Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison

Dell 6Gbps vs. HP 6Gbps RAID controller
performance comparison
Test report commissioned by Dell Inc. February 2010
Executive summary
We compared the performance of the 6Gbps Dell PowerEdge RAID controller (PERC) H700 against that of the
6Gbps HP Smart Array P410i controller with both systems using internal hard disk drives (HDDs). We also
compared the performance of the 6Gbps Dell PERC H800 against that of the 6Gbps HP Smart Array P411
controller using the Dell PowerVault MD1220 and HP StorageWorks D2700 external drive arrays, respectively.
140%
Percentage win
120%
100%
80%
60%
40%
KEY FINDINGS
TEST REPORT
Internal HDD tests
Dell PERC H700 average percentage win in IOPS over
FEBRUARY 2006
The Dell PERC
HP Smart Array P410i across RAID 5 and 10
H700 delivered up
to an average of
130.5% better
IOPS performance
with internal
HDDs than the HP
Smart Array
P410i.
20%
0%
Tests
Figure 1: Percentage wins of the Dell PERC H700 over the HP Smart Array P410i using internal
HDDs.
External drive arrays - Dell PERC H800 single path
average percentage win in IOPS over HP Smart Array
P411 single path across RAID 10 and 50
140%
Percentage win
120%
100%
80%
60%
40%
20%
0%
Tests
Figure 2: Percentage wins of the Dell PERC H800 with the Dell PowerVault MD1220 array in a
single-path configuration over the HP Smart Array P411 with the HP StorageWorks D2700 array in
a single-path configuration.
The Dell PERC
H800 in a singlepath configuration
delivered up to an
average of 117.2%
better IOPS
performance with
external drive
arrays in our tests
than the HP Smart
Array P411 in a
single-path
configuration.
External drive arrays - Dell PERC H800 redundant path
average percentage win in IOPS over HP Smart Array
P411 single path across RAID 10 and 50
140%
Percentage win
120%
100%
80%
60%
40%
20%
0%
KEY FINDINGS
The Dell PERC H800
in a redundant-path
configuration
delivered up to an
average of 115.6%
better IOPS
performance with
external drive
arrays in our tests
than the HP Smart
Array P411 in a
single-path
configuration.
Tests
Figure 3: Percentage wins of the Dell PERC H800 with the Dell PowerVault MD1220 array in a
redundant-path configuration over the HP Smart Array P411 with the HP StorageWorks D2700
array in a single-path configuration.
We present the details of the test access specifications we used in the What we tested section of this report.
Additionally, we present the specific results for individual tests in the What we found section and in Appendix C.
What we tested
Iometer
Iometer measures input/output (I/O) on single and clustered systems. Iometer performs I/O operations on a
system in order to stress the system, and then records the performance of and system stress created by these I/O
operations. Iometer can create and measure workloads on a single system or on networked systems. We used
Iometer version 2006.07.27 on both the Dell PowerEdge R710 and the HP ProLiant DL380 G6 to simulate a file
server workload on the RAID controllers and corresponding storage. We used the same Iometer workload across
all hardware configurations but tuned the number of outstanding I/Os to obtain the maximum possible input/output
operations per second (IOPS) on the controller with the best performance on all three sets of hardware
comparisons being made: the Dell 6Gbps vs. HP 6Gbps internal RAID controllers with hard disk drives, the
single-path configuration Dell 6Gbps vs. single-path configuration HP 6Gbps external RAID controllers with
storage arrays, and the redundant-path configuration Dell 6Gbps vs. single-path configuration HP 6Gbps external
RAID controllers with storage arrays.
Figure 4 displays the details on the Iometer access specifications we used. We present the number of outstanding
I/Os and other specific settings we used during testing in Figure 8 of the Test configurations section.
Note: We tested the Dell hardware in both single-path and redundant-path configurations to demonstrate that the
redundant-path configuration provides a significant performance increase over the HP solution while also adding
fault tolerance to the configuration; even if a single path fails, the server continues to have access to the storage
through the alternate port.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
2
Access specification name and
block size
DB OLTP 8K
Exchange email 4K
Exchange email 8K
Media streaming 64K
OS Drive 8K
SQL server log 64K
Web file server 4K
Web file server 8K
Web server log 8K
Percentage
read
Percentage
write
Percentage
random
Percentage
sequential
70%
67%
67%
98%
70%
0%
95%
95%
0%
30%
33%
33%
2%
30%
100%
5%
5%
100%
100%
100%
100%
0%
100%
0%
75%
75%
0%
0%
0%
0%
100%
0%
100%
25%
25%
100%
Figure 4: Description of access specification settings.
For more information about Iometer, see http://www.iometer.org/.
Testing overview
We used a Dell PowerEdge R710 server for the Dell controllers and an HP ProLiant DL380 G6 server for the HP
controllers, each paired with its corresponding storage, to test the Iometer workload on all RAID levels.
First, at RAID levels 5 and 10, we compared the performance of an internal Dell PERC H700 6Gbps RAID
controller against an internal HP Smart Array P410i 6Gbps RAID controller. The corresponding storage for testing
the internal RAID controllers were six 6Gbps 73GB 15K SAS HDDs.
Then, at RAID levels 10 and 50, we compared the performance of an external Dell PERC H800 6Gbps RAID
controller against that of an external HP Smart Array P411 6Gbps RAID controller, both connected in a singlepath configuration. We used one Dell PowerVault MD1220 drive array with a total of 24 73GB drives to test the
Dell PERC H800, and one HP StorageWorks D2700 drive array with a total of 24 6Gbps 73GB 15K SAS HDDs to
test the HP Smart Array P411.
We then repeated the RAID levels 10 and 50 external HDD tests, but this time with the Dell PERC H800 6Gbps
RAID controller in a redundant-path configuration against that of the HP Smart Array P411 6Gbps RAID controller
in a single-path configuration with external storage. We present the details on the 13 hardware configurations we
used for testing in the Test configuration section below.
What we found
We report the IOPS results of our custom Iometer tests. For all tests, higher IOPS numbers are better. We ran
each test three times and report results from the run that produced the median of the three IOPS results.
We calculated the average percentage wins for each access specification by averaging the performance gain
percentage of all of the RAID levels we tested.
Internal HDD tests
Figure 5 displays the Dell PERC H700 percentage improvement in IOPS performance over the HP Smart Array
P410i. The Dell PERC H700 delivered an average performance increase ranging from 4.3 percent to 130.5
percent.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
3
Internal HDD test access
specifications and block
size
DB OLTP 8K
Exchange email 4K
Exchange email 8K
Media streaming 64K
OS Drive 8K
SQL server log 64K
Web file server 4K
Web file server 8K
Web server log 8K
Dell PERC H700
percentage win in
IOPS over HP Smart
Array P410i
RAID 5
Dell PERC H700
percentage win in
IOPS over HP Smart
Array P410i
RAID 10
10.8%
23.6%
7.6%
51.1%
11.3%
5.5%
2.3%
3.8%
26.7%
14.6%
15.7%
21.2%
22.2%
18.3%
123.8%
6.8%
4.7%
234.3%
Dell PERC H700
average percentage
win in IOPS over HP
Smart Array P410i
across both RAID 5
and RAID 10
12.7%
19.6%
14.4%
36.6%
14.8%
64.6%
4.5%
4.3%
130.5%
Figure 5: The Dell PERC H700 percentage improvement in IOPS performance over the HP Smart Array P410i. Higher numbers are
better.
External drive arrays – Single-path configuration
For our external drive array single-path configuration tests, the Dell PERC H800 RAID controller in conjunction
with the Dell PowerVault MD1220 array delivered significantly better performance than the HP Smart Array P411
controller in conjunction with the HP StorageWorks D2700 array on all access specifications we tested across
both RAID levels. The Dell PERC H800 controller delivered the greatest percentage increase in IOPS in Web
server log 8K, with a 117.2 percent average performance increase. The remaining access specifications also
yielded significant average performance increases ranging from 24.4 percent to 89.7 percent.
Figure 6 displays the results from our external drive arrays single-path configuration.
External drive array test
access specifications and
block size
DB OLTP 8K
Exchange email 4K
Exchange email 8K
Media streaming 64K
OS Drive 8K
SQL server log 64K
Web file server 4K
Web file server 8K
Web server log 8K
Dell PERC H800
percentage win in IOPS
over HP Smart Array
P411 RAID 10
0.8%
2.2%
2.3%
31.4%
1.6%
52.1%
20.3%
17.0%
116.2%
Dell PERC H800
percentage win in IOPS
over HP Smart Array
P411 RAID 50
137.3%
177.3%
153.4%
38.8%
134.3%
6.3%
32.7%
31.8%
118.1%
Dell PERC H800
average percentage
win in IOPS over HP
Smart Array P411
across both RAID 10
and RAID 50
69.1%
89.7%
77.9%
35.1%
68.0%
29.2%
26.5%
24.4%
117.2%
Figure 6: Access specifications results in our external drive arrays single path-configuration tests showing Dell PERC H800
improvements over the HP Smart Array P411. Higher numbers are better.
External drive arrays – Redundant-path configuration vs. single-path configuration
For our external drive array redundant-path configuration tests, the Dell PERC H800 in conjunction with the Dell
PowerVault MD1220 configured in redundant path mode outperformed the HP Smart Array P411 controller in
conjunction with the HP StorageWorks D2700 array configured in single-path mode on all access specifications
across both RAID levels. The greatest average performance increase for the Dell PERC H800 was for the Web
server log 8K access specification with 115.6 percent more IOPS than the HP Smart Array P411. The remaining
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
4
access specifications also yielded significant average performance increases ranging from 24.5 percent to 89.9
percent.
Figure 7 displays the results from our external drive array redundant-path configuration tests.
External drive
array test access
specifications and block
size
Dell PERC H800
percentage win in IOPS
over HP Smart Array
P411 RAID 10
DB OLTP 8K
Exchange email 4K
Exchange email 8K
Media streaming 64K
OS Drive 8K
SQL server log 64K
Web file server 4K
Web file server 8K
Web server log 8K
Dell PERC H800
average percentage
win in IOPS over HP
Smart Array P411
across both RAID 10
and RAID 50
68.9%
89.9%
76.6%
37.7%
67.8%
31.9%
26.6%
24.5%
115.6%
Dell PERC H800
percentage win in IOPS
over HP Smart Array
P411 RAID 50
0.6%
1.8%
1.9%
36.9%
1.7%
58.4%
21.2%
17.7%
113.9%
137.2%
178.0%
151.4%
38.4%
134.0%
5.3%
32.0%
31.2%
117.4%
Figure 7: Access specification results in our external drive arrays redundant-path configuration tests showing Dell PERC H800
improvements over the HP Smart Array P411. Higher numbers are better.
Test configurations
Host servers
®
®
 Dell PowerEdge R710, Intel Xeon Processor X5570, 24GB memory (6 x 4 GB)
 HP ProLiant DL380 G6, Intel Xeon Processor X5570, 24GB memory (6 x 4 GB)
Iometer settings
 1 worker per target (1 total)
 4GB Dataset per target (4 GB total)
 Outstanding I/Os tuned for each access specification
Figure 8 displays the outstanding I/O settings for each access specification we tested.
Access specification
DB OLTP 8K
Exchange 4K
Exchange 8K
Media streaming 64K
OS Drive 8K
SQL server log 64K
Web file server 4K
Web file server 8K
Web server log 8K
Internal HDD testing
outstanding I/Os
64
64
64
16
64
16
128
128
64
External HDD testing
outstanding I/Os
64
64
64
128
64
32
128
128
256
Figure 8: Outstanding I/O settings for each access specification we tested.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
5
Dell PERC H700 vs. HP Smart Array P410i
Internal HDD testing
Figure 9 shows the specific hardware configurations for each RAID controller and RAID level during internal HDD
testing.
Controller
HP Smart Array P410i
HP Smart Array P410i
Dell PERC H700
Dell PERC H700
OS partition HDDs
(RAID 1)
2 x 73GB 6Gbps 15K RPM SAS
2 x 73GB 6Gbps 15K RPM SAS
2 x 73GB 6Gbps 15K RPM SAS
2 x 73GB 6Gbps 15K RPM SAS
Target HDDs
6 x 73GB 6Gbps 15K RPM SAS
6 x 73GB 6Gbps 15K RPM SAS
6 x 73GB 6Gbps 15K RPM SAS
6 x 73GB 6Gbps 15K RPM SAS
RAID level
target HDDs
5
10
5
10
Figure 9: Hardware configurations for each RAID controller and RAID level during internal HDD testing.
Dell PERC H800 vs. HP Smart Array P411
External HDD testing
For external testing with the Dell PERC H800, we configured the Dell PowerEdge R710 with the Dell PERC H700
internal RAID controller with two 73GB 6Gbps 15K RPM SAS drives in a RAID 1 volume for the operating system.
In the case of the HP Smart Array P411 external testing, we configured the HP ProLiant DL380 G6 with the HP
Smart Array P410i internal RAID controller with two of the same 73GB 6Gbps 15K RPM SAS drives, again in a
RAID 1 volume for the operating system.
We tested the Dell hardware in both single-path and redundant-path configurations to connect the storage arrays
to the controllers. This was to demonstrate that the redundant-path configuration provides a significant
performance increase over the HP solution while also adding fault tolerance to the configuration; even if a single
path fails, the server continues to have access to the storage through the alternate port. Additionally, the Dell
PERC H800 controller has automatic I/O load balancing capabilities to detect when a single path starts to become
saturated. The controller then automatically balances the I/O traffic across both paths.
Figure 10 shows the specific hardware configurations for each RAID controller and RAID level during external
HDD testing.
External RAID
controller
HP Smart Array P411
HP Smart Array P411
Dell PERC H800
Dell PERC H800
Dell PERC H800
Dell PERC H800
Enclosure/Connectivity
1 x HP Storage Works D2700
single path
1 x HP Storage Works D2700
single path
1 x Dell PowerVault MD1220
single path
1 x Dell PowerVault MD1220
single path
1 x Dell PowerVault MD1220
redundant path
1 x Dell PowerVault MD1220
redundant path
Target
HDDs per
enclosure
24
24
24
24
24
24
Target HDDs
24 x 73GB
6Gbps 15K RPM SAS
24 x 73GB
6Gbps 15K RPM SAS
24 x 73GB
6Gbps 15K RPM SAS
24 x 73GB
6Gbps 15K RPM SAS
24 x 73GB
6Gbps 15K RPM SAS
24 x 73GB
6Gbps 15K RPM SAS
RAID level
target
HDDs
10
50
10
50
10
50
Figure 10: Hardware configurations for each RAID controller and RAID level during external HDD testing.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
6
How we tested
Installing the operating system
Installing Windows Server 2008 Enterprise R2 x64
1. Boot the server, and insert the Windows Server 2008 R2 x64 installation DVD in the DVD-ROM drive.
2. At the Language Selection screen, click Next.
3. Click Install Now.
4. Select Windows Server 2008 Enterprise (Full Installation) x64, and click Next.
5. Click the I accept the license terms check box, and click Next.
6. Click Custom.
7. Click Drive options (advanced).
8. Delete any existing partitions.
9. Ensure the first drive is selected, and click New.
10. Click Apply.
11. Click OK.
12. Click Next.
13. At the User’s password must be changed before logging on warning screen, click OK.
14. Type your new password into both fields, and click the arrow to continue.
15. At the Your password has been changed screen, click OK.
Windows Server 2008 settings
We installed all recommended and critical Windows updates through 12/7/2009.
Disable Windows Firewall:
1. Click StartAdministrative ToolsWindows Firewall and Advanced Security.
2. Under the Overview heading, click Windows Firewall Properties.
3. Click the drop-down menu beside Firewall state, and select Off.
4. Click the Private Profile tab.
5. Click the drop-down menu beside Firewall state, and select Off.
6. Click the Public Profile tab.
7. Click the drop-down menu beside Firewall state, and select Off.
8. Click OK to close the Properties window.
9. Close the Windows Firewall with Advanced Security window.
Installing Dell PowerEdge R710 software and preparing the target drives
Installing Dell OpenManage Server Administrator 6.2
1. Insert the Dell OpenManage Install DVD.
2. Select Dell OpenManage Server Administrator, and click Install.
3. At the installer dialogue box, click Install again.
4. At the Welcome screen, click Next.
5. Select Install, and click Next.
6. At the Custom Setup window, leave default components selected, and click Next.
7. Click Install.
Setting up a test volume
1. Reboot the system.
2. When the application prompts you to do so, press CTRL+R to enter the PERC BIOS Configuration Utility.
3. Highlight the appropriate RAID controller, and press Enter.
4. Highlight the RAID controller card, press F2, and select Create New VD.
5. Select the appropriate RAID level, and press Enter.
6. Select all 6 drives if testing an internal RAID controller, or 24 drives in a single storage array if testing an
external RAID controller.
7. In Basic Settings, enter a VD Name.
8. Select OK, and select OK again at the warning dialogue box.
9. At the main screen, highlight the newly created Virtual Disk, and press F2.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
7
10. Select Initialization, and in the sub-menu, select Start Init.
11. Select OK when the dialogue box notifies you that the initialization is complete.
12. If testing an external RAID controller, repeat steps 4 through 11 twice to add and initialize the remaining
volumes.
13. After initialization is complete, press ESC twice, and select OK to exit.
14. Press Ctrl+Alt+Delete to reboot when the application prompts you to do so.
Formatting and mounting each test volume
1. Click StartAdministrative ToolsComputer Management.
2. Under Storage, click Disk Management.
3. When the application prompts you to initialize the drive(s), select the GPT partition style, and click OK.
4. For each partition:
a. Right-click Unallocated space, and click New Simple Volume.
b. Leave the default maximum volume size, and click Next.
c. Leave the default drive letter, and click Next.
d. Check the Perform a quick format box.
e. Click Next.
f. Click Finish.
Installing HP ProLiant DL380 G6 software and preparing the target drives
Installing the HP Array Configuration Utility
1. Double click the installer package.
2. Click Install.
Setting up a test volume
1. Open the HP Array Configuration Utility.
2. Under the Configuration tab, select the appropriate RAID controller from the drop-down menu.
3. Under Systems and Devices, highlight the RAID controller, and click Create Array in the right pane.
4. Under Select Physical Drives for the New Array, select either 6 physical drives if doing the internal testing,
or 24 physical drives if doing the external testing, and click OK.
5. Highlight the newly created Logical Drive, and click Create Logical Drive in the right pane.
6. Under Fault Tolerance, select the appropriate RAID level.
7. Leave all other settings at their default value, and click Save.
8. If creating a RAID 50 Logical Drive, allow the Array Configuration Utility to complete initialization once the
volume has been formatted in Windows.
Formatting and mounting each test volume
1. Click StartAdministrative ToolsComputer Management.
2. Under Storage, click Disk Management.
3. When the application prompts you to initialize the drive(s), select the GPT partition style, and click OK.
4. For each partition:
a. Right-click Unallocated space, and click New Simple Volume.
b. Leave the default maximum volume size, and click Next.
c. Leave the default drive letter, and click Next.
d. Check the Perform a quick format box.
e. Click Next.
f. Click Finish.
Setting up Iometer
Installing and configuring Iometer
1. Download the Iometer package from www.iometer.org.
2. Double-click the installer, and click Next at the welcome window.
3. At the License Agreement window, click I Agree.
4. At the Choose Components window, leave the defaults selected, and click Next.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
8
5. At the Choose Install Location window, change the Destination Folder to C:\Iometer 2006.07.27, and click
Install.
Setting up the individual Iometer workloads
We used the following settings for each test:
1. Open Iometer.exe.
2. For each access specification:
a. Create the access specification to match the configurations in Figure 4.
b. Verify that the access specification has the following additional settings:
i. Under Burstiness, set Transfer Delay to 0ms, and set Burst Length to 1 I/O.
ii. Under Align I/Os, select Sector Boundaries.
iii. Under Reply Size, select No Reply.
c. Under Topology, select the computer name, and click the Start a New Disk Worker on Selected
Manager button until you have one worker assigned to each target volume.
d. Under Disk Targets, set the # of Outstanding I/Os according to the corresponding value shown in
Figure 8.
e. Set the disk size to 8388608 sectors (4GB).
f. Under Results Display, make sure that Iometer has selected Start of Test.
g. Under Test Setup, set the Run Time to 4 minutes and the Ramp Up Time to 60 seconds.
3. Exit Iometer.
Running the test
1. Reboot the system.
2. After logging in, open a command prompt.
3. Type cd c:\Iometer 2006.07.27 and press Enter.
4. Type run.bat and wait 5 minutes.
5. Press Enter.
6. After all access specifications are finished running, copy the result files from the server.
7. Repeat steps 1 through 6 two more times, for a total of three runs.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
9
Appendix A ̶ Test server information
Figure 11 provides detailed information for the test servers.
Test server
HP ProLiant DL380 G6
Dell PowerEdge R710
General processor setup
Number of processor packages
Number of cores per processor
package
Number of hardware threads per core
2
2
4
4
2
2
System power management policy
Balanced
Balanced
Vendor
Intel
Intel
Name
Intel Xeon Processor X5570
Intel Xeon processor X5570
Stepping
D0
D0
Socket type
LGA 1366
LGA 1366
Core frequency (GHz)
2.93
2.93
Front-side bus frequency
6.4 GT/s
6.4 GT/s
L1 cache
32 KB + 32 KB (per core)
32 KB + 32 KB (per core)
L2 cache
1 MB (4 x 256 KB)
1 MB (4 x 256 KB)
L3 cache (MB)
8
8
Vendor and model number
Dell PowerEdge R710
HP ProLiant DL380 G6
Motherboard model number
0M233H
PADAB0G9VXC1CQ
Motherboard chipset
Intel 5520
Intel 5520
BIOS name and version
Dell 1.3.6 (10/30/2009)
HP P62 (10/01/2009)
BIOS settings
Default
Default
Vendor and model number
Crucial CT51272BB1339
Crucial CT51272BB1339
Type
PC3-10600R
PC3-10600R
Speed (MHz)
Speed in the system currently running
@ (MHz)
Timing/Latency (tCL-tRCD-iRPtRASmin)
RAM module size (GB)
1,333
1,333
1,333
1,066
9-9-9-24
9-9-9-24
4
4
Number of RAM modules
6
6
Chip organization
Double-sided
Double-sided
Total system memory (GB)
24
24
Build number
Windows Server 2008 R2
Enterprise x64 Edition
7600
Windows Server 2008 R2
Enterprise x64 Edition
7600
Service pack
NA
NA
File system
NTFS
NTFS
CPU
Platform
Memory module(s)
Operating system
Name
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
10
Kernel
Dell PowerEdge R710
ACPI x64-based PC
HP ProLiant DL380 G6
ACPI x64-based PC
Language
English
English
Microsoft DirectX version
11
11
Vendor and model number
Matrox G200
ATI ES1000
Chipset
G2+
ES1000
Type
Integrated
Integrated
Memory size (MB)
8
64
Resolution
1,024 x 768
1,024 x 768
Vendor and model number
Broadcom BCM5709C NetXtreme
II GigE (NDIS VBD Client) x 2
HP NC382i DP Multifunction
Gigabit
Type
Integrated
Driver version
Broadcom 5.0.15.0 (10/02/2009)
Integrated
Hewlett-Packard 5.0.13.0
(7/30/2009)
Test server
Graphics
Network card/subsystem
Optical drive
Vendor and model number
TSSTCorp DVD-ROM TS-L333A
LG GDR-D20N DVD-ROM
Number
4
4
Type
USB 2.0
USB 2.0
Total number
2
2
Wattage of each (W)
870
460
Total number
5
6
Dimensions (mm)
80
80
Voltage (V)
12
12
Amps (A)
1.60
2.45
USB ports
Power supplies
Cooling fans
Figure 11: Detailed test server information.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
11
Appendix B ̶ Test storage information
Figures 12 and 13 provide detailed information for the test storage.
Internal testing hardware
RAID controller
Firmware version
Driver version
Cache size (MB)
ROC (Raid-on-Chip)
Internal storage
Vendor and model number
Number of drives
Speed (Gbps)
Size (GB)
RPM
Type
Dell PERC H700
12.0.1-0083
4.17.2.64
512
LSI 2108
HP Smart Array P410i
2.50
6.18.0.64
512
PM8011
Dell ST973452SS
8
6
73
15K
SAS
HP DH0072FACRD
8
6
73
15K
SAS
Figure 12: Primary internal storage hardware.
External testing hardware
RAID controller
Firmware version
Driver version
Cache size (MB)
ROC (Raid-on-Chip)
Storage arrays
Vendor and model
Connection type
Total number of drives
Hard drives
Vendor and model number
Number of drives
Speed (Gbps)
Size (GB)
RPM
Type
Dell PERC H800
12.0.1-0083
4.17.2.64
512
LSI 2108
HP Smart Array P411
1.58
6.18.0.64
512
PM8011
Dell PowerVault MD1220
SAS SFF 8088
24
HP StorageWorks D2700
SAS SFF 8088
24
Dell ST973452SS
24
6
73
15K
SAS
HP DH0072FACRD
24
6
73
15K
SAS
Figure 13: Primary external storage hardware.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
12
Appendix C ̶ Detailed test results
Figure 14 shows results in IOPS from the HP Smart Array P410i and the Dell PERC H700 RAID controllers during
the internal HDD testing.
Access specification
name and block size
DB OLTP 8K
Exchange email 4K
Exchange email 8K
Media streaming 64K
OS Drive 8K
SQL server log 64K
Web file server 4K
Web file server 8K
Web server log 8K
HP Smart Array
P410i RAID 5
IOPS
1,561.39
1,561.58
1,445.91
3,635.63
1,542.84
10,740.85
4,133.71
3,631.64
53,838.51
Dell PERC H700
RAID 5 IOPS
1,730.63
1,930.64
1,555.09
5,493.46
1,717.03
11,331.78
4,229.71
3,771.30
68,228.92
HP Smart Array
P410i RAID 10
IOPS
2,659.65
2,668.94
2,484.86
4,711.34
2,553.91
3,057.44
4,607.96
4,222.92
16,322.35
Dell PERC H700
RAID 10 IOPS
3,047.16
3,086.86
3,012.07
5,756.03
3,021.09
6,842.36
4,919.90
4,421.75
54,572.70
Figure 14: IOPS results from the HP Smart Array P410i and Dell PERC H700 during internal HDD testing. Higher numbers are better.
Figure 15 shows results in IOPS for RAID levels 10 and 50 from the HP Smart Array P411 in conjunction with the
HP StorageWorks D2700 in single-path configuration and the Dell PERC H800 in conjunction with the Dell
PowerVault MD1220 in single-path configuration during the external HDD testing.
Access specification
name and block size
DB OLTP 8K
Exchange email 4K
Exchange email 8K
Media streaming 64K
OS Drive 8K
SQL server log 64K
Web file server 4K
Web file server 8K
Web server log 8K
HP Smart Array P411
HP StorageWorks
D2700
RAID 10 IOPS
9,939.40
10,260.39
9,671.60
15,041.57
9,771.01
8,361.55
13,828.38
13,129.37
42,856.22
Dell PERC H800
Dell PowerVault
MD1220
RAID 10 IOPS
10,019.46
10,484.08
9,898.73
19,759.23
9,928.24
12,719.04
16,631.20
15,365.89
92,667.27
HP Smart Array P411
HP StorageWorks
D2700
RAID 50 IOPS
3,579.15
3,180.96
3,241.36
14,737.77
3,600.35
10,303.59
12,071.03
11,142.05
38,311.67
Dell PERC H800
Dell PowerVault
MD1220
RAID 50 IOPS
8,493.44
8,819.78
8,212.52
20,458.96
8,436.32
10,948.88
16,019.46
14,686.48
83,568.91
Figure 15: IOPS results from the HP Smart Array P411 and Dell PERC H800 during external HDD single-path configuration testing,
RAID 10 and 50. Higher numbers are better.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
13
Figure 16 shows results in IOPS for RAID levels 10 and 50 from the HP Smart Array P411 in conjunction with the
HP StorageWorks D2700 in single-path configuration and Dell PERC H800 in conjunction with the Dell
PowerVault MD1220 in redundant-path configuration during the external HDD testing.
Access specification
name and block size
HP Smart Array P411
HP StorageWorks
D2700
RAID 10 IOPS
Dell PERC H800
Dell PowerVault
MD1220
RAID 10 IOPS
HP Smart Array P411
HP StorageWorks
D2700
RAID 50 IOPS
Dell PERC H800
Dell PowerVault
MD1220
RAID 50 IOPS
DB OLTP 8K
Exchange email 4K
Exchange email 8K
Media streaming 64K
OS Drive 8K
SQL server log 64K
Web file server 4K
Web file server 8K
Web server log 8K
9,939.40
10,260.39
9,671.60
15,041.57
9,771.01
8,361.55
13,828.38
13,129.37
42,856.22
9,999.57
10,441.34
9,856.91
20,597.45
9,934.76
13,245.92
16,758.50
15,458.34
91,649.68
3,579.15
3,180.96
3,241.36
14,737.77
3,600.35
10,303.59
12,071.03
11,142.05
38,311.67
8,491.14
8,841.78
8,148.10
20,395.60
8,424.26
10,853.01
15,929.53
14,614.18
83,288.55
Figure 16: IOPS results from the HP Smart Array P411 and Dell PERC H800 during external HDD redundant-path configuration testing,
RAID 10 and 50. Higher numbers are better.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
14
About Principled Technologies
We provide industry-leading technology assessment and fact-based marketing services. We bring to every
assignment extensive experience with and expertise in all aspects of technology testing and analysis, from
researching new technologies, to developing new methodologies, to testing with existing and new tools.
When the assessment is complete, we know how to present the results to a broad range of target audiences. We
provide our clients with the materials they need, from market-focused data to use in their own collateral to custom
sales aids, such as test reports, performance assessments, and white papers. Every document reflects the results
of our trusted independent analysis.
We provide customized services that focus on our clients’ individual requirements. Whether the technology
involves hardware, software, Web sites, or services, we offer the experience, expertise, and tools to help you
assess how it will fare against its competition, its performance, whether it’s ready to go to market, and its quality
and reliability.
Our founders, Mark L. Van Name and Bill Catchings, have worked together in technology assessment for over 20
years. As journalists, they published over a thousand articles on a wide array of technology subjects. They
created and led the Ziff-Davis Benchmark Operation, which developed such industry-standard benchmarks as Ziff
Davis Media’s Winstone and WebBench. They founded and led eTesting Labs, and after the acquisition of that
company by Lionbridge Technologies were the head and CTO of VeriTest.
Principled Technologies, Inc.
1007 Slater Road, Suite 250
Durham, NC 27703
www.principledtechnologies.com
info@principledtechnologies.com
Principled Technologies is a registered trademark of Principled Technologies, Inc.
Intel, Xeon, and Pentium are registered trademarks of Intel Corporation or its subsidiaries in the United States
and other countries.*All other product names are the trademarks of their respective owners.
Disclaimer of Warranties; Limitation of Liability:
PRINCIPLED TECHNOLOGIES, INC. HAS MADE REASONABLE EFFORTS TO ENSURE THE ACCURACY AND VALIDITY OF ITS
TESTING, HOWEVER, PRINCIPLED TECHNOLOGIES, INC. SPECIFICALLY DISCLAIMS ANY WARRANTY, EXPRESSED OR
IMPLIED, RELATING TO THE TEST RESULTS AND ANALYSIS, THEIR ACCURACY, COMPLETENESS OR QUALITY, INCLUDING
ANY IMPLIED WARRANTY OF FITNESS FOR ANY PARTICULAR PURPOSE. ALL PERSONS OR ENTITIES RELYING ON THE
RESULTS OF ANY TESTING DO SO AT THEIR OWN RISK, AND AGREE THAT PRINCIPLED TECHNOLOGIES, INC., ITS
EMPLOYEES AND ITS SUBCONTRACTORS SHALL HAVE NO LIABILITY WHATSOEVER FROM ANY CLAIM OF LOSS OR DAMAGE
ON ACCOUNT OF ANY ALLEGED ERROR OR DEFECT IN ANY TESTING PROCEDURE OR RESULT.
IN NO EVENT SHALL PRINCIPLED TECHNOLOGIES, INC. BE LIABLE FOR INDIRECT, SPECIAL, INCIDENTAL, OR
CONSEQUENTIAL DAMAGES IN CONNECTION WITH ITS TESTING, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
IN NO EVENT SHALL PRINCIPLED TECHNOLOGIES, INC.’S LIABILITY, INCLUDING FOR DIRECT DAMAGES, EXCEED THE
AMOUNTS PAID IN CONNECTION WITH PRINCIPLED TECHNOLOGIES, INC.’S TESTING. CUSTOMER’S SOLE AND EXCLUSIVE
REMEDIES ARE AS SET FORTH HEREIN.
Principled Technologies, Inc.: Dell 6Gbps vs. HP 6Gbps RAID controller performance comparison
15