La lecture en ligne est gratuite
Le téléchargement nécessite un accès à la bibliothèque YouScribe
Tout savoir sur nos offres
Télécharger Lire

HP ProLiant BL20p Enclosure Lotus Domino NotesBench Audit –

51 pages
HP ProLiant BL20p Enclosure Lotus Domino NotesBench Audit Report Section 1 - Executive summary .......................................................................................................... 3 Test sponsor............................................................................................... 3 Test methodologies...................... 3 Conclusions................................ 4 Section 2 – Benchmark Objectives..................................... 5 Section 3- Test Methodologies........................................................................... 6 Test configuration........................................................ 6 Test procedure............................................................................................ 6 Section 4- R6Mail Data.................... 8 Lotus NotesMark Values for Benchmarked Configurations .................................................................. 8 Mail-Only Users Per Server ........................................... 9 CPU Performance as Recorded by Lotus Domino............. 10 Section 5- Analysis........................................................................................................................ 11 Price/Performance Ratios:.......... 11 NotesNum Output:.................... 11 Section 7- Statement by Auditor ...................................................................................................... 13 Appendix A - Overall Test Setup and ...
Voir plus Voir moins

HP ProLiant BL20p Enclosure Lotus Domino NotesBench
Audit Report


Section 1 - Executive summary .......................................................................................................... 3
Test sponsor............................................................................................... 3
Test methodologies...................... 3
Conclusions................................ 4
Section 2 – Benchmark Objectives..................................... 5
Section 3- Test Methodologies........................................................................... 6
Test configuration........................................................ 6
Test procedure............................................................................................ 6
Section 4- R6Mail Data.................... 8
Lotus NotesMark Values for Benchmarked Configurations .................................................................. 8
Mail-Only Users Per Server ........................................... 9
CPU Performance as Recorded by Lotus Domino............. 10
Section 5- Analysis........................................................................................................................ 11
Price/Performance Ratios:.......... 11
NotesNum Output:.................... 11
Section 7- Statement by Auditor ...................................................................................................... 13
Appendix A - Overall Test Setup and Software Versions...... 14
Test Setup ................................ 14
Appendix B - System Configuration.................................................................................................. 16
Enclosure (Enclosure Under Test).. 16
Clients..... 16
Network Equipment................................................................................................................... 16
Software Versions..................... 16
Appendix C - Operating System Parameters Values............ 17
Appendix D - Notes Parameters ...................................................................................................... 18
R5Mail Workload Test............... 18
System Under Test Notes.ini.... 18
The Sample Client Notes.ini.... 41
Appendix E - Network Configuration Files......................................................................................... 45
Figure E-1. HP ProLiant BL20p Enclosure Lotus Domino Benchmark........................................................ 45
Network Configuration .............................................................................. 46
Appendix F - Guidelines for Information Usage.................. 47
Appendix G - System Pricing.......................................... 48
Appendix H – Vendor Statement ..................................................................... 50




NotesBench Test Results
Section 1 - Executive summary
Test sponsor
This Lotus® Domino® NotesBench testing was sponsored by Hewlett Packard Corporation.
The NotesBench benchmark was developed and engineered by Lotus Development
Corporation. Testing took place at the HP Solutions Engineering Laboratories in Nashua,
New Hampshire in November, 2003 and was audited by KMDS Technical Associates, Inc.,
in November 2003.
Test methodologies
Blade Computer technology represents a radical change in the definition of what constitutes
a server. Each individual blade contains only part of the essential components that are
required to make it an operational system. In order to be fully operational, the blade must be
placed into an enclosure, which shares other key system components such as network and
SAN connectivity for all the blades it contains. This disaggregating of key system
components blurs the line defining the ‘system’ between the blade and the enclosure,
enables increased configuration and deployment flexibility, and is part of what makes blade
servers attractive. HP realizes that customers are expressing interest in Lotus Domino 6
performance on these configurations. To this end, this report addresses the performance of a
fully configured HP ProLiant BL20p Enclosure running Lotus Domino 6.0.2 using an HP
Enterprise Virtual Array Storage Area Network.
The HP ProLiant BL20p Enclosure was configured to run a single test, Notesbench R6mail.
The Notesbench R6mail workload is designed to replicate an average group of users
performing everyday mail and calendaring tasks. The test measured the performance of
120,000 users participating in Notes R6Mail related activities simultaneously connected to
sixteen Domino R6.02 Servers, spread across the eight HP ProLiant BL20p G2 servers within
the enclosure.
The test was set up using two groups of hardware: the driving systems, and the enclosure
under test. The enclosure under test contained 8 HP ProLiant BL20p G2 servers. Each HP
ProLiant BL20p server was configured with two Intel® 2.8Ghz hyper-threading Xeon™
CPUs, 4 gigabytes of RAM, and two 36.4 SCSI hard disks. Based on Intel’s latest IA-32
processors with Hyper-Threading technology, the HP ProLiant BL20p G2 server is a two CPU
system that configures as a system with four logical processors. As described by Intel, each
Hyper-Threading CPU has two IA-32 architectural states that share the execution resources of
the processor core. Hyper-Threading technology is designed to improve the performance of
the IA-32 processors by exploiting the multi-threading nature of the operating system (OS)
and applications in such a way as to increase the use of the on-chip execution resources
available in the Intel NetBurst microarchitecture. More information is available on Hyper-
threading at WWW.INTEL.COM. Each HP ProLiant BL20p server blade was connected to a
Storage Area Network (SAN) with redundant connections from the HP ProLiant BL20p
Enclosure. Two HP StorageWorks Enterprise Virtual Array 5000’s (EVA5000) were fully
populated with one hundred and sixty eight 36.4 fibre channel disks. Each HP
StorageWorks Enterprise Virtual Array 5000 was configured to present a single 200-
3 gigabyte logical unit to each HP ProLiant BL20p server blade in the enclosure, so that each
blade had two 200-gigabyte logical units, one on each HP StorageWorks Enterprise Virtual
Array 5000. The detailed hardware configuration and test procedure are provided in
Section 3 of this report
Conclusions
The R6Mail workload was run to simulate 120,000 R6Mail users against a Domino 6.0.2
Server. The test ran 7 hours and during which the system under test achieved 175,214
NotesMark (transactions per minute or tpm) with an average response time of
0.384 seconds. Based on these results, the price/performance ratio is $8.67/User (or
$5.94/NotesMark).
4 Section 2 – Benchmark Objectives
A question that is asked by most of our Notes customers is: “What is the maximum number
of active Domino user sessions for our environment?” The answer is determined by three
criteria:
• The application profiling (what type of workload)
• The hardware configuration (for example, the number of system CPUs or the amount of
memory)
• The operating system (OS)
HP feels that the information provided in this report will help our customers in their planning
processes. With the publication of this and future full disclosure reports, HP and the
IBM/Lotus software group are working together to provide information that will help HP
Notes users in their activities associated with performance tuning, enterprise network design,
and many other areas.
This report includes one workload:
• R6Mail on Domino 6.0.2
All tests were performed on the Microsoft® Windows® 2000 Server platform; similar
performance information for other hardware configurations is forthcoming.
5 Section 3- Test Methodologies
Test configuration
The HP Performance Test Lab for Lotus Domino uses HP ProLiant BL10e systems, configured
with IBM/Lotus Notes R6 client with Notesbench. Each HP ProLiant BL10e server has an 800
MHz processor, 40 GB hard disk, and 1024 MB of RAM.
For the testing of R6mail, eighty HP ProLiant BL10e systems were used to create load.
Additionally, a HP ProLiant DL360 system was deployed to act as the Parent system. All
systems were connected over an isolated TCP/IP LAN using100BaseT or 1000BaseT
medium. All drivers were loaded with Microsoft Windows 2000 Server with Service Pack 3;
the blades in the enclosure under test were loaded with Microsoft Windows 2000 Server
with Service Pack 3.
All Windows 2000 Server parameters were constant for the remaining benchmark
sequences. Any Domino Server specific parameter changes are bold in Appendix D’s
NOTES.INI files.
Test procedure
For this workload, HP Engineers first performed several trial runs to determine the best test
duration and confirmation of Steady State. Both test duration and steady state were
determined using real time monitor utilities from Microsoft. During the trial runs, HP
engineers observed the Windows Performance Monitor to determine the maximum number
of users supported by the configuration.
Based on pervious testing for HP ProLiant BL20p servers and IBM/Lotus Domino R6.0.2, HP
engineers knew two things. The first thing identified was that a single HP ProLiant BL20p
server with 2.8GHz CPUs could adequately support a total of 15,000 R6mail users. The
second thing HP engineers identified was that a single Lotus Domino partition under a
R6Mail workload ran best under a load of 7,500 R6mail users. By using these two known
points, each blade in the HP ProLiant BL20p Enclosure was assigned two RAID 0+1 200
gigabyte logical units from the SAN (one from each HP StorageWorks Enterprise Virtual
Array 5000). The Enterprise Virtual Array uses storage system level virtualization; or put
more simply, disks within the EVA are not directly assigned to a particular logical unit. The
controllers within the Enterprise Virtual Array maintain data across a group of disks, and
create logical disks that span the group to be presented to the servers. More information of
disk virtualization and the HP StorageWorks Enterprise Virtual Array 5000 can be found at
http://www.hp.com.
Having set up the storage system, the Lotus Domino partitions were then created across the
blades with the first and ninth partition existing on the first blade, the second and tenth
partition on the second blade, and so forth. Then, knowing that the drivers could support
1500 R6mail users each, and using the understanding that Notesbench starts users in a
sequential fashion, HP engineers created a Lotus Domino Directory in which the sequential
spread of users was across the sixteen domino partitions, thus allowing HP to start load
against all the Lotus Domino partitions in a simultaneous fashion. This sequential spread of
users across the Lotus Domino environment allowed HP engineers to fully ramp the HP
ProLiant BL20p Enclosure to 120,000 R6mail users in only 5 ½ hours from start. Testing
began with the clients being added incrementally to the Lotus Domino environment using a
6 Childstagger value setting at 3 minutes for the first 65 drivers, and 6 minutes for the
remaining 15 drivers as specified in the parent Notes.ini. The Client systems used a
Threadstagger of .6 seconds for all 80 drivers, as specified in each client Notes.ini. All
drivers created 1500 R6mail users each.
The Lotus Domino environment was created with single Lotus Domino directory that
contained 240,000 mail user entries. Each Lotus Domino server had its own copy of that
directory, and mail was routed between all 16 partitions for delivery.
Each test ran for more then seven hours after all client threads were connected to the system
under test, where over 90% of the mail generated during the run was delivered during the
run. Steady state was determined to be achieved during each test run by monitoring server
output, as well as mail routing, calendar and scheduling activities. The test data collected
includes the following files for each partition:
• The SUT Notes.INI file and Log.nsf file
• All client’s Notes.INI file
• The NotesNum utility results file
• Stat Tables for each partition after run completion
7 Section 4- R6Mail Data
Lotus NotesMark Values for Benchmarked Configurations

Figure 1. HP ProLiant BL20p Enclosure NotesMark
BL20p Enclosure - Notesmark
200000
175214
180000
160000
140000
120000
100000
80000
60000
40000
20000
0
Average Notesmark


• NotesMark – transactions per minute or tpm
8
Notesmark (Transactions Per Minute)
Figure 2. HP ProLiant BL20p Enclosure – NotesMark response time

BL20p Enclosure - Response Time
1
0.9
0.8
0.7
0.6
0.5
0.384 Seconds
0.4
0.3
0.2
0.1
0
Average Response Time
Mail-Only Users per Server

Figure 3. HP ProLiant BL20p Enclosure – R6 Mail Users
BL20p Enclosure - R6mail Users
140000
120000 R6 Mail
Users!!
120000
100000
80000
60000
40000
20000
0


9
Notesrv1
Notesrv2
Notesrv3
Notesrv4
Notesrv5
Notesrv6
Notesrv7
Notesrv8
Notesrv9
Notesrv10
Notesrv11
Notesrv12
Notesrv13
Notesrv14
Notesrv15
Notesrv16
Total Users
Seconds
R6 Mail Users
CPU Performance as Recorded by Lotus Domino

Figure 4. CPU Performance
BL20p Enclosure - Maximum CPU Utilization
100
99.9
99.8
99.7
99.6
99.5
99.4
99.3
99.2
99.1
99


10

Un pour Un
Permettre à tous d'accéder à la lecture
Pour chaque accès à la bibliothèque, YouScribe donne un accès à une personne dans le besoin