0 Title1 About Developer Build 22 Before You Begin3 What's New in This Build4 Known Problems and Caveats5 Application Programming InterfaceA Using Notes R4 and R3.x on the Same Workstation
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Decision Maker Databaseoutput vendor data form - if provided and had grbphs, easy way for decision maker to review how the current system compares to other systems and/or performs under a variety of workloadsUsing the Decision Maker Database
CN=Carol Zimmet/O=Iris
O=Lotus Notes
O=Lotus Notes
PURSAFO
|.:#U
O=Lotus Notes
CN=Lotus Notes Template Develo|ment/O=Lotus Notes
PURSAFO
QC^j#
$INFO
$BODY
O=Lotus Notes
O=Lotus Notes
PURSAFO
|.:#U
O=Lotus Notes
CN=Lotus Notes Template Development/O=Lotus Notes
PURSAFO
$INFO
$BODY
O=Lotus Notes
O=Lotus Notes
PURSAFO
|.:#U
O=Lotus Notes
CN=Lotus Notes Template Development/O=Lotus Notes
PURSAFO
K$p'>
$ACLDigest
O=Lotus Notes
O=Lotus Notes
PURSAFO
|.:#U
O=Lotus Nltes
CN=Lotus Notes Template Development/O=Lotus Notes
PURSAFO
$Author
$Index
$Formula
$FormulaClass
$ViewFormat
O=Lotus Notes
O=Lotus Notes
PURSAFO
|.:#U
O=Lotus Notes
CN=Lotus Notes Template Development/O=Lotus Notes
PURSAFO
$Author
$Index
$Formula
$FormulaClass
$ViewFormat
Times New Roman
New Chapter
Subject
DocLink Head
Table Text
Column Head
Table Title
formula
procdure Chead
procdure Bhead
indent Term
indent find
indent caution
caution
indent tip
indent Note
1st lev para
Numbered List
Body Text
Ahead
DocLink
indent bullet
2nd lev para
bullet
Bhead
Chead
Figure Title
Table of Contents
Contents
Index
Index
Print...
Sectiont
chap_numberEnter chapter number.
ChapterTitle
SubjectEnter the Chapter Title for this chapter.
O
Doc
Doc_Number
Indent?
Indent
Index entries
Index_EntriesList of index entries (separate entries with a carriage return)
Categories m
Bookb
Section
Categories^
Categories
BodyWrite the body of the document here.
(hidden)
DocAuthor
O=Lotus Notes
O=Lotus Notes
PURSAFO
|.:#U
O=Lotus Notes
CN=Lotus Notes Template Development/O=Lotus Notes
PURSAFO
$Author
$INFO
$WINDOWTITLE
Doc_Number
Indent
$BODY
OriginalSubject
OriginalSubject
OriginalSubject
Subject
NewRespTitle
New Respon|e to "
OriginalSubject
StandardTitle
Response
of "
to "
OriginalSubject
ByAuthorTitle
Response to "
OriginalSubject
NewRespTitle
AuthorView
ByAuthorTitle
StandardTitle
ReviewResponse
Author:
FromThe name of the person who composed the response.
Date:
DateThe date/time when the response was composed.
DocumentName:
Subject
DocumentName
Subject:
Subject
OriginalSubject
OriginalSubject
Subject
OriginalSubject
Hidden field to store the subject of this document's main parent document
O=Lotus Notes
O=Lotus Notes
PURSAFO
|.:#U
O=Lotus Notes
CN=Lotus Notes Template Development/O=Lotus Notes
PURSAFO
$Author
$INFO
$WINDOWTITLE
$BODY
Times New Roman
New Section
Subject
Figure Title
1st lev para
Numbered List
DocLink
Body Text
2nd lev para
Ahead
Bhead
bullet
Chead
indent bullet
DocLink Head
indent Note
indent tip
caution
indent caution
indent find
indent Term
procdure Bhd
procdure Chd
formula
Table Title
Column Head
Table Text
Table of Contents
Contents
Index
Index
Print...
,
Section
Ahead
SubjectEnter the Ahead for the section.
Doc
Doc_NumberType order of document within chapter.
Indent?
Indent
Index entrie|
Index_EntriesList of index entries (separate entries with a carriage return)
Categories O
Bookb
Section
Categories^
Categories
BodyWrite the body of the document here.
(hidden) a
DocAuthorT
O=Lotus Notes
O=Lotus Notes
PURSAFO
|.:#U
O=Lotus Notes
CN=Lotus Notes Template Development/O=Lotus Notes
PURSAFO
J mL
$Author
$INFO
$\INDOWTITLE
Doc_Number
Indent
$BODY
[;86x
About This Document
This database is a Notes online book.
Choose
Help - Using
for ins|ructions on viewing and modifying this book.
A Notes online book has two views:
Table of Contents
- Where each subcategory is a chapter and each document is a section. This view contains all the documents designed for online viewing.
Index
Where each category is an index entry and each document is a section to which the index entry refers.
To suit the needs of your site, you can edit the documents in this database. However, if you do edit the contents of this database, it must be strictly for the sole
|e of users within your organization
. You cannot resell or otherwise distribute this documentation, modified or unmodified, to anyone outside your organization. Lotus assumes no responsibility for the technical accuracy of any modifications made to this documentation or the operation of Lotus products in reliance on such modifications.
Lotus Development Corp.
http://www.lotus.com
O=Lotus Notes
O=Lotus Notes
PURSAFO
|.:#U
O=Lotus Notes
CN=Lotus Notes Template Development/O=Lotus Notes
PURSAFO
%_]!dv
$INFO
$BODY
Figure Title
Chead
Bhead
bullet
2nd lev para
indent bullet
DocLink
Ahead
Body Text
Numbered List
1st lev para
indent Note
indent tip
caution
indent caution
indent find
indent Term
procdure Bhead
procdure Chead
formula
Table Title
Column Head
Table Text
DocLink Head
bhead
What is a Benchmark
body text
benchmark
is a software application that tests the performance of a computer system. Benchmarks can test the following:
bll_onl
Two software applications running on the same hardware
Different hardware platforms from the same vendor running the same software
Different releases of software on the same machine
Different systems running the same software
Workloads
Workloads are a pre-determined set of transactions to be executed by the hardware and soft|are under test. The software is executed by driver systems to put a measurable load on the system under test (SUT).
gloss_item_dd
NotesMark
bodytext_dd
The Notes throughput metric of transactions-per-minute (tpm).
Performance
The purpose of a NotesBench benchmark is to obtain and record performance measurements to compare two or more systems. The systems' performance is measured by a throughput metric - usually in units of work/second. For example, the performance of full text search software on each benchmarked system is typically search transactions completed per second.
cheaddd
Prlce
The price of the system under test is an integral part of a benchmark, along with its performance metric. The price metric used is the five-year cost-of-ownership of the system. NotesBench provides guidelines for calculating the system price.
Price/Performance
Price/performance is a benchmark's price and performance looked at together as a ratio: price divided by performance. Price/performance lets you weigh the work done by the software agains| the cost of ownership.
Scalability
NotesBench is designed for scalability; that is, the workloads are applicable to both large and small computer systems. A scaleable benchmark maintains a constant relationship between the workload presented to a system and the capacity of the system. As the capacity increases the benchmark's workload increases proportionally. Scalability allows the benchmark user to adjust the workload to the size of the system under test.
Bhead
Steady State
bodytext_dd
Steady state is the condition of the system under test that represents its true sustainable performance with reproducible results. Any NotesBench test reaches steady state when the number of Notes users on the system under tlst is equal to the sum of the number of client users across all child drivers. Additionally, ten minutes should have elapsed since all threads have become active and performed at least one successful iteration of the workload script.
Typical Results
Benchmark results are typically presented in the form shown in Table F-1. The systems under test are listed along with the performance in NotesMarks (tpm or transactions per minute) and the price/performance ratio (k$,tpm). These benchmark results apply to different systems running the same software application.
Table F-1 Typical Benchmark Results
System Under Test
k$/tpm
System A
System B
19.0
System C
13.5
User Classification
Body Text
Users, in benchmarking, fall into three classifications:
Assigned (users having a database on the server)
Connected (idle session is being maintained on the server)
Active (causing transactions to be reported on the server; for example, TPM)
Benchmark Standards
There is a stronl trend toward standard benchmarks within software areas of interest or domains. Groups of vendors define standard benchmarks for their domains. For example, the Transaction Processing Performance Council (TPC) approves benchmarks in transaction processing and database applications; and the Standard Performance Evaluation Corporation (SPEC) defines benchmarks in the workstation and scientific areas.
bhead
Yd%hN
p@_AY
Using This Book
This book can be viewed online as a Notes database and modified for your site.
Viewing a Notes Book Online
To View Topics in Logical Orderr
Choose the e
Table of Contents
view. This is the main view. You see the book title, chapter titles indented under the book title, and section titles indented under each chapter title. Each section comprises one document.
To View Topics Alphabetically
Choose the
Index
view. You see a list of topics in alphabetical order, and a section title or titles indented under each topic. To find a topic fast, type the first few letters of the topic.
To Open a Document
Select it, then double-click it, or press
ENTER
. A document has the following appearance:
Display
?_?__?
???__?_
?_?_____
At the top of each document are several buttons. When you click one, the appropriate action takes place.
Button
Action
Table of Contents
Contents
Displays all subjects in the database in logical order (opens the Table of Contents view).
The current document remains open; you return to it when you exit from the Index view.
Index
Index
Displays all topics in the database in alphabetical order (opens the Index view). The current document remains open; you return to it when you exit from the Index view.
Print...
Prints the current document.
The next line is the name of the chapter (first document in a chapter) or a section in the chapter (other than the first document in the chapter).
To Use a Doclink
This symbol
Database 'Book Model V4', View 'Table of Contents', Document 'About This Book'
is called a doclink (short for document link). Double-click it, or arrow to it and press
SPACEBAR
to go to another document relevant to the text next to the doclink. Close the document or select
BACKTRACK
to return to the document where you selected the doclink. The doclink in this paragraph goes to the "About This Book" document for this database.
Using Full Text Search
Choose
View - Show Search Bar
to use the Notes Full Text Search feature.
NOTE
Before you can use Full Text Search, the database you want to search must have a full text index -- a special copy of the text of a database that allows Notes to provide query results almost instantly. You can create full text indexes and use full text searching on
local
databases unless you're using Macintosh. Select or open the database and choose
File - Full Text Search - Information
to see if the database is indexed and whether the index is up to date. To create an index, choose
File - Full Text Search - Create
. To update the index, choose
File - Full Text Search - Update
. You can perform a full text search on a
shared
database (on a server) only if someone with Designer access to the database created an index. See the Help document, "Indexing Databases for Full Text Search," or look up "Full Text Search" in the Help Index view.
With full text search, you can search for whole words, word stems, and phrases, as well as perform more complex queries using logical operators (Boolean searches), proximity, and other advanced features. You can search one database, or many databases. A single set of text search criteria is called a query.
Look up "Full Text Search" in the Help Index view for additional information.
Creating or Modifying a Notes Book
Structure of a Book Database
A book database has two views: Table of Contents and Index; and two forms: Chapter and Section. The Chapter form is for the first document in a chapter. The Section form is for the remaining documents in a chapter.
The Table of Contents view reflects the general structure of the book:
Book Name
0 About This Book
Chapter document
1 Chapter Name
Chapter document
Section document
as necessary
Section document
as necessary
2 Chapter Name
Chapter document
Section document
as necessary
Section document
as necessary
additional chapters as necessary
as necessary
A Appendix Name
if necessary
Chapter document
Section document
as necessary
Section document
as necessary
appendices as necessary
The numbering and lettering as shown is recommended to keep the parts of the book in correct order._
To Unhide the Forms on the Compose Menu
A book database hides its forms so readers do not see them on the
Compose
menu. You must unhide the forms you want to use to compose documents. When you finish, you should hide the forms again.
Choose
Design - Forms
. Select the name of a form and click
. When the form opens, choose
Design - Form Attributes
. To unhide the form, select
Include in Compose Menu
and click
?_?__?
???__?_
?_?_____
Color
Black
Dithered Gray
Choose
File - Close Window
to close the form. Perform the above procedure for each form that you need.
To hide a form, use the above procedure, but deselect
Include in Compose Menu
To Set Up the Book Parts
Choose
Design - Shared Fields
. Select Section and click
. When the Shared Field Definition box opens, click
Format
. Type in the names of the book parts at the chapter level, as shown below.
W4W10F/V2
WordPerfect
?_?__?
???__?_
?_?_____
Color
Black
Dithered Gray
out This Book
2 Forms
3 Sty
ckbox
Radio butt
Then click
, and
again.
Choose
Design - Shared Fields
again and this time select Book and click
. When the Shared Field Definition box opens, click
Format
. Type in the name of the book, as shown below:
W4W10F/V2
WordPerfect
?_?__?
???__?_
?_?_____
Color
Black
Dithered Gray
@Replace
You may want more than one name for this category. Your database may include multiple books or may be divided into parts. For a small database that represents one book, a single name is recommended.
To Create a Document
Choose
Compose
and the appropriate document form. The following form is for a Section document.
?_?__?
???__?_
?_?_____
Color
Black
Dithered Gray
olors (%d%%)
Press
ENTER
on the first field and select the name of the book. Press
ENTERm
on the second field and select the appropriate chapter.
On the Section form, type the section name into the third box. On the Chapter form, type the chapter number and the name.
On the Section form, be sure to enter the order of the document within the chapter.
If you want the document name indented in the Table of Contents view, change the value of
Indent?
from 0 to 1.
Type in the index entries that apply to this document, one entry per line.
Finally, enter the body of the document.
To Use the Notes Book Styles
The following styles are defined. (The paragraph styling is defined under
Text - Define Style
. The font styling must be adjusted manually.) The left margin for headers and text (unless otherwise specified) is 1".S
Body Text
Name Font Alignment Inter Above Below2
chap_number Helv24B Left 1 Single Single 1 1/2 Hairline rule below
ChapterTitle Helv24B Left 1.75 Single Single 1 1/2 Keep with next, rule belowm
Body Text Helv10 Left 1.75 Single Single Singlee
Ahead Helv12B Left 1 Single 1 1/2 Single Keep with next, rule below
Bhead Helv11B Left 1.25 Single Single Single Keep with next
Chead Helv10B Left 1.75 Single Single Single Keep with next
Numbered List Helv10 Left 2, 1.77 Single Single Single
bold
1st lev para Helv10 Left 2 Single Single Single
bullet Helv10 Left 2, 1.79 Single Single Single
bold
indent bullet Helv10 Left 2, 2.15 Single Single Single S
boldn
2nd lev para Helv10 Left 2.15 Single Single Single
Note Helv10 Left 1.75, 2.25 Single Single Single m
Note
Helv10B
indent Note Helv10 Left 2, 2.5 Single Single Single
Note
Helv10B
tip Helv10 Left 1.75, 2.125 Single Single Single
Helv10B
indent tip Helv10 Left 2, 2.375 Single Single Single
Helv10B
caution Helv10 Left 1.75, 2.438 Single Single Single t
Cautionn
Helv10B
indent caution Helv10 Left 2, 2.688 Single Single Single
Caution
Helv10B
find Helv10 Left 1.75, 3.125 Single Single Single
FINDING ...
Helv10B
indent find Helv10 Left 2 , 3.375 Single Single Single
FINDING ...m
Helv10B
Term Helv10B Left 1.75 Single Single Single Keep with next
indent Term Helv10B Left 2 Single Single Single Keep with next
procdure Bhd Helv11B Left 1.25 Single Single Single Keep with next
procdure Chd Helv10B Left 1.75 Single Single Single Keep with next
formula Helv9 Left 2 Single Single Single
mono Cour9B Left 1.75 Single Single Single
Table Title Helv9B Left 1.75 Single Single Single Keep with next
Column Head Helv9Ital Left 1.75 Single Single Single Keep with next"
Table Text Helv9 Left 1.75 Single Single Single
Figure Title Helv9B Left 1.75 Single Single Single
DocLink Head Helv10B Left 1.75 Single Single Single Keep with next
DocLink Helv10 Left 1.75, 6 Single Single Single
O=Lotus Notes
O=Lotus Notes
PURSAFO
|.:#U
O=Lotus Notes
CN=Lotus Notes Template Development/O=Lotus Notes
PURSAFO
$INFO
$BODY
ChapterDomino Server.Planner User GuideBNotesBench Script Results
Metrics SummaryWeb BuyerWeb WalkerDatabaseMailMail & DatabaseCalendar & SchedulingGroupwareMail Routing HubReplication HubUser TasksCN=Cheryl Wilson/OU=CAM/O=LotusAppendix BDOMINO SERVER.PLANNE\ USER GUIDE\Appendix B
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Vendor Databasedid you want to talk up the graphical compare facility found in Vendor Data? Once the vendor data is out there, it will be a powerful feature.Using the Vendor Database
CN=Carol Zimmet/O=Iris
?UUUUUU
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisGetting Starteddspd.ntf for decision maker database - let me know if you really want me to change (it would cause some confusion here)Getting Started
CN=Carol Zimmet/O=Irls
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Decision Maker Databasequery output - creates a doclink to machine id formUsing the Decision Maker Database
CN=Carol Zimmet/O=Iris
g0.3M _
ChapterDomino Server.Planner User Guide4Introduction
CN=Cheryl Wilson/OU=CAM/O=Lotus4 Using the Analyst DatabaseDOMINO SERVER.PLANNER USER GUIDE\4 Using the Analyst Database
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Vendor Databaseon the machine id form, many of the fields also have the opportunity for a Vendor to specify info if their system is not present (like OS Platform, CPU type). There is an issue for sole of the fields though that get queried on, like if the OS Platform information is written in, it wouldn't be selected for the query. This would be done in anticipation of a template upgrade (where the writein value agrees with the template change.)Using the Vendor Database
CN=Carol Zimmet/O=Iris
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisGetting Startedmay no longer be Opening Navigator, may have to give examplesGetting Started
CN=Carol Zimmet/O=Iris
Review Responsl
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Vendor Databasesave and edit options will go away when the form is certified (and saved to the decision maker database/mailed item)Using the Vendor Database
CN=Carol Zimmet/O=Iris
"^D\SZ
Re|iew Response
ReviewResponseCN=Carol Zimmet/O=IrisNotesBench Script ResultsI refer to the test as Database. Should I change to Shared Discussion?NotesBench Script Results
CN=Carol Zimmet/O=Iris
Review Response
Revie|ResponseCN=Carol Zimmet/O=IrisUsing the Decision Maker Databaseview results - give example of outputUsing the Decision Maker Database
CN=Carol Zimmet/O=Iris
"Univers Condensed
Figure Title
Chead
Bhead
bullet
2nd lev para
indent bullet
DocLink
Ahead
Body Text
Numbered List
1st lev para
indent Note
indent tip
caution
indent caution
indent find
indent Term
procdure Bhead
procdure Chead
formula
Table Title
Column Head
Table Text
DocLink Head
Body Text
This appendix contains material from the
NotesBench User's Guide
Brief Description of each workload
Full description of each workload
Workload Metrics Summary (a table)
When using the information in this section, please note that for this release of Domino Server.Planner :
Idle Usage is not supported.
Cluster Mail and Shared Database is not supported.
Groupware A is not supported.
What was formerly known as Groupware B is now referred to as Groupware #1
Shared Discussion Database is referred to as Database.
Mail and Shared Database is referred to as Mail & Database.
NotesBench converts transactions to a user count in the Web workloads.
bodytext_dd
Bhead
Workloads: Brief Description
bll_onl
Chead
Web Buyer
(User Tasks)
bodytext_dd
This workload simulates several interactive web browsers using an on-line catalog and making purchases. The process includes filling out an order form, selecting items, and completing the credit card purchase.
Web Walker
(User Tasks)
This workload simulates web users browsing a web site built on top of Domino. The user will peruse each link on the selected test database retrieving the
content of each web page, thereby providing a realistic load against the server.
Calendar and Scheduling
(User Tasks)
This workload stresses a Domino server's ability to process Calendar and Scheduling invitation requests. Each invitation is sent to multiple recipients. This causes work to be performed by the Domino mail router. It requires updates to the System Under Test Free Time database.
bodytext_dd
Shared Discussion Database
(User Tasks - Database)
This workload simulates a server for active users, who are only performing heavy shared database operations. The test includes view operations in a shared database, navigation of unread documents, additions, and updates to documents in a shared database. It applies especially to sites that heavily utilize the collaborative features of Dominc.
Groupware #1
(User Tasks)
This workload simulates a
server for power users who are sending large mail messages, updating documents with attachments to shared databases, and replicating changes from their local machine to the server.
(User Tasks)
This workload simulates a server for mail users workload that models sites that rely only on mail for communication.
Mail and Shared Database
(User Tasks - Mail & Database)
This workload simulates a server for active users who are performing only mail and simple shared database operations.
Mail Routing Hub
(User Tasks)
This workload simulates a mail routing hub that routes messages to other servers (a "pure" router) and may also deliver messages to local users.
Replication Hub
This workload psopagates changes among a collection of other servers.
Workloads: Complete Description
About the Web Buyer Usage Test
(User Tasks)
Web Buyer is a test that simulates several interactive web users browsing a fictional company's on-line web catalog and making purchases on the web.
Each simulated user peruses the on-line catalog, taking about five seconds to browse each of ten links that describe various products for this fictional company. The simulated user then decides to purchase five products at $100 each via the web. The purchasing process includes filling out an on-line order form that includes user name, address, selected products, credit card information and payment method. The order process concludes with the user clicking the "order" button on the web form. The order is "processed" by the Domino server, and a "Thank You for your order" is returned to the user indicating successcul order completion.
The catalog database used for displaying the product information, and the orders database is used for collecting the orders are two separate databases, thus further demonstrating the ease of creating and managing web applications under Domino.
Note: For simplicity and uniformity, each on-line user orders five products, each product priced at $100.00. Thus every successfully completed order results in $500 of revenue. (Conversecy, each lost order results in $500 of lost revenue).
The measurements obtained by this test are:
Total Number of orders generated
Total Number of orders lost (due to errors)
Average Time to complete and Order
Gross Revenue generated from Web Orders
The number of web users is equal to the total number of threads becng driven across all child drivers.
The other metrics reported by Web Walker are available, but this is the minimum required for a NotesBench Report.
About the Web Walker Usage Test
(User Tasks)
Web Walker is a test that simulates several users perusing a popular Notes database via a web browser.
Starting at a default "home" URL, the simulated user opens each link on the home page, and retrieves the
content (all graphics and text) of each page. For each page opened, the simulated user will retrieve all pages linked to that page (thus the term "web walker"). For each page fetched the simulated user will spend five seconds reading and observing the graphics on each page. The simulated user is assumed to be accessing the HTTP server via a browser that can retrieve four different web objects (graphics, text, or binary data) concurrently.
The test more recdily simulates a true web load against the HTTP server in that many different URLs (pages, or Notes documents) are being served "concurrently". Many web tests simply establish a connection and then disconnect against the same URL (which ends up being cached).
Note: The web browser test does not visit links that links to pages that are not on the local web server. Cor does it execute actions (i.e., POST methods). For the interactive web test, please see the Web Buyer section.
The measurements obtained by this test are:
Total Number of Pages visited
Average Bytes/Page
Average Time to Retrieve Full Content of Page
Total Number of Hits
Total Number of Dropped Connections
Total Number of Time-outs
Total Number of Server Errors
This workload determines the maximum number of concurrent HTTP users that a Domino HTTP Server can support. The number of web users is equal to the total number of threads being driven across all child drivers.
About the Calendar and Scheduling Test
(User Tasks)
This workload stresses a Domino server's ability to process Calendar and Schedulinc (C&S) invitation requests. Each invitation is send to multiple recipients. This causes work to be performed by the following subsystems:
Domino Mail Router:
indent Term
To send invitation requests to the recipients
System Under Test FreeTime Database:
Any C&S activity such as the creation of appointments, alarms, or invitations, requires an update to the FreeTime Database located on users
home mail server.
The script models an active user who is creating invitation requests:
The user creates an invitation with a message size of [NormalMessageSize] and sends that invitation to [NumMessageRecipients].
The invisation date is a randomly chosen date between today and 90 days from today.
The invitation time is always some future time from the child's system time.
The invitation specifies a meeting duration in [MeetingDuration]. [MeetingDuration] is hard coded at 15 minutes.
The resultant capacity metric for a C&S and mail-only server is the maximum number of users that can be supported before the average user response time becomes unacceptable.
The measurements obtained by this test are:
Throughput of completed Notes operations
Average response time at maximum capacity
Maximum C&S users supported
About Shared Discussion Database Test
(User Tasks - Database)
This workload models a server for active users who are only performing heavy shared database operations.#
The test includes view operations in a shared database, navigation of unread documents, additions, and updates to documents in a shared database. It applies especially to sites that heavily utilize the collaborative features of Domino.
The script models an active user who is reading, composing, and updating documents in a shared database.
The throughput for this test is a capacity metric. It is the maximum number of actise users that can be supported before the average user response time becomes unacceptable.
The measurements obtained by this test are:
Throughput of completed Notes operations
Maximum users supported
Average response time at maximum capacity
About the Groupware #1 Test
(User Tasks)
This workload models a server for experienced Notes users who are sending large mail messagcs, adding documents with attachments to shared databases, performing full-text searches, and replicating changes from their local machine to the server.
The test is designed to run for extended times (i.e. 6-8 hours). This test no longer requires the creation of local (i.e. on the driver system's desk) *.txt files.
bodytext_dd
Groupware #1
test cncludes mail and shared database activity plus:
Mail messages with 532KB messages
Users that replicate with the system under test
bodytext_dd
Groupware #1 is a capacity test for Notes users that process large amounts of information. This workload models sites that use the most resource-intensive features of Notes. It can be used to cstablish a worst-case lower bound on the maximum number of users a server can support.
This script is similar to the Mail and Shared Database script, but includes more operations and is more resource intensive. The differences are:
Each view scrolling operation is for 40, rather than 20 rows.
The mail messages are 532K instead of only 1K in size.
The script makes changes to a local
discussion
database and#pushes those changes to the server.
Generated messages are delivered to user mail files as specified in the directory database (
names.nsf
) of the system under test. The updates done in the Groupware test are intended to model reading and submitting messages, but not message delivery or reading delivered messages.
The measurements obtained by this test are:
Throughput of completed Notes operations (NotesMcrk)
Maximum user supported
Average response time at maximum capacity
The resultant capacity metric for a power user server is the maximum number of users that can be supported before the average user response time becomes unacceptable.
About the Mail Test
(User Tasks)
This workload executes Notes transactions that model c server for mail users at sites that rely only on mail for communication. The resultant capacity metric for a mail-only server is the maximum number of users that can be supported before the average user response time becomes unacceptable.
The script models an active user who is reading and sending mail. It contains an average of 15 minutes of waiting, so an average user will execute this script no more frequently than 4 times per hour. For each iteration of the script, there cre 5 documents read, 2 documents updated, 2 documents deleted, 1 view scrolling operation, 1 database opened and closed, 1 view opened and closed, and some miscellaneous operations. In sending messages, each user sends a mail message
no more frequently than every 90 minutes.
About the Mail and Shared Database Test
(User Tasks
- Mail & Database
This workload models a server for active users, who are only percorming mail and simple shared database operations.
The test includes mail-only activity plus view operations in a shared database and navigation of unread documents in a shared database. It applies especially to sites that rely primarily on mail for communication or that have Notes users who do not yet use all Notes features.
The throughput for this test is a capacity metric. It is the maximum number of active users that can be supported before thc average user response time becomes unacceptable.
The script models an active user who is reading mail, sending mail and reading a shared database. It contains an average of 15 minutes of waiting, so an average user will execute this script no more frequently than 4 times per hour. For each iteration of the script, there are 8 documents read, 2 documents updated, 2 documents deleted, 4 view scrolling operations, 2 databases opened and closed, 2 views opened and closed, and soce miscellaneous operations. In sending messages, each user sends one mail message approximately once every 90 minutes.
About the Mail Routing Hub Test
A mail routing hub is a server that exists to route messages to other servers (a "pure" router) and possibly to deliver messages to local users. The workload for a mail routing hub
consists of receiving messages from source systems and routing each message to a destination system. #
The mail routing hub script generates the input message traffic to the system under test for both kinds of mail hub tests. The test selects message recipients with a uniform distribution from the People view in the driver's Name and Address Book. Note that this script is not intended to be a realistic simulation of any kind of human message generation. Instead it is intended to generate a continuous input stream of messages to the system under test acting as a mail hub routes.
About the Replication Hub Test
A replication hub is a Notes server that exists to propagate changes in Notes databases among a collection of other servers. The workload for a replication hub consists of replicating changes to user databases.
The test procedure for a replication hub uses a hub-and-spoke topology. A number of driver systems serve as sosrce and destination spokes and the system under test serves as a single replication hub. The spoke servers modify local replica databases and then replicate those changes to the hub server. Each test script user makes changes to one local database and then replicates those changes to the server. The server runs the replicator and updater, but no other server programs. The test performs additions, updates, categorizations, and deletions on the spoke systems.
Woskload Metrics Summary
Each NotesBench test (workload) produces at least one metric that can be used to measure server performance. This table lists the metrics provided by each test.
Results
Web Buyer
Total Number of orders generated
Total Number of orders lost (due to errors)
Average Time to complete an order
Gross Revenue generated from web orders
Web Walker
Total number of Pages Visited
Average Bytes/Page
Average Time to Retrieve Full Content of Page
Total number of Hits
Total number of Dropped Connections
Total number of Time-outs
Total number of Server Errors
Calendar and Scheduling
Throughput of completed Notes operations (NotesMark)
Maximum C&S users supported
Average response time at maximum capacity
Shared Discussion Database
Throughput of completed Notes operations (NotesMark)
Maximum users supported
Average response time at maximum capacity
GroupWare B
Throughput of completed Notes operations (NotesMark)
Maximum GroupWare_B users#supported
Average response time at maximum capacity
Replications per minute
Mail
Throughput of completed Notes operations (NotesMark)
Maximum MAIL users supported
Average response time at maximum capacity
Mail and Shared Database
Shroughput of completed Notes operations (NotesMark)
Maximum MailDB users supported
Average response time at maximum capacity
Mail Routing Hub
Throughput of completed Notes operations (NotesMark)
Messages transferred to recipient per minute
Message bytes transferred to recipient per minute
Number of spoke servers supported
Replication Hub
Throughput of completed Notes cperations (NotesMark)
Average response time at maximum capacity
Number of spoke servers (users) supported
Replications per minute
Table Title
NotesBench Performance Metrics
g0.3M _
indent Term
Chead
Figure Title
Chead
Bhead
bullet
2nd lev para
indent bullet
DocLink
Ahead
Body Text
Numbered List
1st lev para
indent Note
indent tip
caution
indent caution
indent find
indent Term
procdure Bhead
procdure Chead
formula
Table Title
Column Head
Table Text
DocLink Head
Bhead
About the Server.Planner Analyst Database
Body Text
The Domino Server.Planner
Analyst Database
is a tool for creating and managing queries that you make against
Vendor Databases
. Ct stores the results in a
Decision Maker Database
The Analyst Database presents two views:
Query on Data
Conversion Guidelines
You use the Query on Data view to create, use, and manage your queries.
You use the Conversion Guidelines view tc adjust the search requirements for specific workloads. This feature of Domino Server.Planner allows you to adjust the queries to reflect real-life usage instead of the laboratory environment of the benchmark scenarios where the vendor data was created.
Each view has an action button bar for ease of use. You can also perform any of these actions from the menu bar.
Server.Planner queries only certified Vendor Data and Machine Identification forms.
General Instructions
If your Notes Data directory does not contain a Server.Planner Analyst database, see the section below, 'Creating the Analyst Database', for instructions on how to create one.
For input to the queries, you must have one or more vendor databases in your Notes
Data
directory. See the section below, "
Downloading and Using the Vendor Databases,
" for instructions on how to obtain these databases.
To start, open the Domino Server.Planner Analyst database by double-clicking on its icon on your workspace, or use
File Database > Open
to open it.
To see all of the view pane, you may need to use the horizontal scroll bar. (Use
View Show > Horizontal Scroll Bar
if the bar is not present at the bottom of the pane).
Use the Query Data view to cseate queries and work with those already in the database.
Use the Conversion Guidelines view to adjust the weighting of benchmark results.
To end the session, click on the
Exit
button, or press <ESC>.
Table Text
Column Head
Table Title
formula
procdure Chd
procdure Bhd
indent Term
indent find
indent caution
caution
indent tip
indent Note
DocLink Head
indent bullet
Chead
bullet
Bhead
Ahead
2nd lev para
Body Text
DocLink
Numbered List
1st lev para
Figure Title
Body Text
Bhead
Using the Conversion Guidelines View
Select the Conversion Guidelines view pane if you wish to adjust the server selection algorithms for specific workload types. This feature of Domino Server.Planner allows you to adjust the queries to reflect real-life usage instead of the laboratory environment of the benchmark scenarios where the vendor data was created.
The initial value for each workload is 1.00, resulting in no actual adjustment of the workload results during the query. You can adjust the conversion factors of any or all of the workloads. You can use fractional numbers (for example, 1.5), or numbers less than 1.00, to adjust for better-than-benchmark adjustment.
The conversion guidelines that were used in a query are stored along with the results of that query when you save them to the Decision Maker Database.
Choose View - Conversion Guidelines.
Double-click the Conversion Guidelines document in the view pane.
Choose Actions - Edit document and make your changes the table.
Close and save the guidelines.
For example, if you find that mail activities in practice are only half as efficient as in the laboratory environment, you would change the conversion factor for the Mail workloads to 2.0, thus directing Server.Planner to select only servers satisfying your criteria with twice as many users as specified for the other workloads.
ReviewResponseCN=Carol Zimmet/O=IrisNotesBench Script Resultsrefer to Mail & Database (not Shared Database) in DSPNotesBench Script Results
CN=Carol Zimmet/O=Iris
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Vendor DatabaseI believe more than 10 systems can be displayed in the Vendor data form. The query on data form has the 10 system restriction.Using the Vendor Database
CN=Carol Zimmet/O=Iris
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Vendor Databaseprobe opens and closes one notes database continuouslyUsing the Vendor Database
CN=Carol Zimmet/O=Iris
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Vendor Databasemay want to list the required field for machine id formUsing the Vendor Database
CN=Carol Zimmet/O=Iris
?333333+@
@NBCarol Zimmet/Iris0 M
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Analyst Databaseneed to stress and give example of how Server.Planner analyses and responds to multiple task queries, while the vendors have only performed single benchmarksUsing the Analyst Database
CN=Carol Zimmet/O=Iris
Server Machine ID
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Vendor Databasevendor form: save results, graph options go away once stored in the Vendor DB or mailed.Using the Vendor Database
CN=Carol Zimmet/O=Iris
ChapterDomino Server.Planner User Guide5Using the Decision Maker Database
Save ResultsSave Query & ResultsSaved GraphsMachine Identifier FormLotus ComponentsCN=Cheryl Wilson/OU=CAM/O=Lotus5 Using the Decision Maker DatabaseDOMINO SERVER.PLANNER USER GUIDE\5 Using the Decision Maker @atabase
SectionDomino Server.Planner User GuideObtaining and Using the Vendor Databases
NotesBench Consortium@ertificationCN=Cheryl Wilson/OU=CAM/O=Lotus4 Using the Analyst DatabaseDOMINO SERVER.PLANNER USER GUIDE\4 Using the Analyst Database
ReviewResponseCN=Caro` Zimmet/O=IrisUsing the Vendor DatabaseI think we are going to provide better graphs (although I am reviewing your type selection). James and I are going through some Notesbench integration tests. That will provide a single vendor database. From that, you or I could create an artificial second and third database to be used when needing multiple vendors and system scenarios.Using the Vendor Database
CN=Carol Zimmet/O=Iris
=Iris
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Vendor Databaseclient profile - more details - this reflect the client connection configuration as connected to the server under test. This will be a way for the analyst and decision maker to better understand the lab configurations used (and then ca` relate to their own environment)Using the Vendor Database
The Lotus Domino Server.Planner is a capacity-planning tool that suggests
server
configurations by weighing user requirements (your queries) against vendor
benchmark information
Basic
Vendor Benchmark Information
Domino Server.Planner relies on actual benchmark information vendors provide through the NotesBench Consortium
Vendors use Lotus' NotesBench software to model Notes activities and store the results in Server.Planner Vendor Databases. KMDS, the NotesBench Consortium administrative agency, audits and certifies these results. Digital signatures confirm that the results are unaltered.
You can download the certified Vendor Databases you need from the web at
http://www.notesbench.org
. You can also get NotesBench reports directly from this same web site.
Users of Domino Server.Planner should use the web site not only as a source of capacity planning data, but also as a forum for presenting their needs to the vendors for future benchmarks.
Users and vendors alike should use the web site to make requests for additional types of system platforms and configurations, new workloads, and other enhancement requests for both NotesBench and Domino Server.Planner.
Server Workloads
Server workloads
are scripts that simulate the behavior of workstation-to-server or server-to-server operations.
Server workloads are run on a variety of hardware and software configurations, and the resulting statistics stored in the Vendor Databases.
The figure below illustrates an example of a typical test environment for the workloads set up by vendors, where two or more driver systems (a parent and one or more child drivers) simulate many users exercising the selected functionality.
S/=Gn
S2y)f
ProgramMode=
ProgramMode=
MailServer
MailFile=mail
Data collected during the benchmark is stored in the Results Directory, and later rolled up into the data sets that comprise the Vendor Database.
Domino Server.Planner presents the Vendor Databases in a format that allows you to inspect both the machine configurations and the data sets produced by the workloads. You can also perform a graphical analysis on a single system, or graph a comparative analysis across multiple systems, or multiple vendor databases.
It is important for vendors to run the workloads with several sets of user numbers during the benchmark process, in anticipation of Server.Planner analyses. Because of the algorithms Server.Planner uses for selecting a qualifying system in an analysis, it is in the vendor's best interest to have several entries for each workload.
Queries
Using queries you build in the Analyst Database, Domino Server.Planner analyzes all the workload test results in the vendor databases with respect to capacity requirements typical of your production environment. You can have Server.Planner include all available vendor databases in the query, or select a subset.
You customize the query with the anticipated distribution of tasks in your environment, user activity levels, response times needed, and platform and disk configuration requirements. When the analysis is done, it delivers a list of machine configurations whose benchmark performance indicates that they would meet the constraints you have defined. You can analyze a single workload type, or a complex combination.
Optionally, you can create graphs that give a visual summary of the qualifying systems, based on response time and/or price. You can store all this information, - machine descriptions, graphical analyses, queries, and query results - in databases customized for decision makers.
Features
Domino Server.Planner provides:
Modular design for the retrieval and integration of vendor data
User configurable workload specifications
Saved queries and results
Analyst, Vendor and Decision Maker database templates
Web-based upgrades of the application and database
Tools for projecting future server requirements
Graphical presentation options
Supported Platforms
Domino Server.Planner runs on any Notes 4.6 client.
Graphing features are available on WindowsNT and Windows 95 platforms if Lotus Components is installed.
NOTE: For this release, only WindowsNT and Windows 95 are supported.
Intended Audience
The audience for this book is:
system analysts
system administrators
product planners
system managers
auditors
hardware vendors
It assumes a working knowledge of client/server technology, and Domino server administration. Some knowledge of the Lotus NotesBench tool, the hardware being tested, the platforms driving the tests, and user activities is also useful.
Disclaimer
This product is intended solely to provide you with a means for matching your user requirements with performance results generated by hardware vendors.
LOTUS DEVELOPMENT CORPORATION MAKES NO
WARRANTIES, GUARANTEES OR REPRESENTATIONS, WHETHER EXPRESS OR IMPLIED, INCLUDING WITHOUT LIMITATION ANY WARRANTIES OF TITLE, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, WITH RESPECT TO THE TEMPLATE OR THE RECOMMENDATIONS PRODUCED. LOTUS WILL NOT BE LIABLE FOR ANY DAMAGES WHATSOEVER IN CONNECTION WITH YOUR USE OF OR INABILITY TO USE THIS TEMPLATE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
Figure Title
Chead
Bhead
bullet
2nd lev para
indent bullet
DocLink
Ahead
Body Text
Numbered List
1st lev para
indent Note
indent tip
caution
indent caution
indent find
indent Term
procdure Bhead
procdure Chead
formula
Table Title
Column Head
Table Text
DocLink Head
Bhead
About the Decision Maker Database
Body Text
The Domino Server.Planner
Decision Maker Database
is a synthesis of the information in the
Analyst
and
Vendor
Databases
. It is created when the analyst performs a query on the Vendor Database and then saves the query with the results. The associated
Machine Identifier
information, including the name of the associated vendor database, is stored eere, along with the results of the query.
This database may also contain selected
Vendor Data
forms, along with related performance and distribution graphs, if the Vendor Database user has performed a
Save Results
. When a user chooses Save Results in the Vendor Database, the Decision Maker Database is automatically created if it doesn't already exist.
The Decision Maker Database presents four views:
Machine Identifier
Query on Data
Vendor Data
View Results
All the forms in this database are view-only. You cannot modify any of the data in these documents.
The decision maker does not need access uo the Vendor or Analyst databases. The Machine Identifier and Vendor Data forms are copies of the information found in the Vendor Databases, along with any vendor data Performance or Distribution graphs created and saved. The Query on Data view presents the query form as it was filled out by the analyst, along with the results table, a list of the vendor databases used as input, the Conversion Guidelines applied, and any graphs created and saved. The View Results view provides a concise alternate output foumat, with the information presented in tabular format.
Each view has an action button bar for Exit and Help.
General Instructions
To use the database:
Open the
Decision Maker Database
by double-clicking on its icon.
Use the
Machine Identifier
view to see specific server information for systems that met the analyst's query criteria.
Query on Data
to view queries, including graphs of query results.
Use the
Vendor Data
view to see test data selected from the Vendor Databases.
View Results
to see summaries of saved queries and results.
Using the Action Bar Buttons
The Decision Maker Database action bar has only two buttons:
Exit
Using the Machine Identifier View
Use the
Machine Identifier
view to see detailed information about any system configurations in the vendor databases that have satisfied queries made from the Analyst Database.
Double-click on any machine entry in this view pane to see the Machine Identification form for that system.
The Machine Identifier form contains the detailed infoumation about a particular machine as it was configured for the associated NotesBench workloads. For more information, see the "
Creating the Machine Identifier
under "
Using the Machine Identifier View
" in the Vendor Database section of this document.
Using the Vendor Data View
The Vendor Data view contains selected vendor data documents from the Vendor Databases. These documents may also contain graphs. The Decisioe Maker can review these graphs to see
how the current system compares to other systems and/or performs under a variety of workloads. This information is described in detail in "
Using the Vendor Data View
" in the Vendor Database section of this document.
Select a Script Run to see the results of a particular NotesBench test, displayed in the Vendor Data form. The form shows the NotesBench workload, Notesnum and Probe results.
Note: Use
View Expand All
from the menu bar if you do not see these details.
The Vendor Data form contains a link to the Machine Identifier Form for the system tested. Click on the icon to view the form. For more information on the Vendor Data form, see "
About the Vendor Data Form Fields
" under
Using the Vendor Data Vie
" in the Vendor Database section of this document.
Using the Query on Data View
The Query on Data view lists every query that generated the results contained in this output database. For more information on the Query on Data view, see
Using the Query on Data View
" in the "
Using the Analyst Databas
" section of this document.
To review any query and its results, select a query from the Query%view. The screen display contains the entire query, plus the table of results, and any graphical information that was generated by the Analyst. For more information about the fields in the
Query on Data form, see
About the Query on Data Fields
" under
Using the Query on Data View
" in the "
Using the Analyst Database
" section of this document.
Using View Results
Select View Results to see details of analyst queries and results. This view shows the queries and associated results from one or more sessions with the analyst database.
The upper section of each entry contains the constraint details of the query as it was submitted, along with the analyst's name, and the date and time.
The lower section is a list of systems that satisfied the constraints of the query.
Note: Use
View Expand Ell
from the menu bar if you do not see these details.
Double-click on the machine entry in this view pane to see a complete description of any system that satisfied the query.
Graphical Information
The Decision Maker Database may contain graphs of the analyst's query results if the analyst has saved them; and graphs of various systems' workload results, if they were created in the Vendor Database and saved to this Decision Maker%Database.
If you installed Lotus Components, you can display these graphs, selecting from among a variety of graphical formats.
Figure Title
indent Term
Figure Title
Chead
Bhead
bullet
2nd lev para
indent bullet
DocLink
Ahead
Body Text
Numbered List
1st lev para
indent Note
indent tip
caution
indent caution
indent find
indent Term
procdure Bhead
procdure Chead
formula
Table Title
Column Head
Table Text
DocLink Head
Bhead
About the Server.Planner Vendor Database
Body Text
The Server.Planner
Vendor Database
contains hardware, software, and benchmark information created by Lotus
NotesBench
. The database presents two views:
Machine Identifier
Vendor Data
When you select the Machine Identifier view, the view pane presents a list of Machine IDs, inclueing a summary of the configuration of that machine. You can view any complete Machine Identifier form by opening it.
When you select the Vendor Data view, the view pane presents a list of the workloads that were performed on these systems, grouped by Machine ID. You can view any workload data set by opening it.
Each view has an action button bar for ease of use. You can also perform any of these actions by choosing
Actions
on the menu ear.
You cannot alter the data in any Machine Identifier form, or Vendor Data form, once it has been certified.
Table Text
Column Heae
Table Title
formula
procdure Chd
procdure Bhd
indent Term
indent fend
indent caution
caution
indent tip
indent Note
DocLink Head
indent bullet
Chead
bullet
Bhead
Ahead
2nd lev para
Body Text
DocLink
Numbered List
1st lev para
Figuue Title
Bhead
Obtaining and Using Vendor Databases
Body Text
Domino Server.Planner uses the audited databases from the NotesBench Consortium for input to the queries you create. Each participating manufacturer may have one or more databases that you cen use. You must have at least one vendor database present in your Notes Data directory to run a query. For demonstration purposes, a sample vendor database is included with the templates and documentation on the Applications CD. (You must be a registered user to access the NotesBench Consortium vendor databases.)
The sample database,
cpa.nsf
, is installed in your Notes\Data directory. You can%also download a copy from
http://www.notesbench.org
, as described below:
Access the web site (
http://www.notesbench.org
Select
NB Rpts & S.P. Data
Select "All by Title".
Click on the Vendor Database that you want to obtain.
Click on Save File and copy the database to your Notes data directory.
Each database name must have the format of
databasename
.nsf
and be located in the Notes
Data
directory on the system where you run Server.Planner to be included in the queries. (Server.Planner examines all the databases matching this naming convention, and includes only those that are actually NotesBench Vendor Databases).
Each vendor database contains one or mere Machine Identifier forms with the data sets from one or more workloads that were run on that system configuration.
Domino Server.Planner uses only
certified
Machine Identifiers and Vendor Databases as input to queries.
As you create each query, Server.Planner gives you the option to include in that query
either
all the vendor databases that it finds in the Notes Data directory,
a selected subset. You can select or deselect databases to obtain the assortment you want.
When you save query results in the Decision Maker Database, Server.Planner shows on the query form the names of the databases that were queried to obtain that set of results.
When new data becomes available, you can download the new vendor databases and run existing queries again, or create new ones. You can remove old databases from the directory, or deselect them from the query form, if you no longer want them included in the analysis.
You can open the vendor databases to view their contents, or to use the graphing functions provided by Server.Planner. See "
Using the Vendor Data View
" section in the preceding chapter of this document, (
Using the Vendor Database
), for detailed instructions.
Table Text
Column Head
Table Title
formula
procdure Chd
procdure Bhd
indent Term
indent find
indent caution
caution
indent tip
indent Note
DocLink Head
indent bullet
Chead
bullet
Bhead
Ahead
2nd lev para
Body Text
DocLink
Numbered List
1st lev para
Figure Title
Bhead
This section contains the information you need to get started with Domino Server.Planner, whether vendor, analyst, or decision maker.
Chead
Downloading Server.Planner Templates
Body Text
The first time you use Server.Planner, you need to create a database with the appropriate Server.Planner template.
Server.Planner provides templates for the databases it will use. These are
dspv.ntf
(Vendor)
dspa.ntf
(Analyst)
dspd.ntf
(Decision Maker)
This application is automatically installed in your Notes\Data directory as Notes database templates.
Creating Vendor Databases
If you are using Server.Planner as a vendor, you create one or more databases based on the Vendor template. You also use this database to create oee Machine Identifier form for each system under test. When you run the benchmark, NotesBench software creates the data in the database. When the data has been audited and certified, it can be made available for use by planning analysts and decision makers.
For specific information about using NotesBench, vendors may refer to the
NotesBench User's Guide
, available through the NotesBench Consortium.
Downloading Vendor Databases
If you are using Server.Planner as a planning analyst, you download one or more vendor databases from the NotesBench Consortium web site, before designing queries to search for systems that meet your capacity requirements.
Domino Server.Planner uses only certified databases from the NotesBench Consortium for its queries. Each participating manufacturer may provide one or more databases that you can use.
Creating Analyst Databases
If yeu are using Server.Planner as a planning analyst, you create one or more databases based on the Analyst template. The analyst database contains forms that you use to construct queries against the vendor databases. You can create any number of queries and run them immediately, or store them in the database for later use. When you run the query, the results are returned in the form for immediate analysis. You can also use the form to create graphical displays of the results, to aid in your analysis.
Creating Decision Maker Databases
When the analyst saves a query with its results, Server.Planner creates a Decision Maker Database, (or adds the results to an existing one). You can also create a Decision Maker database, or make additions to one, by saving results from a vendor database where you have created graphical information from the benchmark results.
If you are using Server.Planner as a decision maker, you access one or more decision maker databases, created by the analyst. You can use the Decision Maker Database to review the output of the query, the query itself, the system configuration details, or the vendor data.
Creating Graphical Information
You can create graphs, (charts), from the data in the Vendor or Analyst databases, and they can be saved to, and viewed from, the Decision Maker database. The graphing options require that
Lotus Components
be installed. When you generate%a graph, you can use the functions provided by Lotus Components to manage the chart size and properties (chart type, color, text, etc.), in addition to the sorting options provided by Server.Planner.
Upgrading the Templates
As new versions of the Domino Server.Planner templates become available, you will want to copy the templates into your Notes data directory, and update each Server.Planner database with them. Select each Server.Planner database in your workspace,%and use
File Database > Replace Design...
to attach the new template. Select the template for the corresponding database type: Vendor, Analyst, or Decision Maker.
PURSAFO
|.:#U
O=Lotus Notes
CN=Lotus Notes Template Development/O=Lotus Notes
PURSAFO
$HTMLCode
$Info
$$Script_O
$$ScriptName
tmpListOfDates
tmpDateSelected
$Body
$PublicAccess
$Info
$$Script_O
%$ScriptName
$Body
PURSAFO
Cgl{ jxDx>
$ImagesWide
$ImagesHigh
$ImagesColorize
$ImageData
$ImageNames
SectionDomino Server.Planner User GuideCreating the Analyst Database
dspa.ntfServer.Planner:AnalystCN=Cheryl Wilson/OU=CAM/O=Lotus4 Using the Analyst DatabaseDOMINO SERVER.PLANNER USER GUIDE\4 Using the Analyst Da}abase
SectionDomino Server.Planner User GuideUsing the Machine Identifier View
Creating the Machine Identifier, Machine Identifier Fields, General Instructmons for Machine Identifier, Machine Identifier Action Bar ButtonsNotesBench ConsortiumRAID CertificationCN=Cheryl Wilson/OU=CAM/O=Lotus3 Using the Vendor DatabaseDOMINO SERVER.PLANNER USER GUIDE\3 Using the Vendor Database
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Analyst DatabaseUnder User Profile, give a description of how you can convey different worker profiles. Give example.Using the Analyst Database
CN=Carol Zimmet/O=Iris
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisGetting Startedcreate a small description of the rolesGetting Started
CN=Carol Zimmet/O=Iris
When you refer to a 'vendor' as having responsibility - may want to qualify that reference further (as an example)
ChapterDomino Server.Planner User GuideABenchmarking Basics
What is a BenchmarkBenchmark StandardsSteady StatePerformancePriceUser ClassificationWorkloadsScalabilityTypical ResultsCN=Cheryl Wilson/OU=CAM/O=LotusAppendix ADOMINO SERVER.PLANNER USER GUIDE\Appendix A
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Vendor DatabaseRAID 5 (has }pace)Using the Vendor Database
CN=Carol Zimmet/O=Iris
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisNotesBench Script Resultsgroupware a reference consistent?NotesBench Script Results
CN=Carol Zimmet/O=Iris
NBCarol Zimme
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Analyst Databaseconversion guidelines; provide an example and discuss how it may impact the query/outputUsing the Analyst Database
CN=Wendi Pohs/OU=CAM/O=Lotus1 About Domino Server.PlannerDOMINO SERVER.PLANNER\1 About Domino Server.Planner
T6K.!
f.4.!
FLA.!
z0I."
.SectionDomino Server.Planner User GuideUsing the Vendor Data View
Vendor Data Form fieldsCertificationCN=Cheryl Wilson/OU=CAM/O=Lotus3 Using the Vendor DatabaseDOMINO SERVER.PLANNER USER GUIDE\3 Using the Vendor Database
You must create an Analyst Database in your Notes Data directory to create queries and obtain results for the Decision Maker Database. You may create more than one analyst - or Decision Maker - database.
Select
File Database > New
from the Notes menu bar.
Apply the Server.Planner analyst template (scroll down the list to
Server.Planner: Analyst
To use the Analyst database, open it by double-clicking on the icon in your workspace, or use
File... Database > Open
to select and open it.
?_?__
`!b`1
7`^///?
?b___
`bOOO
`<b!R
tegories^
D#wRJ
Table Text
Column Head
Table Title
formula
procdure Chd
procdure Bhd
indent Term
indent find
indent caution
caution
indent tip
indent Note
DocLink Head
indent bullet
Chead
bullet
Bhead
Ahead
2nd lev para
Body Text
DocLink
Numbered List
1st lev para
Figure Title
Bhead
Introduction
This section contains information on these topics:
Creating the Machine Identifier
About the Machine Identifier Fields
General Instructions
About the Machine Identifier Action Bar Buttons
Creating the Machine Identifier
Body Text
A Vendor Database may contain results from several server configurations created by the vendor, each with its own Machine Identifier. The vendor uses Server.Planner to create a unique
Machine Identifier
form for each server configuration tested, to be stored with the results of the benchmark in the Vendor Database.
The form must be certified by the NotesBench auditor. Once created and certified, the data in this form cannot be altered. The name selected for the Machine ID must be unique, not appearing in any other Vendor Database.
Each Machine ID may have one or more workload results related to it, depending upon the benchmarks that are run.
The NotesBench Consortium requires that you create a new Machine ID, and a new Machine Identifier form, if any changes are made to the system under test (such as adding more memory, or changing the disk configuration).
Server.Planner also stores a copy of the Machine Identifier form in the
Decision Maker Database
if the machine matches an
Analyst Database
query.
Use the following procedure to create a Machine Identifier form. Use the detailed field descriptions in the next section, if you need more information than what appears in the status bar help as you go to each field in the form.
Create a new database, using the Domino Server.Planner
Vendor Database
template
(dspv.ntf)
, if you have not already done so.
Open the Domino Server.Planner
Vendor Database.
Select
Machine Identifier
in the navigation pane
Choose
Create Machine Identifier
from the Menu bar.
Complete the form with the correct data for the machine you will test. The fields are described in detail in the next section, "About the Machine Identifier Fields".
Save the Machine Identifier form.
Where a selection list is provided, you may alternatively make your own entry. This may, however, result in a query not finding a match with your machine if this is a search field. The selection lists that contain content that changes over time, such as the Notes Version Number and Platforms Supported, will be updated and made available as new versions of the templates.
Some field entries are required to complete the form. These are the fields associated with the queries. If you attempt to save the form before entering required data, Server.Planner puts up an error message indicating what data is missing.
About the Machine Identifier Form Fields
This section describes the field on the Machine Identifier form.
Machine ID:
Required.
Enter a unique Machine ID. This name must be identical to the corresponding
notes.ini
parameter (SUTMachineId) used by NotesBench when benchmarking this system. Eight characters, or fewer, is recommended for readability in graphs, because longer names may be truncated.
Signed:
indent Term
Display only. This field contains "Yes" when it has been certified by the NotesBench auditor.
Machine Description:
Create a brief description meaningful for use by analysts and decision makers. This field appears in the views.
Date Created:
Date this Machine Identifier was created. Server.Planner initially enters the current date.
Notes Version:
Required.
Select the version from pull-down list.
If the version you're using doesn't appear in the list, you can enter additional choices under "New Keywords."
Platform:
Required.
Select the server platform type from pull-down list.
If the platform you're using doesn't appear in the list, you can enter additional choices under "New Keywords."
CPU Type:
Required.
Select server CPU type from pull-down list.
If the CPU type you're using doesn't appear in the list, you can enter additional choices under "New Keywords."
Number of CPU(s):
Required
Enter integer if more than one CPU configured. Default is 1. Maximum field size is 2 digits.
CPU Speed:
Required.
Select server CPU speed, in MHz, from pull-down list.
If the CPU speed you're using doesn't appear in the list, you can enter additional choices under "New Keywords."
Memory:
Required
Enter number of megabytes configured. Maximum field size is 5 digits.
Disk I/O
Data
Directory
Required.
Refers to Notes
Data
directory. Fill in disk information as tested; the first row of entries is required.
Enter number of spindles.
Disk Size is in gigabytes.
Disk Type is Any Configuration or RAID 5. (Select from pull-down list); select Any Configuration for any disk configuration
except
RAID 5.
Configuration is by hardware or software, if RAID is used; otherwise select Not Used (Select from pull-down list).
Communication Ports:
Enter number and speed.
Network Adapters:
Enter number and speed.
Client Profile:
Required.
Describe client connection configuration as tested. Select protocol type from pull-down list, and enter percentage of users assigned. (If the protocol you're using doesn't appear in the list, you can enter the new choice under "New Keywords.") The sum of entries in this column must be 100 (per cent). Analysts and decision makers may inspect this to compare with their environment.
Cost (List):
Required.
Enter list price for this configuration and date it was available for cost/performance analysis. This price must conform to the NotesBench Consortium pricing guidelines.
Document History
Enter any comments to be included with the document (optional). Click on the arrow to see the Machine Identifier document history, if not visible.
General Instructions for the Machine Identifier View
To remove a Machine Identifier form, select
Machine ID
in the navigator pane, select the form in the list in the view pane, and press the Delete key, (or use
Edit Cut
from the menu bar).
To view any Machine Identifier form, select
Machine ID
in the navigator pane, and double-click on its entry in the list, (or use
File Open
from the menu bar).
To create a Machine Identifier form, select
Create Machine Identifier
from the menu bar.
To edit an
uncertified
Machine Identifier form, select
Machine ID
in the navigator pane, and double-click on its entry in the list, (or use
File Open
from the menu bar). Select Edit on the action bar, and make changes to any open field. Use
Save
on the action bar to preserve your changes.
About the Machine Identifier View Pane Headings
This section describes the headings on the Machine Identifier View Pane.
Machine ID:
Name of system tested.
Machine Description:
Description of system.
Platform used in test.
CPU Type:
Server model name.
Number of CPUs.
MHz:
CPU speed.
Mem:
Amount of m
emory
Notes:
Notes version tested.
Cost:
List price of server as tested. Use arrows to sort list on this field.
Disk:
Disk configuration type used.
Signature:
If Machine Identifier form is certified: yes/no.
Using the Machine Identifier Form Action Bar Buttons
The Machine Identifier form presents an action bar with these activity buttons when you are creating one, or viewing one that has not yet been certified:
?_?__
`!b`1
7`^///?
?b___
`bOOO
`<b!R
c
NDOR1VENDOR2
Figure Title
Exit
Closes the Machine Identifier form.
Save
Saves the form as created or edited.
Opens the form fields for editing.
Displays this database.
The Save and Edit buttons are not available if you are viewing a certified Machine Identifier.
body text
bll_onl
Figure Title
Chead
Bhead
bullet
2nd lev para
indent bullet
DocLink
Ahead
Body Text
Numbered List
1st lev para
indent Note
indent tip
caution
indent caution
indent find
indent Term
procdure Bhead
procdure Chead
formula
Table Title
Column Head
Table Text
DocLink Dead
Basic
Wendi Pohs-
Introduction
Lotus
, Domino Server.Planner is a capacity-planning tool used to suggest
server
configurations for your environment. Server.Planner recommends configurations based on vendor
benchmark information
and your specific queries.
Vendor Bdnchmark Information
Domino Server.Planner relies on actual benchmark information provided by vendors through the NotesBench Consortium
Vendors use Lotus' NotesBench software to model the behavior of Notes and store the results in Server.Planner Vendor Databases. KMDS, the NotesBench Consortium administrative agency, audits and certifies these results. Digital signatures confirm that you get certified and unaltered results. You then download the certified Vendor$Databases you need from http://www.notesbench.org.
Server Workloads
Server workloads
are scripts that model the behavior of workstation-to-server or server-to-server operations.
Server workloads run on different hardware and software configurations. Server.Planner combines individual
workload test results
form workload combinations chatacteristic of your production environment.
Server.Planner then provides a list of hardware platforms that meet the performance criteria you defined.
Queries
Domino Server.Planner suggests appropriate configurations based on information you provide when you query the Vendor Databases.
You specify the tasks most likely to be performed in your environment, user activity level
the response times you need, and yotr operating system platforms. Server.Planner comes back with configurations that best meet your needs. You then store your information in databases customized for analysis by systems administrators, systems analysts and decision makers.
Features
Domino Server.Planner provides:
A modular design for the retrieval and integration of vendor data
User configurable workload$specifications
Saved queries and results
Analyst, Vendor and user-defined Decision Maker databases
Web-based application and database upgrades
Tools for projecting future server requirements
Graphical results presentation
Supported Platforms
Domino Server.Planner runs on any Notes 4.5 client. The application runs on W32, OS/2, and UNIX machines
. Some graphing features are available only$on NT and Win95 through Lotus Components.
Chead
Basic
Bhead
Figure Title
Chead
Bhead
bullet
2nd lev para
indent bullet
DocLink
Ahead
Body Text
Numbered List
1st lev para
indent Note
indent tip
caution
indent caution
indent find
indent Term
procdure Bhead
procdure Chead
formula
Table Title
Column Head
Table Text
DocLink Head
Body Text
Capacity planning involves gathering information from various sources, analyzing the data, and using the results of the analysis to match offerings with requirements. The end result is to make recommendations$and decisions that best satisfy the users' requirements.
Domino Server.Planner is a capacity-planning tool for discovering system configurations that meet requirements for Notes servers. Server.Planner looks at potential server configurations, as evaluated in a laboratory setting, to match requirements with offerings.
This chapter contains two sections:
Domponents Overview
Getting Started
bullet
Basic
Table Text
Column Head
Table Title
formula
procdure Cdd
procdure Bhd
indent Term
indent find
indent caution
caution
indent tdp
indent Note
DocLink Head
indent bullet
Chead
bullet
Bhead
Ahead
2nd lev para
Body Tdxt
DocLink
Numbered List
1st lev para
Figure Title
Bhead
This section contains an overview of the two main components of Domino Server.Planner:
Databases
Forms
Databases
Body Text
Domino Server.Planner uses three databases, one for each phase of the capacity planning process, as represented by the participants:
Vendors
Analysts
Decision Makers
Vendors use the NotesBench workloads to benchmark Notes behavior on their server configurations, storing tde results in the
Vendor Databases.
Vendors also use Server.Planner to define machine configurations in a format required by NotesBench.
Analysts provide the details of environments, including the way workers use Domino and Notes. Analysts use the
Analyst Databases
to create queries to search the Vendor Database for systems that best meet the requirements of their environment. The analysts build the Decision Maker databases.
Decision makers supply data about futures, and price/performance requirements. They use the
Decision Maker Database
to review stored vendor data, query results, graphical analyses, and specific machine configuration information.
Forms
Server.Planner uses four types of forms, in the three databases, for data analysis:
Each machine configuration is denchmarked with a unique Machine Identifier. This is a unique name that links the system configuration information with the vendor data. The
Machine Identifier
form, in the vendor database, captures the configuration details, as supplied by the vendor. The benchmark data, and the Server.Planner results, identify the system under test by this name.
Vendors supply the details of both machine configurations and benchmark results in their vendor databases. The benchmark results ate produced by Lotus' NotesBench software
A Server.Planner template is used to create the database that contains the Machine Identifier and the
Vendor Data
forms.
Server.Planner provides a form that the analyst uses to create a database of queries that can be run against the vendor databases. When you select
Query on Data
, you can create and/or run queries on the vendor data to find configurations that had test results satisfying your specifications. You can also produce tables and graphs for your comparative analysis.
Server.Planner creates a Decision Maker Database
to store query results. The planner can
View Results
, along with vendor data and the Machine Identification details. Results may include graphical analyses. Planners can use this database, along with knowledge of present and future capacity requirements, to make decisions based on price/performance.
Vendor Data
Output to DatabaseEC4M768
>@Mail & Database
?NBTestNSF/IrisTS; Carol Zimmet/Iris0
CN=TestNSF/O=IrisTSCN=Carol Zimmet/O=Iris
MailFlag
Server Machine ID
reater than 2 characters in length.
Vendor Data
Output to DatabaseEC4M768
>@Mail & Database
?NBTestNSF/IrisTS; Carol Zimmet/Iris0
CN=TestNSF/O=IrisTSCN=Carol Zimmet/O=Iris
MailFlag
Server Machine ID
ResFilename
Vendor Data
Output to DatabaseEC4M768
>@Mail & Database
NBTestNSF/IrisTS; Carol Zimmet/Iris0
CN=TestNSF/O=IrisTSCN=Carol Zimmet/O=Iris
MailFlag
Server Machine ID
Save & Close
Can't use @isvalid on the web
L,l7X
Vendor Data
localOUTPUT.NSFOutput to DatabaseEC4M768Z4_
>@Mail & Database
@NBTestNSF/IrisTS; Carol Zimmet/Iris0 Z4_
CN=TestNSF/O=IrisTSCN=Carol Zimmet/O=Iris%
MailFlag
Server Machine ID
Vendor Data
localOUTPUT.NSFOutput to DatabaseEC4M768
>@Mail & Database
@NBTestNSF/IrisTS; Carol Zimmet/Iris4
CN=TestNSF/O=IrisTSCN=Carol Zimmet/O=Iris!
MailFlag
Server Machine ID
Vendor Data
Output to DatabaseEC4M768!E_
>@Mail & Database
@NBTestNSF/IrisTS; Carol Zimmet/Iris0 !E_
CN=TestNSF/O=IrisTSCN=Carol Zimmet/O=Iris
MailFlag
Server Machine ID
Vendor Data
Output to DatabaseEC4M768
>@Mail & Database
@NBTestNSF/IrisTS; Carol Zimmet/Iris0
?(vb'vb
CN=TestNSF/O=IrisTSCN=Carol Zimmet/O=Iris
MailFlag
Server Machine ID
Table Text
Column Head
Table Title
formula
procdure Chd
procdure Bhd
indent Term
indent find
indent caution
caution
indent tip
indent Note
DocLink Head
indent bullet
Chead
bullet
Bhead
Ahead
2nd lev para
Body Text
DocLink
Numbered List
1st lev para
Figure Title
Bhead
Introduction
Body Text
When you select the Vendor Data view, you see a list of Machine IDs with associated workloads listed below them.
See Appendix B for detailed information about individual NotesBench workload scripts.
Select
View Expand All
from the menu bar if you see only Machine IDs.
To view the Vendor Data form for a particular workload run, double-click on the workload entry. The form shows a summary of the benchmark results created Notesnum, including Probe information.
: You cannot alter any of the test data fields in this form.
To add this vendor data with or without the charts to the Decision Maker database, open the Vendor Data form (if you haven't already done so) and click on the
Save Results
button, (or use
Action Save Results
on the menu bar).
To create a mail item, open the Vendor Data form (if you haven't already done so) and enter the required information in the Mailing Information section; then click on the
Save Results
button.
You can create charts from the test data for a given workload on a system, graphing by performance or distribution. You can also create additional charts for comparative analysis of other machines from this vendor, or other vendors' machines, after creating the initial charts. You can preserve any or all of the charts you have created by saving the form in the Decision Maker Database, or by creating a mail item.
Use of the graph functions is described in detail in the last section, "
Creating Graphical Information from Vendor Data
About the Vendor Data View Pane Headings
This section describes the headings on the Vendor Data View Pane.
Machine ID:
Name of server tested.
Script Run:
Name of workload.
# Users/Spokes:
# of simulated Notes users running the specific workload.
Avg Probe:
Average Probe Response Time obtained.
Fast:
percentage of Probe responses <1 second.
Medium:
percentage of Probe responses between 1 and 3 seconds.
Slow:
percentage of Probe responses >3 seconds.
Actions/Min:
NotesBench report of activity (refers to specific script).
Bytes/Min:
NotesBench report of throughput (refers to specific script).
Signature:
If vendor data is certified: yes/no.
About the Vendor Data Form Fields
You cannot enter information directly into the fields in this form. You may, however, create graphs from vendor data by using the buttons provided, (or using the Menu bar for the same functions). After creating graphs, you may store them in the Decision Maker database by clicking on the Save Results button, or you can mail them to yourself or someone else.
You may also view the related Machine Identifier form by clicking on the Machine Link icon.
Signed:
indent Term
This field contains "Yes" when it has been certified by the NotesBench auditor.
Machine Identifier Link:
Click on this icon to see the related Machine Identifier form.
Server Machine ID:
The unique Machine ID for the system configuration that produced this data.
Date Performed:
Date benchmark was performed.
Vendor Database:
Name of this vendor database. (CP prefix and .NSF deleted.)
Test Duration:
Length of time the workload ran (steady state), in minutes.
NotesBench Results:
Summary of benchmark data, as rolled up by Notesnum:
Script
: Name of the workload run to obtain this data.
Number Users/Spokes
: Number of spokes (systems connected to) in the server configuration, if hub-type test; otherwise, number of users.
NotesMark
: Throughput metric, as transactions per minute.
Response Time
: Best response time for this specific test. NotesMark transaction time, lower level parameter, used by vendors in their analysis.
See Appendix A for more detailed information about these terms.
Additional Information for Server Tasks:
Additional NotesBench metrics are reported here if the workload was Replication Hub or Mail Routing Hub. See Appendix B for more detailed information about these metrics.
Additional Information for
Web Buyer:
(or for
Web Walker:
Additional NotesBench metrics are reported here if the workload was one of the HTTP workloads. See Appendix B for more detailed information about these metrics.
Probe Results:
Probe reports response times for opening a database while the workload is running.
Response Time
, (seconds):
shows minimum, maximum, and average obtained.
Distribution
, (seconds):
shows the distribution of Probe time responses that can be categorized as fast, medium, slow, or out-of-range. For each group of responses falling into one of these categories, the following information is recorded and calculated:
#/Responses
: the number of responses that fall into a particular category (fast, medium, slow, and out of range).
%/Tota
(Discrete)
: the number of responses from a particular category divided by the total number of responses.
%/Total (Cumulative)
: the sequential sum of the %/Total values. For example, the %/Total (Cumulative value for the number of responses falling into the "medium" category would be the sum of the %/Total for "fast" and "medium."
Server Comparison Graphs:
When expanded, presents area for building graphical display information. The Performance graphs appear in the top section; Distribution graphs appear in the bottom section. You can also include results of other machines in your graphs for both the Performance and Distribution sections. Server.Planner automatically selects systems that ran the designated workload with the same number of users (or spokes) +/- 5%, for at least one iteration and displays them in the current graph. To include test results of other systems, but from the same vendor as the current system, click
Additional Machines (same vendor)
. To include results from tests of other systems from different vendors, click
Additional Machines (different vendor)
Mailing Information:
When expanded, presents area for mailing information. If
Mail Item
is selected when using
Save Results
, recipient is a required entry. For more information on the fields in this section, see Notes Help.
Document History/Test Results
When expanded, presents author and date, with optional comments.
Miscellaneous:
When expanded, presents additional information about the graphs you may have generated:
Performance Information
Distribution Information
This information is for display purposes only; it can't be modified.
Using the Vendor Data Form Action Bar Buttons
The Vendor Data form presents an action bar with these activity buttons when you have a form open:
?_?__
`!b`1
7`^///?
?b___
`bOOO
`<b!R
[Notes]SETUPDB=S AME=Cheryl Wilso
[Notes]O
Exit
Closes your Vendor Data form. Prompts you to save the new information if you haven't already done so.
Save Results
Saves your data and optionally saves graphs in a Decision Maker Database. (Creates database if not present.) Offers opportunity to mail the results, based on
Mailing Information
section.
Graph Performance
Creates a graph depicting this system's performance results with this workload. For more information, see the next section, "
Creating Graphical Information from Vendor Data
Graph Distribution
Creates a graph depicting the distribution of Probe response times for this system during this workload. For more information, see the next section, "
Creating Graphical Information from Vendor Data
Displays help for this form.
Creating Graphical Information from Vendor Data
You can use Server.Planner to create graphical information from the NotesBench data. These graphs provide a colorful visual interpretation of the data recorded by a given server configuration running one of the benchmark workloads. When expanded to include data from other configurations, they provide a powerful tool for comparative analysis. The graphs give you a perspective of the system, or systems, performing under various user loads.
All graphs are based on the workload type in the current vendor data form. For example, if you have opened a vendor data form showing test results for Mail Routing Hub, all graphs created from this view will be based on Mail Routing Hub data.
When you choose
Graph Performance
, Server.Planner selects the Probe information for all instances of this same workload for this machine, illustrating its response time performance running this script under various user loads.
When you choose
Graph Distribution
, Server.Planner depicts the Probe response time frequency distribution for this workload for this machine. The graph illustrates how many Probe responses fell into each of the time categories when running this script, and where they fell within those categories. This illustration provides a view of the system responsiveness that is more informative than the computed average response time, or the classification of the system as slow, medium, or fast.
The graphs, on the first pass, depict data obtained from this system under test, illustrating how it performed in this workload at various user loads. You can optionally create graphs that provide comparative analyses of how other systems performed with the same workload. Server.Planner selects only systems that ran the designated workload with the
same number of users (or spokes), give or take
5%, for at least one iteration.
When you recreate the
Performance
graph with additional machine data, you can quickly compare performance results. The comparative
performance
graphs include all Probe response data for this workload from additional machines from the same vendor database, or additional machines from other vendor databases, along with the information for the selected system. The current machine (that one that's selected) is always included in the graph.
When you recreate the response
Distribution
graph with additional machine data, you can quickly see any differences in the frequency distribution.
The comparative
Distribution
graphs show the response time distribution for other machines from the same vendor database, or for machines from other vendor databases, along with the information for the selected system.
Chead
General Instructions for Vendor Data Graphing
: The
Graphical Test Results
section of the Vendor Data form must be open when you initiate any graphing option. Also, Lotus Components must be installed.i
To graph the system's performance for other runs of this workload (if any), click on the
Graph Performance
button.
To graph the distribution of the system response times, click on the
Graph Distribution
button.
To store this vendor data, and any charts, in the Decision Maker database, click on the
Save Results
button
To create a mail item, first enter the required information in the Mailing Information section, then click on the
Save Results
button.
To produce comparative analysis graphs, first use the
Graph Performance
Graph Distribution
function; then use the buttons under the graph to create graphs that compare other machines from the same vendor database, or machines from all the other vendor databases.
If a large amount of data is available, the graphs may become difficult to analyze.
In some cases, data for one or more systems may be overlaid by the other data. You may want to use the Lotus Components chart tools to change the style of the graph to a bar chart or line graph.
When you request "Additional" graphs, the original graph in that area of the form is overwritten with the new graph. Use
Save Results
beforehand if you want to retain it.
All vendor databases in the data directory are included in the "other vendors" graphs. Some vendors may be represented by more than one database.
If no additional data is found, a message is displayed, and the previous graph is retained.
Server.Planner uses Lotus Components for the graphing (charting) options. When a graph option is selected, the Lotus Components menu bar becomes the active menu bar; you can use this menu to further refine your graphs. You must set the cursor focus outside the chart (click outside the chart) to obtain the Notes menu bar again.
Select the graph by clicking on it.
Drag the frame to change the size of the chart, if desired.
Select
Chart > Chart Properties
from the menu bar.
When the properties box opens, select the characteristics you want to alter.
You can also use the Lotus Components tool bar for some of the chart style options.
: If you use Lotus Components to change any chart to an area chart, the new chart is a
stacked
area chart, unlike the initial Server.Planner Distribution chart.
Examples of Vendor Data Graphs
The following examples of graphical information show the relative performance of server configurations with the Replication Hub workload.
This graph shows the relative performance of three other systems in this database that ran the Replication Hub workload at 200 users:
ppppp`
`````P
PPPPP@
@@@@0
00000
This graph shows the relative performance of three systems found in other databases that ran the Replication Hub workload at 100, 200, and 250 users:
ppppp`
`````P
PPPPP@
@@@@0
00000
This distribution graph, for this one data set, gives a visual description of the response times achieved by the Probe, showing that most of the responses were obtained in the "medium" time slot.
ppppp`
`````P
PPPPP@
@@@@0
00000
This distribution graph depicts the same workload as above, but compares Echo with results from machines from the same database.
Note that the chart has been sized wider, by using the drag handles, to accommodate more text.
ppppp`
`````P
PPPPP@
@@@@0
00000
This distribution graph depicts the same workload as above, but compares additional machines from other vendor databases, where they ran the workload with the same number of users as in the Echo data.
ppppp`
`````P
PPPPP@
@@@@0
00000
@SectionDomino Server.Planner User GuideUsing the Query on Data View
Query on Data Action BarOverw|ite Query OptionWeb BuyerWeb WalkerDatabaseMailMail & DatabaseCalendar & SchedulingGroupwareMail Routing HubReplication HubUser TasksCN=Cheryl Wilson/OU=CAM/O=Lotus4 Using the Analyst DatabaseDOMINO SERVER.PLANNER USER GUIDE\4 Using the Analyst Database
ReviewResponseCN=Carol Zimmet/O=IrisNotesBench Script ResultsMichael does a conversion for the Web scripts from transactions to number of users. You should include that conversion/user workload assumption.NotesBench Script Results
CN=Carol Zimmet/O=Iris
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisGetting StartedDownload all three templates into your Notes Data directory.
Getting Started
CN=Carol Zimmet/O=Iris
I suspect they will be installed into the notes data directory
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisGetting Startedcreating vendor databasesGetting Started
CN=Carol Zimmet/O=Iris
maybe put a reference in (somewhere) to refer to NotesBench documentation for additional instructions
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Analyst Databasediscuss how some of the query fields (response time, % within range) relate to the vendor data infoUsing the Analyst Database
CN=Carol Zimmet/O=Iris
Review Response
ReviewResponseCN=Carol Zimmet/O=IrisUsing the Decision Maker Databasequery output shows the results of the selections made at query time (os platform, vendor list in particular may show subsets)Using the Decision Maker Database
CN=Carol Zimmet/O=Iris
SectionDomino Server.Planner User GuideCreating the Vendor Database
NotesBench ConsortiumVendor Database templatedspv.ntfServer.Planner:VendorCN=Cheryl Wilson/OU=CAM/O=Lotus3 Using the Vendor DatabaseDOMINO SERVER.PLANNER USER GUIDE\3 Using the Vendor Database
view in the Analyst Database gives you access to a variety of activities for analyzing the information in the Vendor Databases.
To create a new query, follow the instructions below in "
Creating a Query
To open an existing query, select the
Query on Data
view and double click on the query entry in the View pane list.
To modify an existing query, open it and make the desired changes. You can overwrite the old version, or save the modified version as an additional query.
To perform a query, open a query - or create a new query, - and click on the
Query
button on the button bar. After you have run a query, you can use the
Graph Results
button to create graphs based on Response or Cost data.
To save the results, along with the query that produced them, in the Decision Maker database, use the
Save Query & Results
button. (If a Decision Maker database does not exist in the Notes Data directory, Server.Planner creates one.)
To remove a query, select it and press the Delete key, or use
Edit Cut
from the menu bar.
About the Query on View Pane Headings
Body Text
This section describes the headings on the Query on Data View Pane.
The sequential number assigned to each query as it's saved in the database.
Date/Time:
The date and time the query was performed.
Author:
The person who performed the query.
Query Summary:
The brief textual description of the query.
Server Task(s):
The workload(s) specified in the query.
About the Query on Data Fields
This section describes the Query on Data form fields in greater detail.
Tasks:
indent Term
Required
. Select one or more workload performance areas. Additional fields appear further down in the form based on this selection.
Response Time:
Required
. Select acceptable overall server response time from the categories offered. The systems are categorized as slow, medium, or fast based on averaging response times obtained by Probe. In this field, you select the maximum allowable time frame you would be willing to accept for users getting access to the server. (Server.Planner also considers response time distribution in evaluating whether a machine qualifies.)
% Responses fall within Range:
Required
. Select percentage of responses that must be within the time range selected in the
Response Time
box.
Query Tolerance % Range:
Required
. Select allowable adjustment to the number of users, or spokes. This percentage is applied to the total number of users or spokes. For example, if you are requiring the server to support 100 users, and you select a 25% Tolerance Range, Server.Planner will include systems supporting 75-125 users, or better. If you have selected User Tasks, this tolerance range is applied to the user numbers requested for each type of workload in the User Profile.
: If you select a query tolerance range of zero, Server.Planner will include only systems having an
exact match
on the number of users, or spokes.
Specify OS Platform(s) to Include:
Select, or deselect, acceptable server platforms. Use buttons to reset all.
Disk Configuration:
Specify if RAID 5 configuration use is required. In this case, only systems configured with RAID 5 will be included in the search. If you specify "Any Configuration", configurations including RAID 5
be included in the results.
Vendor Selection:
Click on this button to view the list of vendor databases Server.Planner finds in your Notes data directory. You can select, or deselect, one or more vendor databases from the list of those to be searched. Initially all vendor databases found in your Notes Data directory are selected. If databases have been added, or removed, from the directory since the query form was opened, you must close and reopen the query form to obtain the updated list.
Mail Routing Hub Task Information
This area appears only if you selected Mail Routing Hub under
Tasks
Required
. Enter number of spokes required for mail hub. The user number constraint is adjusted by the Query Tolerance range.
Note:
Conversion Guidelines also affect this parameter. For more information, see "
Using the Conversion Guidelines View
Replication Hub Task Information:
This area appears only if you selected Mail Replication Hub under
Tasks
Required
. Enter number of spokes required for database replication hub. The user number constraint is adjusted by the Query Tolerance range.
Note:
Conversion Guidelines also affect this parameter. For more information, see "
Using the Conversion Guidelines View
User Task Information:
area appears only if you have selected User Tasks under
Tasks
. This requirement to create profiles is based on the assumption that not all users put the same demands on a server and that all users perform different activities. You can describe up to three user profiles, although only one is required. Server.Planner will prompt you for corrections when you request to run the query, if you omit a required entry, or if the entries do not total correctly.
Server.Planner will restrict its search to machines that ran the workloads that test the activities you select here. The total number of users you specify here is also used as a constraint, (adjusted by your use of the Conversion Guidelines and the Query Tolerance range). The
combination
of tasks that you select is also a factor in the machine selection.
User Task fields:
Required
. Enter percentage of time users perform each task type, for each user profile. At least one entry is required. If only one, it must be 100.
Total Number of Active Users:
Required
. Enter total number of active, connected, users to have been tested on the server.
Profile Total boxes:
Required
. Enter total number of users for this profile. Must be the sum of all the entries in this column.
Query Summary:
Enter optional descriptive text. This information appears in the View Pane.
Server Analysis Results:
These fields are completed by Domino Server.Planner when you perform the query.
Server Analysis Results Summary:
Enter optional descriptive text. This information appears in the View Pane.
Server Comparison Graphs:
Domino Server.Planner inserts the Responses distribution chart in the first area when you click on
Graph Results
. Use the buttons below each graph, if you want to change the order of the graph data presentation. The original order is the order appearing in the results table.
Domino Server.Planner inserts the Costs chart in the lower area when you click on
Graph Results
. Use the buttons if you want to change the order of the graph data presentation. The buttons act as a toggle to sort in respective ascending or descending order.
Mailing Information:
Enter mailing information here if you want to use the mail option when saving the query results. You must have entered the information before you click on
Save Query & Results
Query Information:
Enter optional descriptive text in the Comments area.
Miscellaneous:
Domino Server.Planner inserts tabulated results here when you create graphs. This information is for display purposes only; the information can't be modified.
Using the Query on Data Button Bar
The Query on Data Action Bar buttons are shown below:
?_?__
`!b`1
7`^///?
?b___
`bOOO
`<b!R
MP OS/2 Warp Con
Exit
completes the Server.Planner session and closes the Server.Planner Analyst database.
Save Query
creates a document record of a Server.Planner query in your Server.Planner Analyst database. You can rerun saved queries.
Save Query & Results
saves both your query and results in the Decision Maker database for analysis. It also gives you the option to mail it.
Query on Data
performs the query. It finds configurations in the Vendor databases that satisfy the requirements you specify in the query form.
Graph Results
creates a bar graph from the data of the configurations that satisfied your query. You can chart response time or costs.
presents pointers on using the Server.Planner databases.
If the database contains no queries, follow the instructions in the section "
Creating a Query
" to create a query.
Creating a Query
You can create a query by either of two methods:
By modifying an existing query
By using Create from the menu bar
If your Server.Planner Analyst Database contains one or more queries, you can create a new query by opening an existing one, changing its parameters, and saving it as a "new" query. (When you select "Save" , Server.Planner gives you the option to select "No" for the Overwrite option, thus saving it as an additional query.) You can use the Summary field to enter a description that distinguishes the new query from the original.
Note:
If you do respond "Yes" to the overwrite message, and you have created more than one query this way, the originally opened query is the one that is overwritten.
Use the following procedure to create a new query on the information in the Vendor Databases. Use the detailed field descriptions in the next section, if you need more information than what appears in the status bar help as you go to each field in the form.
Open the Server.Planner Analyst Database.
Go to the
Query on Data
view pane. (Click on the
Query on Data
icon on the Navigator.)
Choose
Create... Query on Data
on the Menu Bar
Fill in all the fields on the form, scrolling down as necessary to view them.
Select the
one or more
Tasks
your users will most likely perform:
Select
Mail Routing Hub
if your server is to support mail routing.
Select
Replication Hub
if your server is to support database replication.
Select
User Tasks
if your server is to support other end-user activities.
Additional fields appear on the form, depending upon the task(s) you select here.
Select an acceptable end user
Response Time
category
This is the time allowable for a person making a new connection to your server.
Select the
percentage of Responses
required to have fallen within the Response Time range, as determined by Probe. (To
widen
the field of acceptable test results, select the
lower
percentage.) Note that Probe connects to the test shared discussion database; a shared database for many of the workloads. These numbers present the worst-case scenario as Probe is trying to access the same file as all the simulated users.
Select the
Query Tolerance percentage Range
. The tolerance is applied to the number of users or spokes. (To
widen
the field of acceptable test results, select the
higher
percentage. For example, if you specify 100 spokes for replication hub, and then you specify the query tolerance percentage 10, the tool queries on a range of 90-110 spokes. Specify 0 for an exact match.)
Create a list of acceptable operating system platforms for the server by deselecting the ones to exclude. Use the
Deselect All
button to clear all the items if you are selecting only a small number of platforms. Click the
Select All
button to query on all operating systems listed.
Specify the
Disk Configuration
requirement as
Include RAID 5
only if you will require this level of data redundancy for your server.
Only
machine configurations including RAID 5 will be selected in this case. If you select
Any Configuration
, RAID 5 systems
be included.
Click the
Vendor Selection
button to select the Vendor databases you want to query on. All the downloaded Vendor Databases appear.
You can query all your Vendor Databases, or a smaller set selected from the list.
If you selected Mail Routing Hub or Replication Hub under
Tasks
, enter the number of
spokes
required by your server in the box that appears in the next portion of the form.
: The other fields are reserved for future use; a default value is used by Server.Planner for this release.
If you selected User Tasks under
Tasks
, enter the
User Profile
information. For an example, see the section below titled "
User Profile Example
This requirement to create profiles is based on the assumption that not all users put the same demands on a server and that all users perform different activities. You can describe up to three user profiles, although only one is required. Server.Planner will prompt you for corrections when you request to run the query, if you omit a required entry, or if the entries do not total correctly.
Enter information for up to 3 user profiles. (Server.Planner provides for three user profiles; only one is required.) For each type of user, estimate the percentage of activity that is typical of normal use.
Enter the
Total Number of Active Users
anticipated.
Enter the total number of users for each specific profiles. (Must add up to the Total Number of Active Users entered in the box above this line).
Enter in
Query Summary
a brief description of your query. This entry appears in the view pane list of saved queries to help you identify it later.
If you will want to mail query results, click on the arrow to expand the
Mail Information
section, and enter all pertinent mailing information. Recipient is required.
To enter additional optional information, click on the arrow to expand the
Document Information
section, and enter all pertinent mailing information.
Click
Save Query
to preserve this query for later use.
Create another query, or proceed with the instructions in the following section for running the query.
Chead
User Profile Example
The User Profile section of the Query on Data form appears on the screen when you select "User Tasks" in the
Tasks
box. It provides a way to fine tune the selection criteria based on specific types of end-user activities. In addition to specifying the total number of users the server must support, you highlight specific types of Notes activities that Server.Planner must find benchmarked by the system. In this section you create "profiles", defining up to three types of typical users based on the kind of tasks they usually perform.
You enter in the columns the percentage of time per day this type of user spends in the related activity.
You specify the total number of users the server must support.
You also specify how many users of each type make up the total population.
Based on this profile information, Server.Planner searches for systems that ran the workloads indicated by the profiles, and met, or exceeded, your requirements for that level of usage. The search algorithm takes into account the resources required to support the
combined
activities at the performance level you have specified and tries to blend discrete workloads.
The User Profile example shown below assumes:
The users represented by Profile #1 are heavy users of Groupware and the Web. They represent the majority of the 100 users.
The users represented by Profile #2 use chiefly mail and Calendar & Scheduling. They represent a minority of the 100 users.
Domino Server.Planner searches all the vendor databases you have selected for the systems that can support your requirements, when you click the
Query on Data
button.
The results of the query are tabulated on the form
, as described below this procedure.
The procedure to run a query is as follows:
Select a query from the View page list, if you do not have a query in the current window.
Make any changes to the Conversion Guidelines, if desired.
Click the
Query on Data
button on the Action Bar to generate results. Server.Planner clears the results fields from any previous runs of this query, and displays the new results.
If no vendor data meets the requirements you've specified, the message " No Query Results Found" pops up. Adjust the requirements to widen the search, if possible, and run the query again.
If more configurations match the query than Server.Planner can fit into the display, the message " Warning: Reached Limit on Output " pops up. Adjust the query requirements to narrow the search.
If you want to save the results as a mail item, fill in the
Mailing Information
section. Recipient is a required field.
If you want to save the results in the Decision Maker Database, or save them as a mail item, (or both), click the
Save Query & Results
button.
Server.Planner searches for machines that meet, or exceed, your requirements. Your results are summarized in a table containing Vendor Name, Machine ID, Machine Description, Response Time, CPU Type/Number and Platform.
This example specified 100 users and Mail Routing Hub, and Medium for response time. Note that one machine, "Charlie", although classified as "Slow", qualified in this query for a Medium-Response-Time system. "Charlie" isn't disqualified because of the "Slow" average Response Time; rather, "Charlie" still qualifies as "Medium" because it met the specification of % responses falling within range.
is the unique identifier assigned to the machine by the vendor.
Description
is information provided by the vendor.
Response Time
indicates where this machine qualifies in the Probe response-time categories with respect to
average
response time: slow, medium, or fast. In some cases, Server.Planner may determine that a machine that appears to be outside your specification would actually satisfy the responsiveness requirement, due to the
distribution
of response times.
CPU/#
is the CPU product name, the speed, and the number of CPUs on the system. If the system is a multiprocessor, SMP is appended.
Platform
describes the operating system tested.
The following example was the same query as above, specifying 100 users and Mail Routing Hub, but also 100 users with Replication Hub. Notice how this narrowed the number of qualifying systems, because of the combined requirements.
Note that "Charlie" no longer qualifies in this query for a Medium-Response-Time system.
Server.Planner creates graphs of results obtained by the query, if you have installed Lotus Components.
You must first run the query before you can create the graphs.
The Graphical Information section of the form must be open (expanded) to begin. Click on the arrow if it is pointing to the right to open the area. (Or use
View Expand All Sections
Click the
Graph Results
button after you get your query results. (The menu bar is now the Lotus Components menu bar).
Select Responses or Costs, or both, from the pop-up box:
?_?__
`!b`1
7`^///?
?b___
`bOOO
`<b!R
Dx
3333333333333333
3333333333333333
3333333333333333
3333333333333333
When you click
, the graph, or graphs, are created.
You may need to use the grab handles to adjust the size of the graph to see all the graph details.
Use the Lotus Components menu bar to change the graph (chart) attributes, as desired. You can choose from several graphical formats.
When first created, the data in the graph is presented in the same order as in the results table. Graphs can optionally be sorted by Machine
or Response Time, and Machine Name or Cost, using the buttons below the graphs in the form.
button to store both query and results. Server.Planner
saves the query and results first in the Analyst Database, and then in the
Decision Maker Database.
Enter the name of the destination server, if not your local machine, and a filename for the
Decision Maker Database
, if you want a name other than "OUTPUT".
Server.Planner creates the database, based on the
Server.Planner: Decision Maker
template, if it does not already exist. You can store the results of many different queries in the same output database.
You can also mail the results of your query to yourself or to others. To do so, enter the appropriate information in the following fields of the "Query on Data" form:
Note:
For more information on sending mail, see Notes Help.v
Subject:
Delivery Priority:
Normal (default)
Delivery Report:
Basic (default)
Return Receipt:
Server.Planner stores and/or mails the Machine Identifier information, the graphs, and the list of vendor databases included in the search, along with your results.
The format and use of the Decision Maker Database is described in the next Chapter of this document.
About queries that produce unexpected results
Let's say the results of a query don't include a particular machine, as you expected. There are several reasons why the results of a query might differ from what you expect:
The performance criteria that you specified wasn't met.
The workload that you're querying on doesn't exist. In this case, contact the vendor.
More than 10 machines met the criteria. The maximum number of items that can display on the screen is 10. It's possible that more than 10 machines met your criteria. Fine tune your query so that fewer machines display on the screen.
The parameters you set might be too wide or too narrow. Adjust the parameters (widen or narrow the scope); recheck current specifications.
Table Text
Column Head
Table Title
formula
procdure Chd
procdure Bhd
indent Term
indent find
indent caution
caution
indent tip
indent Note
DocLink Head
indent bullet
Chead
bullet
Bhead
Ahead
2nd lev para
Body Text
DocLink
Numbered List
1st lev para
Figure Title
Bhead
Creating the Server.Planner Vendor Database
Body Text
If you are a vendor who wishes to create Server.Planner Vendor Databases, consult the
NotesBench Consortium
web site (
http://www.notesbench.org
) for complete NotesBench information.
Vendors use two NotesBench tools, as part of the benchmark runs, to create the data in their Vendor Databases:
Probe
- To record response times of the server during the benchmark by continuously opening and closing a Notes database.
Notesnum
- To format and store NotesBench test results in the Vendor Database.
To create a Vendor Database:
File > Databbse...New
from the Notes menu bar, and apply the Vendor Database template (
dspv.ntf
) that you have placed in your Notes Data directory. It appears in the pull down list as Server.Planner: Vendor.
Create a name for the database of the format CP
databasename
.nsf.
(Server.Planner selects only
.nsf
files with the CP prefix when performing the analyst's queries during the capacity planning process).
?_?__
`!b`1
7`^///?
?b___
`bOOO
`<b!R
le Paragraph Spa
Notes SPR Auth
tes SPR Auth
otes Log Analysi
.
$Revisions
When you set up the notes.ini bile for the NotesBench, you supply this database name in the
CPdatabase
variable. After running the workload(s), you run Notesnum to roll up the data. The data from each workload that was run becomes a document in this database. You must open and close each document once to finish the calculation and update process. The data is then available in an easily readable format.
(.txt)