Professional Documents
Culture Documents
SAP AG
400 Technology Square, Floor 9
Cambridge, MA 02139
September 2007
Table of Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Audience for this Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Topic Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Chapter 1: E-Sourcing Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Chapter 2: Optimizer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Chapter 3: Contract Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Chapter 4: Deploying SAP E-Sourcing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Confidential i
Table of Contents
Installation Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Installing the E-Sourcing Binary Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Configuring the Software for Deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Updating the Java Home Directory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Installing the Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Creating or Updating Database Login Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Updating the EAR File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Including Custom Jar Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Defining the Application Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Set Session Timeout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Post-installation Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Task 1: Add SAP-Required Libraries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Task 2: Define Java Memory and Configuration Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Task 3: Define JAXP Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Task 4: Configure JMS Messaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Configuring JMS Queues and JNDI Aliases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Defining System Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Task 5: Update Session Timeout (Optional) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Task 6: Upgrade Database Schema from Previous Release. . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Task 7: Deploy the Application EAR File to the Application Server . . . . . . . . . . . . . . . . . . . 19
Task 8: Generate License File for E-Sourcing Explorer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Task 9: Verify Production Readiness Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Environments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Tracking Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Naming Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Troubleshooting Object Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Export Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Import Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Preface
Confidential 1
Audience for this Guide Preface
Overview
Depending on the license and distribution media, the SAP E-Sourcing installation can include the
following components:
` E-Sourcing Platform
` Optimizer
` Contract Generation
` Exchange Calendar Synchronization
The figure below provides an overview of the SAP E-Sourcing components and the required and
optional services. Items in bold are required.
Topic Organization
Chapter 2: Optimizer
The SAP Optimizer is also a J2EE application that requires a J2EE-compliant application server
that supports the EJB2.0 spec. The E-Sourcing application makes requests to the Optimizer service.
The Optimizer has no direct end-user interface. The Optimizer is typically deployed on a separate
physical application server from the E-Sourcing application. This prevents the
computation-intensive processing of the Optimizer from impacting the performance of user
interaction with the E-Sourcing application.
This chapter describes how to install the Optimizer application.
Confidential 5
Overview 1 Installing the E-Sourcing Platform
Overview
Before you can install the SAP E-Sourcing software, you must install software on the application
server and the database server, and configure a number of database settings on the database server.
To install SAP E-Sourcing, you must perform three sets of tasks:
` Pre-installation procedures
` Automatic installation and configuration of the software
` Post-installation procedures
If this is a new installation, you will also need to implement the SAP E-Sourcing application as
defined in the Online Help under Setup. Although the configure utility creates a baseline database
for a new installation, this baseline database lacks the business-specific data required for a fully
functioning system.
The Setup section of the Online Help describes how to define clusters for load balancing, support a
multiple-enterprise context, and define master data and user accounts. You can define this data
interactively or via a deployment workbook. The deployment workbook is an Excel spreadsheet
that you use to define the data and then import it into the application. Instructions for its use are
included in Chapter 6: Deploying SAP E-Sourcing.
Performance note: As with other web-enabled applications, the performance of the E-Sourcing
client is highly dependent upon the network infrastructure between the web browser client and the
application server. E-Sourcing performance may suffer if sufficient bandwidth is not provided.
Pre-installation Procedure
The following items must be configured before installing the E-Sourcing software.
Universal Tasks
security.provider.n
Oracle Tasks
Note: If you are upgrading from Sourcing 2.x, you must obtain this new JDBC driver.
DB2 Tasks
Installation Procedure
The two primary steps of the installation procedure are:
1 Installing the E-Sourcing and third-party binary files.
2 Configuring the software for deployment.
Note: Application and software data are commingled within the same database schema.
Software configuration settings are found in the FCI.HOMEDIR/config directory
of the local file system (fcisystem.properties and fcilocal.properties).
Note: The installer performs a full search for the JVM. This may take a while in some
cases; you can specify the location of the JVM by using the following:
-is:javahome /yourpathname
The first step of the installation procedure is responsible for the deployment of the software
binaries and requires that you do the following:
` Accept the license terms of the software.
` Specify the directory location for the FCI software. This will be the FCI home directory.
Note: The path for the FCI home directory must be all lower case or the
installation/upgrade will not be successful.
` Select the appropriate database: Oracle or DB2.
■ If you select Oracle, you see a page asking you to provide the Oracle JAR file (normally
oracle.jar).
■ If you select DB2, you see a page asking you to provide the two DB2 JAR files (db2cc.jar
and db2cc_license_cu.jar).
` Specify a single or multibyte database character set.
Note: Again, the installer will perform a full search for the JVM. You can specify the
location of the JVM by using the following:
-is:javahome /yourpathname
The configure utility can perform the following functions.
For Oracle:
Enter the following database login information:
` Driver Class (required): oracle.jdbc.driver.OracleDriver
` URL to connect to the database and the username and password to connect to the database
schema (required). Modify the URL to include the host name, port, and SID for your server.
The URL must be in the following format:
jdbc:oracle:thin:<hostname>:<port>:<SID>
For example:
jdbc:oracle:thin:@localhost:1521:XE
` Username and Password: the user ID and password of the Oracle account used by the
application to access the database (required). This is the connection information you obtained
during the pre-installation procedure.
` Schema (typically, the user name -- required for upgrade, recommended but not required for
installation)
` Optionally, the names of custom database tablespace(s) used for storing tables (Table Space),
indexes (Index Space), and blobs (Blob Space). If you leave these fields blank, the default
Oracle settings are used.
For DB2:
Enter the following database login information:
` Driver Class (required): com.ibm.db2.jcc.DB2Driver
` URL to connect to the database and the username and password to connect to the database
schema (required). Modify the URL to include the host name, port, and database for your
server. The URL must be in the following format:
jdbc:db2:<hostname>:<port>/<database>
For example:
jdbc:db2://localhost:5912/ES
` Username and Password: the connection information you obtained during the pre-installation
procedure (required).
` Schema (required)
` The names of database tablespaces used for storing tables (Table Space), indexes (Index
Space), and blobs (Blob Space) (required)
In some cases, this setting may be overridden by a session timeout that is set at the application
server level. For example, WebSphere has a session timeout property at the application server level.
In this situation, do one of the following:
` Using the WebSphere admin console, use the application server property to control session
timeout.
` Using the WebSphere admin console, disable the timeout at the application server level.
` Configure the SAP application timeout setting to override the application server setting via the
Session Management page for SAP E-Sourcing.
Post-installation Procedure
Once the E-Sourcing application has been installed and configured, the following post-installation
tasks must be performed:
1 Add SAP-required libraries to server class path.
2 Define Java virtual memory and configuration settings.
3 Define JAXP settings.
4 Configure JMS messaging.
5 Optionally, update the default session timeout.
6 Upgrade the database schema from previous release.
7 Deploy the application EAR file to the application server.
8 Generate license file.
9 Verify production readiness criteria.
The libraries required by the SAP E-Sourcing application are packaged in the EAR file that is
deployed by the application server. However, the mechanism used by the application server to load
XML parsers requires some of the libraries to be available to the application server. The following
libraries must be added to the application server’s classpath:
` <FCIHOMEDIR>/lib/activation.jar
` <FCIHOMEDIR>/lib/avalon-framework.jar
` <FCIHOMEDIR>/lib/batik-all.jar
` <FCIHOMEDIR>/lib/endorsed/serializer.jar
` <FCIHOMEDIR>/lib/endorsed/xalan.jar
` <FCIHOMEDIR>/lib/endorsed/xercesImpl.jar
` <FCIHOMEDIR>/lib/endorsed/xml-apis.jar
` <FCIHOMEDIR>/lib/fcifopext.jar
` <FCIHOMEDIR>/lib/fop.jar
` <FCIHOMEDIR>/lib/jakarta-oro-2.04.jar (WebSphere only, if using the
upp.extservletauthenticator.username_regex property to implement a single sign-on solution)
` <FCIHOMEDIR>/lib/jcschart.jar
` <FCIHOMEDIR>/lib/mail.jar
` <FCIHOMEDIR>/lib/soap.jar
` <FCIHOMEDIR>/lib/ssce.jar (WebLogic only)
For example, if you define <FCIHOMEDIR> to be c:/fci, you might define an FCI_CLASSPATH
variable with the following values:
SET FCI_CLASSPATH=
c:/fci/lib/xercesImpl.jar;c:/fci/lib/fop.jar;c:/fci/lib/fcifopext.jar;c:/
fci/lib/avalon-framework.jar;c:/fci/lib/batik-all.jar;c:/fci/lib/saxon.ja
r;c:/fci/lib/soap.jar;c:/fci/lib/mail.jar;c:/fci/lib/activation.jar;c:/fc
i/lib/jcschart.jar;c:/fci/lib/cocoon-fci.jar;c:/fci/lib/ssce.jar
Note that the FCI_CLASSPATH environment variable must still be added to the server’s classpath
environment variable.
The Java API for XML Processing (JAXP) provides a mechanism for defining the XML and XSL
processors used by the E-Sourcing application. SAP E-Sourcing ships with the required XML and
XSL processors. The following JAXP properties must be defined to allow SAP E-Sourcing to use
the required processors:
` javax.xml.parsers.DocumentBuilderFactory=org.apache.xerces.jaxp.DocumentBuilderFactoryIm
pl
` javax.xml.parsers.SAXParserFactory=org.apache.xerces.jaxp.SAXParserFactoryImpl
` javax.xml.transform.TransformerFactory=org.apache.xalan.processor.TransformerFactoryImpl
-
---- this one changed
The JAXP settings can be defined in a JAXP specific property file, using the Java
–D<name>=<value> option, or using an application-server-specific mechanism for defining JAXP
settings.
Note: The setup of these queues depends on the application server. Most will
auto-prefix queue names with “queue/”.
` Integration API
Queue: fs.destination.queue.IapiEventQueue
JNDI: fs.queue.IapiEventQueue
` Optimization (see the Installing the SAP Optimizer chapter for more details)
Queue: fs.AllocationStatusQueue
JNDI: fs.optimization.AllocationStatusQueue
Queue: fs.AllocationRequestQueue
JNDI: fs.optimization.AllocationRequestQueue
Queue: fs.HaltAllocationQueue
JNDI: fs.optimization.HaltAllocationQueue
Field Value
jms.common.queueconnectionfactory jmsfactory/default/fs.common.QueueConnectionFactory
jms.destination.queue.jmx_queue jmx_queue jmsqueues/default/fs.queue.JmxQueue
jms.destination.queue.iapi_event_queue jmsqueues/default/fs.queue.IapiEventQueue
jndi.java.naming.factory.initial com.sap.engine.services.jndi.InitialContextFactoryImpl
jndi.java.naming.factory.url.pkgs com.sap.tm.jndi.providers
jndi.java.naming.provider.url localhost:50004
Field Value
optimization.config.queue_connection_fact jmsfactory/default/fs.optimization.QueueConnectionFacto
ory ry
optimization.config.halt_destination jmsqueues/default/fs.optimization.HaltAllocationQueue
optimization.config.request_destination jmsqueues/default/fs.optimization.AllocationRequestQueu
e
optimization.config.status_destination jmsqueues/default/fs.optimization.AllocationStatusQueue
jndi.java.naming.factory.initial com.sap.engine.services.jndi.InitialContextFactoryImpl
jndi.java.naming.factory.url.pkgs com.sap.tm.jndi.providers
jndi.java.naming.provider.url localhost:50004
This chapter describes how to configure and deploy the Optimization module and
how to configure SAP E-Sourcing to use it.
This chapter covers the following topics:
` Overview on page 22
` Pre-installation Procedures on page 22
` Installation Procedures on page 25
` Post-installation Procedures on page 26
Confidential 21
Overview 2 Installing the E-Sourcing Optimizer
Overview
The following examples are based on the JBoss server, but can be extrapolated to other application
servers.
Before deploying the Optimization module, set up the standard SAP E-Sourcing environment.
Note: For more information on the JBoss software, see the JBoss Configuration Guide:
http://www.jboss.org/online-manual/HTML/
Pre-installation Procedures
The Optimizer must be installed on a machine that is physically separate from the SAP E-Sourcing
server. Technically, the system will still function if the Optimizer is run on the same machine in a
different Java VM, but this is strongly discouraged.
Http LDAP
Web Server
J2EE LDAP Directory
Client Http
SSL Application Server
JDBC
SMTP
JMS
SMTP Gateway
Windows 2000
JBOSS 2.4.3
Jetty 3.1.3-1
J2EE
Application Server
Optimizer
Note: The alias is the coded name that is used for locating each queue. Having both an
actual queue and an alias is the best practice for working with JMS. However, the
aliases are not strictly required if the queues themselves use the coded names.
It is also very important that each queue be distinct. There should not be a single
queue with multiple aliases. Furthermore, these queues should be used only for
SAP E-Sourcing.
<attribute
name="ToName">queue/fs.AllocationStatusQueue</attribute>
<attribute
name="FromName">fs.optimization.AllocationStatusQueue</attribute>
</mbean>
<mbean code="org.jboss.naming.NamingAlias"
name="DefaultDomain:service=NamingAlias,fromName=fs.optimization.Alloc
ationRequestQueue">
<attribute
name="ToName">queue/fs.AllocationRequestQueue</attribute>
<attribute
name="FromName">fs.optimization.AllocationRequestQueue</attribute>
</mbean>
<mbean code="org.jboss.naming.NamingAlias"
name="DefaultDomain:service=NamingAlias,fromName=fs.optimization.HaltA
llocationQueue">
<attribute
name="ToName">queue/fs.HaltAllocationQueue</attribute>
<attribute
name="FromName">fs.optimization.HaltAllocationQueue</attribute>
</mbean>
<!-- END FS OPTIMIZATION ALIASES -->
Note: If you directly edit the configuration files, you must restart JBoss. This is not
necessary when working with the management console.
Installation Procedures
Post-installation Procedures
After the Optimizer is deployed, you must configure SAP E-Sourcing to indicate that it is available
for use. Perform the following two post-installation tasks:
` Identify the Optimization server.
` Enable the optimization functions for each specific module, such as RFx and Auctions.
Confidential 27
Pre-installation Procedures 3 Installing Contract Generation
Pre-installation Procedures
The contract generation service is a Microsoft .NET service. A typical installation would install the
service on a separate machine independent of the application server. However, the service can be
installed on any server that meets the requirements.
The following must be installed and configured before installing the contract generation service:
` Microsoft Windows 2000/2003 server
` Internet Information Services (IIS)
` Microsoft .NET Framework 2.0
` Run Windows Update and install all patches and updates, including those for Windows, IIS,
and the .NET framework.
` Read the following article to understand how to install and configure Office applications to run
as a specific user: http://support.microsoft.com/default.aspx?scid=kb;EN-US;288367
Note: When performing this configuration, you do not need to perform step 18, which
describes how to set up an NT service.
` Install Microsoft Word 2003 as follows:
n Run the Office 2003 setup and select only the Word 2003 option for installation.
n Select Choose Advanced Customization of Application.
n On the following page, expand and select all the available nodes in the tree, especially all
the .NET Programmability Support nodes.
` There is a known issue with the following Microsoft Office Update: Applying this update
prevents the contract generation service from working.
Excel 2002 Update: October 16, 2002
Prevents users from unintentionally updating links to external data within Microsoft Excel
spreadsheets.
http://www.microsoft.com/downloads/details.aspx?FamilyID=09708964-A8D6-4F4
E-9D82-280E08F36E8A&displaylang=en
DO NOT apply this update to the server running the contract generation service.
Installation Procedures
You must have administrator privileges to install the contract generation service. Be sure to log in
as a member of the administrator group. The Setup program for contract generation is provided on
the distribution media. Once the prerequisites have been met, simply run the install program
WordServiceInstaller.msi.
You will be prompted to enter the Virtual Directory and Site where the web service should be
installed. This information will be required to properly configure the FCI application to
communicate with the service.
Post-installation Procedures
1 Log into the Web Service machine as the OfficeAutomationUser and start Microsoft Word. In
Word, do the following:
a Set the default file format to the format that is appropriate for your deployment. You can
find this setting by clicking the following:
Tools -> Options -> Save - Default Format -> Save Word Files As
b Add the FCI Schema to the Schema library. Click the Add Schema button, browse to the
directory where the web service is installed, and select the fci.xsd file. You can find the
Schema library by clicking the following:
Tools -> Templates and Add-ins -> XML Schema
c Set the default Web encoding to Unicode (UTF-8). You can find this setting by clicking
the following:
Tools -> Options -> General -> Web Options -> Encoding
2 Create the fci directory on the server. Create logs and temp subdirectories under the fci
directory. It is recommended that you create this directory directly under the main drive (for
example, C:\FCI, C:\FCI\logs, C:\FCI\temp).
3 Locate the web config file under the following directory: inetpub\wwwroot\<virtual
directory>\ and set the fciDir application key value to the fci directory you just created under
the <appSettings> section (for example, check and set the first node to <add key=”fciDir”
value=”c:\FCI”/>
4 Change the values for the other appSettings section parameters in the web.config file as
follows:
n showWord: Make MS Word visible or not. Normally set it to false in deployment and true
in debug.
n logLevel: Indicate the level of detail in the log information. Set either to debug or info
level.
n wordRunNumber: The contract generation server will regularly shut down MS Word and
restart it after the specified number of operations to improve the Word performance.
n requestTimeout: the contract generation server can only process one request at a time.
Other requests are queued. If a request has been in the queue a specified length of time, it
will time out. Set the timeout number in seconds.
n A timeout mechanism is used to catch Word message box exceptions. For example, if
Word cannot open a document in 30 seconds, it is assumed that Word is frozen by a
message box waiting for user interaction, which is not supported. In this case, an
exception will be thrown and Word will be restarted. Common Word operations such as
open file, save file and insert file are monitored.
n openFileTimeout: maximum time (in seconds) allowed to open a file using Word.
n saveFileTimeout: maximum time (in seconds) allowed to save a file using Word.
n closeFileTimeout: maximum time (in seconds) allowed to close a file using Word.
n insertFileTimeout: maximum time (in seconds) allowed to insert a file using Word.
n compareDocsTimeout: maximum time (in seconds) allowed to compare two Word
documents using Word.
5 To clean up obsolete Word instance and the temporary directories, it is necessary to
impersonate the automation user with the buildin ASPNET account. DPAPI is used to encrypt
the identity sections in ASP.NET 2.0 to protect sensitive account information. Please refer to
the following link:
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnpag2/html/paght000005.asp
In summary, the steps are as follows:
a Create a new Web site named MachineDPAPI. Ensure that this directory is configured as a
virtual directory.
b Add a Web.config configuration file to this directory. Add the following content to the
file, substituting the correct username and password:
<?xml version="1.0"?>
<configuration>
<identity impersonate="true"
userName="domain\automationUserAccount here"
password="account password here" />
</configuration>
c Open the .NET 2.0 command prompt (start->All Programs->Microsoft .NET Frameword
SDK 2.0->SDK Command Prompt, or open the Command Prompt and cd to directory
Windows\Microsoft.NET\Framework\v2.0.xxx on the Windows 2003 server) and enter
the following command:
aspnet_regiis -pe "identity" -app "/MachineDPAPI" -prov
"DataProtectionConfigurationProvider"
You should see the output as:
Confidential 33
Overview 4 Deploying SAP E-Sourcing
Overview
The following diagram outlines the basic deployment process.
Define
Define Define Define Define
Numbering
Cluster Context Extensions Accounts
Tables
` System setup phase - The first task when developing a new E-Sourcing enterprise is to define
the system-level components that will ultimately manage your data and user deployment. An
enterprise consists of at least one cluster and one context. Clusters serve to map users to a
deployment instance when an environment contains multiple instances of E-Sourcing.
Contexts serve to create a logical definition of how data should be shared or not shared in a
complex business environment. In short, contexts manage access to information across a
corporate deployment. Define Contexts and Clusters using the default system administrator
login, portal and instructions defined in the following sections of this manual.
` Enterprise setup phase - The second task is to install the master data that will be shared by all
sourcing companies within the enterprise. Most master data falls in this category. This data is
generally installed using the built-in enterprise administration account (although ongoing
maintenance should be handled by actual activated users).
` Company setup phase - The last phase of any deployment is to install master data specific to
each company. This phase is distinguished from the second phase because it must be applied to
each installed company with its own context.
Note: System Administrator rights are required to create initial Contexts and Clusters.
Define Contexts and Clusters using the login and passwords defined below.
System Administrator system Use for creating and managing contexts, clusters, and
global system properties.
Enterprise Administrator enterprise Used to bootstrap the enterprise by installing everything
from numbering tables to accounts. Once real accounts are
installed, they should be used for any remaining
configuration.
Note: To ensure that your system remains secure, disable the use of the system accounts
after creating custom accounts specific to your organization,.
Clusters
A cluster is a logical group of machines. A cluster defines how end users access the E-Sourcing
application. For example, you define a cluster to map a large volume of related users to a specific
instance of a deployment when multiple instances exist within the same server. The system requires
that at least one cluster exist within an enterprise. Clusters define how one sourcing context is
differentiated from another, and serve as the root of the entire deployment. Create your initial
cluster by logging into the administration portal page as the system administrator and selecting
Cluster Configuration from the Setup page.
For more information about clusters, see the Setup section of the Online Help.
Contexts
Contexts define how master data and security are managed across a complex enterprise. Use
Contexts to restrict a series of master data components from being shared across an enterprise.
Contexts provide the ability to logically separate and duplicate master data components so that they
can be managed differently across a conglomerate. This is most commonly used in a complex or
global environment when international requirements or corporate divisions require that master data
is managed separately. At least one sourcing context must be set up for an enterprise.
Master Data
Define the components of master data that exist within each context after contexts and clusters have
been defined. The following table lists the master data object types requiring definition for the
enterprise. A data classification of Configurable indicates that default data is provided but can be
customized for deployment. A data classification of Custom indicates that no default data is
provided and the deployment must provide the data.
Numbering table definitions for Configurable Defines the format of IDs. For example, an RFP
enterprise-level objects. might follow the format: RFP-nnnn, where nnnn
increments for each new rfp created.
Default list of extendible Configurable Lists those enterprise-level objects which can be
enterprise-level objects. extended.
Default list of objects that can Configurable One type of extension allows a reference to
used referenced in an extension. another type of object. This list identifies the
allowed kinds of references.
Value List Types Configurable Default value list types.
Value List Values Configurable Default value list values, such as countries, time
zones, etc.
Units of Measure Configurable Standard units of measure used throughout the
system
Currencies Configurable A list of all currencies used by your
organization
Currency Conversions Configurable Default values for any currency conversion rates
Document Link Definitions Configurable Lists the allowable kinds of document links.
Scheduled Task Types Configurable Used to define categories of scheduled tasks.
Scheduled Tasks Configurable Lists scheduled/background tasks.
Metrics Configurable Used to define metrics for scorecards.
Contract Terms Configurable Used to define terms on contracts
Contract Types Configurable Types define what set of contract features are
available.
Project Types Configurable Types define what set of Project features are
available.
RFx Types Configurable Types define what set of RFx features are
available.
Auction Types Configurable Types define what set of Auction features are
available.
DA Types Configurable Types define what set of Demand features are
available.
Security Profiles Configurable Identifies the security profiles installed with the
system.
Rights on security profiles. Configurable Defines the rights for each security profile.
Time Periods Configurable Defines information about quarters, years, etc.
Reporting Calendars Configurable Used for managing how data is reported with
relation to the calendar.
Collaborator Roles Configurable Defines the role (rights and responsibilities) of a
person/group on a specific document instance.
Document Security Templates Configurable Sets up the default collaborators and security
when new documents are created.
Shared Discussions Configurable A list of discussion forums available as a
channel and not relating to a specific document.
Workbench Channels Configurable Channels available on a user's desktop.
Workbench IFrames Configurable Data to create iFrames
Workbench Channel Configurable Channel configuration parameters.
Configurations
After configuring the enterprise master data, you must next configure the company master data.
This data must be created once for each sub-context. This example only uses a single context, so
this data must only be created once.
To install master data via the client DB Install tool, do the following:
1 Locate the home directory, typically c:/fci or d:/fci, and go to the bin directory.
2 Run the program dbinstall.bat or dbinstall.csh, whichever is appropriate for your operating
system. The DB Importer tool will open.
3 Click the Lookup icon for the File field and select the file to import.
4 Click the Lookup icon for the Context field. Choose the destination context (for example,
Acme Enterprise) where the data should be inserted or updated.
5 If system users have already been created, log in as a user with administrative rights. If users
have not been added, use the enterprise administrator built-in account. Use the Username and
Password text fields to set the login parameters.
6 Click Run.
7 Click the Status tab to see the progress of the import.
After a few minutes, the import will complete. The Script tab should show nothing but green
check marks.
8 Close the DB Import tool.
The deployment workbooks can be downloaded from the web user interface by anyone with system
administrator access rights. The deployment workbooks are included in the online Reference Guide
(RG).
When starting a new deployment project, preserve the original workbook templates by saving them
with a new name. Import data using your modified templates only. This ensures continuity if
additional deployments are required in the future. Keep the following in mind when using
deployment workbooks:
` The Configuration Sheet controls the import process. Refer to the Help Sheet for input
specifications of Configuration Sheet columns.
` The system can import the workbooks directly, but there are limitations to what Excel
constructs are supported. The Help Sheet describes the rules that must be observed.
` The same importing process supports both CSV files and Excel workbooks.
Note: This guide assumes that the default values are accepted for the remaining data
types.
Note: If you want to build all custom data in the UI, simply deactivate the import of the
above data types in the deployment workbook. See the Help page for instructions
on how to skip a specific worksheet.
Installing a Cluster
1 Log into the system as follows:
a For Username, enter system.
b For Password, enter manager. (You may have to see your system administrator for an
updated password.)
c Select Login.
2 Click Cluster Configurations on the Setup page and then click Create.
Note: Configuring a cluster can be a complicated task. See the Setup section of the
Online Help for more information about setting up clusters. The steps below assume
a basic knowledge about clusters and system administration.
3 Configure the cluster with the following parameters:
n Cluster Identity: acme.cluster.id (for example; you can use any id, but it can't have spaces
or special characters)
n Display Name: ACME Cluster (for example; you can use any name you wish)
n Hostname: *
n Add your host to the cluster member list. In the Service Hostname box, enter the name of
your host. Select a service and make sure the Is Active checkbox is selected.
4 Click Save Document.
5 Click Setup to return to the Setup page.
Note: The steps above will set the public hostname of the cluster to "*". This is a special
token which matches any HTTP request to the server. Make sure that only one
cluster is configured with the "*" token.
Installing a Context
1 Click Enterprise Contexts on the Setup page and then click Create.
2 Create a new context with the following parameters:
n Display Name: ACME E-Sourcing Context (as an example; any name will do)
n External ID: acme.context.id (as an example; any id will do it can't have spaces or special
characters)
n Sys Admin Password: Enter something you can remember.
n Default Cluster: ACME Cluster (the one created above)
3 Click Save Document. The system creates a new Enterprise E-Sourcing context to hold your
data. This may take several minutes.
4 Log out.
/images/companylogo/<company.id>/<image.gif>
Where this is true:
n <company.id> specifies the External ID from the company where the logo is attached
n <image.gif> specifies the name of the image which was attached to the company
An example would be:
/images/companylogo/acme/acme_logo.gif
11 Click Save Document.
12 Reload the property cache as follows:
a Click Setup.
b Under Administrative Reports, select System Information from the drop-down list.
c Click OK.
d Click the Cache tag and select Refresh for the properties cache.
13 Log out.
The new company logo is displayed.
Confidential 49
Overview Appendix A: Performance Tuning
Overview
The following are the recommended performance configurations for SAP E-Sourcing 3.2 on the
Solaris 8 Platform. These recommendations assume the use of the 1.4 compliant JVM that has been
certified with WebLogic 8.1.
Note that these configurations were tested in the SAP QA LAB on a Sun E450 with 2G of memory.
The following configurations are recommended for this environment:
` WebLogic Native I/O Performance Pack
` WebLogic Socket Readers
` WebLogic Execution Threads
` Sun JVM Threading Model
` Sun JVM Heap Size
MEM_ARGS="-XX:+AggressiveHeap
Confidential 53
Overview Appendix B: Running Multiple Instances on a Single Machine
Overview
In some customer environments, it may be desirable to run more than one instance of SAP
E-Sourcing on a single machine. For example, a customer may want to be able to run the current
version of the software for training purposes while also running the next version of the software in
a staging environment. In another example, a customer may want to leverage some untapped
capacity on application server boxes. In either case, the application server's resource management
will greatly influence the success of this configuration. Depending on the proprietary resource
management of a customer’s chosen application server vendor, this multiple instance configuration
may present a performance degradation or a performance optimization.
This appendix describes the steps required to configure SAP E-Sourcing to operate successfully
within a multiple instance configuration. The decision to use this configuration should be based
upon the application server vendor’s claims for performance.
Assumptions
This appendix assumes that you have experience with deploying SAP E-Sourcing and that you
understand the performance implications of running multiple instances of the Java Virtual Machine
on one physical application server, as described in the literature provided by your application server
vendor.
The steps required to configure the system properties in each instance of SAP E-Sourcing are
described here, so that multiple instances can run safely on one physical machine. It is
recommended that any machine that will be used to run multiple instances of SAP E-Sourcing have
at least 4GB of RAM.
For example, one machine would have the following home directories if intending to run three
concurrent instances of SAP E-Sourcing:
■ C:\fci_instanceA\home
■ C:\fci_instanceB\home
■ C:\fci_instanceC\home
2 SAP Installation: Perform a complete product install to each of the directories created in step
(1) above.
3 SAP System Properties: For each installation instance of SAP E-Sourcing, there will exist a
unique fcilocal.properties file located under <FCI_HOME>\config. In this example, the
following files would exist:
■ C:\fci_instanceA\home\config\fcilocal.properties
■ C:\fci_instanceB\home\config\fcilocal.properties
■ C:\fci_instanceC\home\config\fcilocal.properties
By default, these files will be empty. However, each of these files must be edited to add a
unique value for the following properties:
PROPERTY VALUE
system.service.name= <user-defined unique name>
system.daemons.events.listen.port= <user-selected available port>
In the example, add the following to C:\fci_instanceA\home\config\fcilocal.properties:
system.service.name=FCI-Instance-A
system.daemons.events.listen.port=5028
In the example, add the following to C:\fci_instanceB\home\config\fcilocal.properties:
system.service.name=FCI-Instance-B
system.daemons.events.listen.port=5029
In the example, add the following to C:\fci_instanceC\home\config\fcilocal.properties:
system.service.name=FCI-Instance-C
system.daemons.events.listen.port=5030
Please note that when only one instance of SAP E-Sourcing is running, it would choose a
default port number for the events daemon. To prevent multiple instances from choosing the
same port, it is recommended that you enter a unique port number in each of the files as
described above. Note also that the service name is just a description and need not match any
particular directory name.
4 Application Server Startup script: If the Application Server startup script contains the java
option -DFCI.HOMEDIR=/<path_to_FCI_HOME>, it must be modified to point to the
appropriate <FCI_HOME> for each instance of SAP E-Sourcing. If the startup script does not
use this parameter, this step can be skipped.
5 Reboot the Application Server: The Application Server must be restarted to respond to any
changes in the startup script and to respond to the new information in each fcilocal.properties
file.
6 Cluster Configuration: After restarting the Application Server, modify the Cluster
Configuration through the SAP E-Sourcing UI. From SetUp >> System Setup >>
Configuration Dropdown >> Cluster Information List, edit your cluster information as follows:
a Select the cluster containing the physical machines that have just been reconfigured to run
multiple instances of SAP E-Sourcing.
b Add the new member services to the cluster by the name that was entered as the service
name in the fcilocal.properties file. In the example, the following new cluster members
would be added:
FCI-Instance-A
FCI-Instance-B
FCI-Instance-C
This list is used to determine the actual host to which the user request will be redirected.
During load balancing, users will be distributed across all available members in the list.
Note that if the cluster is set to Automatically Add New Services to this Cluster, the
services will be listed automatically.
Make sure to enter the correct host names.
Future Upgrades
Note that the release upgrade steps provided with each new SAP E-Sourcing release will not
remind you to carry forward this multiple instance configuration. To maintain this configuration
with future releases, you must upgrade each of your multiple installation directories with the new
release and repeat all subsequent steps described in this document.
Confidential 61
Data Storage Security Appendix C: Security Information
Confidential 65
Backup Recommendations Appendix D: Operation and Maintenance
Backup Recommendations
You need to regularly back up your system landscape to ensure that you can restore and recover it
in the case of failure.
The backup and restore strategy for your system landscape should not only consider SAP systems
but should also be embedded in overall business requirements and incorporate your company's
entire process flow.
In addition, the backup and restore strategy must cover disaster recovery processes, such as the loss
of a data center through fire. It is most important in this context that you ensure that backup devices
are not lost together with normal data storage (separation of storage locations).
In the E-Sourcing application, all data is contained in the database, with the exception of the
configuration files located in FCI_HOME/config that detail how to access the database. Therefore,
to back up E-Sourcing, do the following:
` Back up the E-Sourcing database.
` Back up the FCI_HOME/config/*.properties files.
Use your organization’s standard backup procedures as defined by your corporate backup policies.
` For information about backup and restore for NetWeaver Application Server for Java, see the
SAP Help Portal at help.sap.com: SAP NetWeaver-->SAP NetWeaver 7.0 (2004s)-->select
language-->SAP Library-->SAP NetWeaver Library--> Administrator's Guide-->Technical
Operations Manual for SAP NetWeaver--> Administration of SAP NetWeaver Systems-->AS
Java (Application Server for Java)--> Management--> Tasks-->Backing up and Restoring AS
Java.
` For information about backup and restore for other application servers, see the documentation
for your application server.
In addition to these backups, ensure that you have access to the installation media in the event that
you need to recover.
The following is a sample backup strategy for E-Sourcing:
` Databases are real-time mirrored for high availability.
` Nightly database backups are performed.
` Weekly database backups are sent off-site and stored for one month.
` Monthly database backups are sent off-site and stored for one year.
` Application configuration files are backed up and stored nightly.
` Application log files are backed up nightly and are stored for at least 90 days.
Component Copy
Component copy functionality is not natively present in the E-Sourcing application. Promoting
objects from system to system is typically handled through our Object Migration capabilities. See
the Object Migration appendix of this guide for further details.
Restore Process
A database administrator should perform the following procedures to restore the E-Sourcing
system in the event that this is necessary:
1 Restore the database backup using the standard processes defined by your database vendor.
2 Reinstall the E-Sourcing application on an application server that is configured to run
E-Sourcing. Refer to the appropriate sections of this guide for installation and configuration
details.
3 Restore the fcisystem.properties and fcilocal.properties files from backup and place them in
the FCI_HOME/config directory.
4 If the database connection information has changed, use the FCI_HOME/bin/configure utility
to establish connectivity to the new database.
5 Start up and access the E-Sourcing application.
Load Balancing
For details about load balancing in E-Sourcing, see the E-Sourcing Online Help: Setup --> System
Setup --> Configuration --> Cluster Configuration.
Confidential 69
Overview Appendix E: Object Migration
Overview
This document provides the reader with a detailed description on the capability within SAP
E-Sourcing known as Object Migration. The object migration technology in E-Sourcing is a
capability that allows businesses to easily move master data from one instance of E-Sourcing to
another instance by exporting specified data from one installation and importing that data into a
second installation.
It is assumed that the reader of this document is familiar with E-Sourcing and how it is typically
deployed by customers.
Introduction
The object migration capabilities are accessed from the Setup > System Administration >
Import/Export Tools area within E-Sourcing. Import Data is used to load data into E-Sourcing and
Export Data is used to extract data from E-Sourcing.
The object migration methodology is based on the concept that one or more business objects within
E-Sourcing will be moved between systems. Such a model necessitates that the user have a sound
understanding of the business objects and their relationships within E-Sourcing, but does not
require that database table structures be understood. For example, E-Sourcing represents an
instance of a Vendor as a business object. While the data associated with a vendor may span many
database tables within the E-Sourcing schema, the user need only know that they want to export a
Vendor business object.
Object migration packages may be executed multiple times by adding to the Exports list. However,
the instructions within an object migration package can not be changed after the first export has
been performed.
Object Dependencies
Most business objects within E-Sourcing follow typical relational constructs. That is, a business
object may be the parent of many subordinate objects (children) and may also refer to other
business objects, defined elsewhere in the system.
For example, a fairly simple object in E-Sourcing is the Vendor. The vendor object consists of
header level data such as the vendor name and address. The vendor object also has subordinates;
the list of Internal Categories, for example, is a collection contained within the vendor object. An
example reference to another business object is the actual Internal Category which is referenced by
each element in the internal category collection. A good rule of thumb is that if the data is
maintained while editing the object, it is part of that object. If, however, the data is maintained on a
separate page, as is the case with internal categories, the object is a reference.
As a rule, when an object is exported, only the object itself and its children will be exported. The
referenced objects must be migrated separately. Therefore, users of the object export capability
must be careful to consider inter-object dependencies when building export packages.
For example, if a new vendor is created and it refers to a new internal category, both the vendor and
category must be migrated to the target system. If, however, a new vendor is created that refers to
an existing internal category, only the vendor need be migrated.
Failure to include any referenced objects in the export package which do not already exist on the
target system will result in the subsequent import failing.
More information is provided below on common object dependencies.
Alterations to the external id data member are not recommended as such changes could have
undesirable effects on the object migration. Such changes could result in creation of new objects
when that was not the intent. Failures can also occur if other members of the object represent the
unique key, which may not be getting changed as part of the creation, and therefore, could cause a
unique key constraint error.
Import Data
After data is exported, the OMA file may be downloaded to a computer and imported into another
instance of E-Sourcing. Users need only access the Import Data tool and upload the OMA file. It is
generally recommended that the import be run in the background and progress monitored by
refreshing the page.
Importing data requires that objects be created in a proper order so that dependencies between
objects are properly handled. The E-Sourcing import data tool handles this for the user as long as
all of the objects are contained in a single OMA file. That is, use of a single export package and,
therefore, a single import package enables E-Sourcing to properly handle dependencies between
objects. Use of multiple export packages and, therefore, multiple imports requires that the user
ensure that objects are imported in the proper order, which is strongly discouraged.
Introduction
The use of a “partial object migration” strategy is recommended, in which only selected objects are
migrated from one system to another. Such a strategy requires that developmental changes occur in
a STAGING system where the users can freely modify and test changes without impacting
PRODUCTION users. At the time of release (when the changes will be promoted to the
PRODUCTION system for use), an object migration should be performed which moves only those
objects required for the release.
These practices allow E-Sourcing users to plan, develop, and promote releases of application
functionality, using traditional project management techniques.
Environments
Most E-Sourcing customers deploy two environments: STAGING and PRODUCTION. Under this
configuration, development is performed in the STAGING system and objects are exported from it
and loaded into PRODUCTION.
Customers that also deploy a TEST environment may utilize it to validate the object migration
packages. That is, before moving objects from STAGING to PRODUCTION, they can be exported
from STAGING and loaded into the TEST environment. This approach provides an additional
check in that it allows users to confirm that the contents of the export package are complete and that
they will import into the PRODUCTION system without error (NOTE: this is based on the
assumption that the STAGING and TEST environments were initially set up with a copy of the
PRODUCTION database).
Tracking Changes
In order for the partial object migration to be successful, changes to the STAGING system must be
tracked. It is recommended that one user (the E-Sourcing system administrator) be charged with
maintaining the list of changes that will be included in a release package. The system administrator
should work with all E-Sourcing developers and keep a list with the following information:
` The object type being migrated (e.g., Query, Query Group, Project Template, etc.)
` The name of the object.
` A brief description of the change (or a reference to a requirements document that describes the
change)
This information will be used by the system administrator to develop the final export package for
the release.
Naming Conventions
The use of naming conventions, particularly for lower level objects such as localized resources and
query definitions is a valuable way to ensure that the proper objects are migrated.
It is recommended that all localized resources added to the STAGING system for customization
purposes utilize a bundle named custom. This technique simplifies the object migration strategy by
allowing the entire custom bundle to be exported, without a need to track each individual localized
resource change.
It is also recommended that custom queries and query groups have an internal name that begins
with CUSTOM-. Although it is recommended that individual queries and query groups be migrated
using the sourcing object list technique, naming objects this way simplifies searches and query
management.
Export Data
In most cases, object migration exports should work without error. If errors are encountered during
the export, the E-Sourcing log files (accessed from the System Information area of the Setup page)
should be consulted. Contact SAP Technical Support for assistance in reviewing the logs.
Import Data
Importing data errors most commonly occur because dependent objects do not exist in the system
to which the data is being imported. For example, if a new RFx Template is imported into a system
that depends on a new RFx Question Library that is not imported, then an error will be reported in
the trace file for the import.
Dependency errors such as the above can be corrected by re-creating the export package with all of
the required objects and performing the import another time.