- ページ 15
ストレージ Dell FS7500のPDF マニュアルをオンラインで閲覧またはダウンロードできます。Dell FS7500 18 ページ。 Equallogic best practices series network attached storage
Dell FS7500 にも： 概要 (7 ページ), テクニカル・マニュアル (22 ページ), ハードウェア・セットアップ (2 ページ)
ページを印刷する
Dell EqualLogic FS7500 – Unified block and file storage for virtual workloads
Virtual Workload Configurations Tested on FS7500 over NFS
Table 1.
Virtual Workload Configurations Tested on FS7500 NFS
Workload
Application
Name
Exchange
Mail server
2007
Web 2.0
Olio
E-
DVD
Commerce
Store 2
Test Results
The simulations were designed to be highly reproducible, meaning that we could reproduce test results
if identical circumstances were provided. In each configuration we held everything else constant
except the number of EqualLogic PS6000XV arrays. We used the various application throughputs from a
single PS6000XV array configuration as the baseline. The test results from the other two configurations
were then compared to the baseline to determine the performance of each configuration relative to
the baseline. This comparison allowed us to show the performance improvements in each configuration
as we added additional PS6000XV arrays to the storage pool.
Figure 5 shows the simulation test results for three hardware configurations where the CPU, memory,
and network subsystems were kept constant and only the storage capacity in the EqualLogic FS7500
was increased by adding additional PS6000XV arrays. The height of each bar in this chart shows the
normalized application throughput in a VMware environment via NFS. These test results clearly show
that in a VMware environment, with multiple virtual applications running on an EqualLogic unified
scale-out NAS, a near linear increase in application transactional performance is achieved when
additional storage capacity is added. We were easily able to provision additional capacity in the
EqualLogic FS7500. Each time the PS series array was automatically discovered and added to the
storage pool, and data was load-balanced across all the disks in the pool while the applications were
running. As we increased the workloads, the application performance increased at a near-linear rate
with capacity.
5
SUSE Linux Enterprise Server
Workload
Virtual Machine
Config.
Platform
1000 Heavy
Windows 2008
Profile Users
64-bit
Back end: MySQL
database on
Two-Tier
Novell SLES
Java based;
64-bit
400
Front end:
Concurrent
Tomcat based tc
Users
Server on Novell
SLES 11 64-bit
Two-Tier
Back end: MySQL
Three
database on
Webservers
Novell SLES 11 64-
and One
bit
Front end (3): SLES
Database
11 64-bit
Server
Virtual Machine
Configuration
4
8GB
vCPU
RAM
5
11
2
2GB
vCPU
RAM
4
8GB
vCPU
RAM
4
4GB
vCPU
RAM
2
4GB
vCPU
RAM
Performance
QoS
Metric
No of
<500ms
Transactions
Transaction
80GB
Executed by
response
disk
Mail Server
time
Total
16GB
<600ms
completed
disk
operations
Operation
response
per minute
time
68GB
disk
No of
<700ms
Transactions
Transaction
49GB
Executed per
disk
minute
14GB
response
disk
time
Page 13