials

Item No Qty Vendor Model/Name Description
1 8 Fujitsu PRIMERGY RX100 S3 Server with one Intel Pentium D 945 processor

Load Generators

LG Type Name LG1-8
BOM Item # 1
Processor Name Intel Pentium D 945
Processor Speed 3.40 GHz
Number of Processors (chips) 1
Number of Cores/Chip 2
Memory Size 2 GB
Operating System WIndows Server 2003
Network Type 2 onboard Gigabit Ethernet controllers Broadcom BCM5721 (1 used)

Load Generator (LG) Configuration

Benchmark Parameters

Network Attached Storage Type CIFS
Number of Load Generators 8
Number of Processes per LG 21

Testbed Configuration

LG No LG Type Network Target Filesystems Notes
1..4 LG1 1 1, 2, 3 N/A
5..8 LG1 2 1, 2, 3 N/A

Load Generator Configuration Notes

The filesystems were mounted on all clients. Four of the clients were connected to the first network, the other four to the second network.

Uniform Access Rule Compliance

The measurement was done to find out the performance of a PRIMERGY TX300 S5 server equipped with the maximum number of internal SAS disks. The server has two boxes for 2.5" SAS or SATA hard disks available in its chassis: one box for 12 disks which can be connected via an expander to one channel of the 1st SAS controller, and one box for 8 disks, four of which each can be connected to the two channels of the 2nd SAS controller. Two disks of the 2nd box were used for the OS. To make use of all other disks three equally dimensioned file systems were built: 2 on the 12 disks of the first box and 1 on the 6 disks of the second box. 12 processes were utilized on each of the 8 clients. The processes were spread over the filesystems equally. Four of the clients were connected to the first network, the other four to the second network.

Other Notes

(No other report notes)

Config Diagrams


Generated on Fri Jun 05 16:21:35 2009 by SPECsfs2008 HTML Formatter
Copyright © 1997-2008 Standard Performance Evaluation Corporation

First published at SPEC.org on 05-Jun-2009