atg search installation and configuration guide - oracle
TRANSCRIPT
Search
Version 10.1.2
Installation and Configuration Guide
Oracle ATG
One Main Street
Cambridge, MA 02142
USA
ATG Search Installation and Configuration Guide
Product version: 10.1.2
Release date: 12-17-12
Document identifier: SearchInstallConfigGuide1404301402
Copyright © 1997, 2012 Oracle and/or its affiliates. All rights reserved.
Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of their respective owners.
Intel and Intel Xeon are trademarks or registered trademarks of Intel Corporation. All SPARC trademarks are used under license and are
trademarks or registered trademarks of SPARC International, Inc. AMD, Opteron, the AMD logo, and the AMD Opteron logo are trademarks or
registered trademarks of Advanced Micro Devices. UNIX is a registered trademark of The Open Group.
This software and related documentation are provided under a license agreement containing restrictions on use and disclosure and are
protected by intellectual property laws. Except as expressly permitted in your license agreement or allowed by law, you may not use, copy,
reproduce, translate, broadcast, modify, license, transmit, distribute, exhibit, perform, publish, or display any part, in any form, or by any
means. Reverse engineering, disassembly, or decompilation of this software, unless required by law for interoperability, is prohibited.
The information contained herein is subject to change without notice and is not warranted to be error-free. If you find any errors, please
report them to us in writing.
If this software or related documentation is delivered to the U.S. Government or anyone licensing it on behalf of the U.S. Government, the
following notice is applicable:
U.S. GOVERNMENT END USERS:
Oracle programs, including any operating system, integrated software, any programs installed on the hardware, and/or documentation,
delivered to U.S. Government end users are "commercial computer software" pursuant to the applicable Federal Acquisition Regulation and
agency-specific supplemental regulations. As such, use, duplication, disclosure, modification, and adaptation of the programs, including any
operating system, integrated software, any programs installed on the hardware, and/or documentation, shall be subject to license terms and
license restrictions applicable to the programs. No other rights are granted to the U.S. Government.
This software or hardware is developed for general use in a variety of information management applications. It is not developed or intended
for use in any inherently dangerous applications, including applications that may create a risk of personal injury. If you use this software or
hardware in dangerous applications, then you shall be responsible to take all appropriate fail-safe, backup, redundancy, and other measures
to ensure its safe use. Oracle Corporation and its affiliates disclaim any liability for any damages caused by use of this software or hardware in
dangerous applications.
This software or hardware and documentation may provide access to or information on content, products, and services from third parties.
Oracle Corporation and its affiliates are not responsible for and expressly disclaim all warranties of any kind with respect to third-party
content, products, and services. Oracle Corporation and its affiliates will not be responsible for any loss, costs, or damages incurred due to
your access to or use of third-party content, products, or services.
The software is based in part on the work of the Independent JPEG Group.
ATG Search Installation and Configuration Guide iii
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Audience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Document Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
More Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. ATG Search Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
How Oracle ATG Web Commerce Search Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Oracle ATG Web Commerce Search Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Security in Oracle ATG Web Commerce Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3. Planning Your Oracle ATG Web Commerce Search Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Questions for Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Defining Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Oracle ATG Web Commerce Search Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Architecture for Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Architecture for Production Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Oracle ATG Web Commerce Search Installation Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Oracle ATG Web Commerce Search Installation Roadmaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Oracle ATG Web Commerce Search Evaluation Roadmap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Oracle ATG Web Commerce Search with Oracle ATG Web Commerce, ATG Content
Administration, and Oracle ATG Web Commerce Merchandising . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Oracle ATG Web Commerce Search with Oracle ATG Web Knowledge Manager and Oracle
ATG Web Knowledge Manager Self Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4. Installing Oracle ATG Web Commerce Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Installation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Installing HTMLFilter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Installing PDF Extract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Adding Search Engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Configuring Remote Indexing Engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Configuring Search with CIM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Configuring a Multi-Server Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Configuring Search Routing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Configuring the DeployShare Directory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Configuring the Lock Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Configuring the IDGenerator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Configuring the IndexingOutputConfig Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Configuring Customization Adapters for Search Merchandising . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Configuring SearchSQLRepository Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Indexing Repositories Managed by ATG Content Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Configuring Remote Routing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Configuring Data Sources for Oracle ATG Web Commerce Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Creating the Oracle ATG Web Commerce Search Database Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Creating Base Oracle ATG Web Commerce Search Database Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Importing Initial ATG Search Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Creating Database Tables for Oracle ATG Web Commerce Search with Multisite . . . . . . . . . . . . . . . . . . . . . 43
Creating Database Tables for Oracle ATG Web Commerce Search with Oracle ATG Web
Commerce Merchandising . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Deleting Oracle ATG Web Commerce Search Database Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Accessing Search Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
A. Oracle ATG Web Commerce Search Component Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5. Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
1 Introduction 1
1 Introduction
This guide explains how to install and configure Oracle ATG Web Commerce Search, and how to integrate your
Search installation with other Oracle ATG Web Commerce applications. It is written for system administrators
and other experienced persons involved in planning and performing a Search installation. Review this guide
thoroughly before beginning your installation.
This chapter includes the following sections:
Audience (page 1)
Document Conventions (page 1)
More Information (page 2)
Audience
This manual is intended for system and site administrators responsible for installing and configuring Oracle ATG
Web Commerce Search. To use this guide, you should be familiar with:
• Windows or UNIX administration
• Oracle ATG Web Commerce architecture
• Server environments and security infrastructure in your organization
Document Conventions
This guide uses the following conventions:
• <ATG10dir> refers to your Oracle ATG Web Commerce installation directory.
• <SearchEngineDir> refers to the directory where you have installed a standalone search engine.
2 1 Introduction
More Information
For more information on Oracle ATG Web Commerce Search, see the following:
• ATG Search Administration Guide—Administering your Search installation and indexing content.
• ATG Search Query Guide—Constructing search queries and handling the results.
• ATG Service Installation and Configuration Guide—Using Search as part of Oracle ATG Web KnowledgeManager.
2 ATG Search Overview 3
2 ATG Search Overview
Oracle ATG Web Commerce Search is an industry-leading search engine that lets users zero in on the information
they need, regardless of format, location, or language. Multiple levels of analysis pinpoint answers and deliver
them with subsecond response times.
Search features include:
• Natural language processing of end-user queries
• Multiple ways to find information (browse, keyword, etc.)
• Support for faceted navigation in client applications
• Multiple language support
• Industry-specific lexicons
• Automated indexing and categorization
• Predefined reports using Oracle ATG Web Commerce Business Intelligence
• Multisite support
Review this guide thoroughly before beginning your installation. This chapter includes the following sections:
How Oracle ATG Web Commerce Search Works (page 3)
Oracle ATG Web Commerce Search Components (page 4)
Security in Oracle ATG Web Commerce Search (page 6)
How Oracle ATG Web Commerce Search Works
“Searching” means retrieving information using an input query. Commonly, the information is textual, it is
retrieved from a collection of documents, and the queries are words or phrases entered by an end-user. The
search results are typically those documents most relevant to the query, plus some indication of why the
documents were retrieved. In order for searching to be efficient, the document collection is indexed by its terms
in a secondary storage component, typically called the index.
Oracle ATG Web Commerce Search generalizes documents and other content into an abstraction called an index
item. An index item can be an actual document such as an HTML file, a repository item such as an Oracle ATG
Web Commerce product, or a piece of structured data from a database. An index item consists of two elements:
4 2 ATG Search Overview
searchable text content and metadata. Metadata includes the title, summary, and other properties, and is used in
the following Commerce Search features:
• Faceted navigation
• Rank configuration
• Secondary sorting
The index items are fed into Search, which analyzes the content and stores a representation for each item in the
index. The index objects are organized into hierarchical sets much like a directory scheme. Search creates some
sets from the physical organization of the items, some sets from the metadata of the items, and other sets from
topic categorization results. These item sets enable users to search within subsets of the collection and browse
the collection without query input.
Text content is processed through the natural language components, which identify the structural elements
(such as sentences, headers, and tables) as well as the terms in the content. This processing is driven by a
dictionary and other language data, which are also stored in the index. The terms are divided into statement
vectors representing the sequence of terms in a structural portion of the content. The terms are also added to an
index, similar to one found in the back of a large book. This back-of-the-book index allows for efficient searches
and provides a global view of all the content.
Rather than simple text queries, Search accepts complex requests that specify what actions should be
performed. A request may include parameters, processing options, constraints, security settings, and other
information. The two primary requests are the Search Query and the View Item request. Search returns
responses that contain varied information depending on the type of request. For a Search Query, the response
contains a list of results plus other information to drive the search application or user interface.
For more information on how Search processes content into a searchable index, see the ATG Search
Administration Guide.
Oracle ATG Web Commerce Search Components
Oracle ATG Web Commerce Search has the following components:
• Search Administration—The interface where you create projects, index content, deploy indexes, etc. Runs
within a standard Oracle ATG Web Commerce installation. The SearchAdmin.AdminUI module is required to
run Search Administration. Your installation should include only one instance of Search Administration.
Project Administration is the part of the Search Administration interface dedicated to project-related tasks. The
search project contains information about the searchable content set you define, such as its location, what
rules are followed for its indexing, what customization data is applied. If you are using multisite, the project
also contains information on which content should be indexed and searchable from which sites.
Note: A search project bears no relationship to an ATG Content Administration project. Search projects do not
contain deployable assets or have an associated workflow, but are a means to organize and configure a given
set of indexable content.
The project links the index with search environments. Environments define the available resources for indexing
content and deploying completed indexes.You can also use projects to associate an index with customization
data. Some customization data affects how topic is indexed; other data is applied when the end-user submits
a query.
2 ATG Search Overview 5
When you install Search, a default environment is created automatically; this means that you don’t have to
create any environments before indexing your content for evaluation. You can add hardware resources to the
default environment, and create environments as required to handle end-user requests. See the ATG Search
Administration Guide for information on creating and configuring environments.
The Search Workbench is where you create and work with customization data and view information about
existing indexes. Customizations either add information to the index itself (such as additional dictionaries and
term weights) or affect the way your content is processed or accessed by end users (for example, Query Rules,
Topic Sets, and Facet Sets). Such customizations optimize your Search implementation for higher accuracy
and improved usability.
For information on Search Administration and customization data, see the ATG Search Administration Guide.
• Search engine—Engine instances can either index content or serve answers to end-user queries. The
DAF.Search.Routing module starts engine instances in the environments associated with a project as
needed.
Search engines do not require an Oracle ATG Web Commerce instance (and are therefore also called
“standalone” engines). Your installation can have any number of search engines, but should run only one per
CPU. ATG recommends at least two cores per engine. A single search engine requires at least 1.5GB of memory
on a 32-bit operating system.
• Routing—The DAF.Search.Routing module starts search engines and coordinates communication among
all search components. This module is included in Search Administration and in client applications.
• Client application—Interface through which your users place their search queries, such as an Oracle ATG Web
Commerce site. This application must include the DAF.Search.Routing and DAF.Search.Query modules.
• Search index—The searchable content deployed on your sites. An index is composed of one or more logical
partitions, each of which is associated with content configured in Search Administration. Each logical partition
is composed of one or more physical partitions. Each physical partition requires a search engine to serve
answers from that partition. Each engine has its own copy of the physical partition it serves.
Search Index Concepts Related
6 2 ATG Search Overview
• Search database—Oracle ATG Web Commerce Search uses the following three repositories:
• Routing—Stores search engines, index structure, and index deployment information.
• Search Administration—Stores other Search Administration data.
• IncrementalItemQueue—Stores incremental indexing data.
Search Administration requires access to all three repositories. Your client application requires access to the
IncrementalItemQueue and the Routing repository.
See the Oracle ATG Web Commerce Search Architecture (page 8) section for detailed information on how
these components can be arranged in your installation.
Security in Oracle ATG Web Commerce Search
Oracle ATG Web Commerce Search includes the Search Access role, which allows a user assigned that role to
access Search Administration and the Workbench. This role is granted by default to the admin user; in Oracle
ATG Web Knowledge Manager, by default it is also assigned to the service user.
The role can be found in the SearchAdminRoles folder of the Oracle ATG Web Commerce Business Control
Center. For information on using the Business Control Center, see the ATG Business Control Center Administration
and Development Guide.
3 Planning Your Oracle ATG Web Commerce Search Installation 7
3 Planning Your Oracle ATG Web
Commerce Search Installation
This chapter introduces installation and configuration procedures for Oracle ATG Web Commerce Search. Review
this chapter thoroughly before beginning your installation. It includes questions that you must consider when
planning your installation, and roadmaps that provide high-level outlines of the steps required to install and
configure Search with other Oracle ATG Web Commerce products.
This chapter includes the following sections:
Questions for Planning (page 7)
Defining Servers (page 8)
Oracle ATG Web Commerce Search Architecture (page 8)
Oracle ATG Web Commerce Search Installation Prerequisites (page 10)
Oracle ATG Web Commerce Search Installation Roadmaps (page 11)
Questions for Planning
Before you start installing Oracle ATG Web Commerce Search, consider the following questions:
• What Oracle ATG Web Commerce products do you plan to use along with Search? It’s a good idea to review all
of the applicable documentation before beginning any installation.
• Do you plan to use multisite? This decision does not affect your Search installation, but may affect other
aspects of your installation planning. See the ATG Multisite Administration Guide for information on multisite
planning.
• How will you apportion your hardware? Indexing and searching are resource-intensive functions that benefit
strongly from dedicated hardware with robust specifications.
• How will your applications be distributed? Some applications can share an Oracle ATG Web Commerce server
instance, while others cannot. Will your system include clusters? See the ATG Multiple Application Integration
Guide for architectural information.
• How will administrative functions such as lock management be handled? With a dedicated Oracle ATG Web
Commerce server instance, or as part of some other function?
8 3 Planning Your Oracle ATG Web Commerce Search Installation
• What database software will you use? Where will the Oracle ATG Web Commerce platform and application
schemas be located? How will they be accessed?
• Where will your Search deployment share be located? The deployment share is a single, scalable, shared
directory where master copies of indexes are stored. Your installation can have only one deployment share,
which should be located on a dedicated high-performance disk, separate from the Search Administration
installation. The Search Administration installation and all search engines must have access to this directory.
Note: The size of the index that is deployed does not bear any direct relationship to the size of the raw
information being indexed—dictionaries, topics, and other customization data can all add to the size of the
index, as can the nature of the content itself. For example, content consisting mostly of pictures with some
metadata might form a very small index relative to the raw content size, while a product catalog with many
small, unique pieces of information might be relatively large.
Defining Servers
The term server can be confusing due to its multiple meanings. In this guide, server refers to an Oracle ATG
Web Commerce application instance that is deployed to your application server in the form of an EAR file. Each
server can (and usually does) include multiple Oracle ATG Web Commerce products and a variety of modules
depending on the server’s purpose. Conversely, one product may require multiple servers.
You can create any number of server instances based on a single physical installation of Oracle ATG Web
Commerce. Server instances are a way to provide server-specific configuration information without undue
duplication; see the ATG Installation and Configuration Guide for information if you are not familiar with this
feature.
To reduce confusion, application servers and Oracle ATG Web Commerce servers are always specified as such in
this guide, and the term server never refers to hardware.
Oracle ATG Web Commerce Search Architecture
The purpose of this section is to outline the principles involved in the Oracle ATG Web Commerce Search
architecture and enable you to make the best decision for your sites. The examples that follow deal solely
with search components; everything else is represented as the “Client Application.” For diagrams that show
Search integrated with other Oracle ATG Web Commerce products, see the Oracle ATG Web Commerce Search
Installation Roadmaps (page 11). In the diagrams that follow, the light gray boxes represent Oracle ATG Web
Commerce servers, which can be, but are not necessarily, analogous to hardware resources. Dark gray boxes,
when present, represent standalone search engines.
Architecture for Evaluation
This architecture runs all of your search components within a single Oracle ATG Web Commerce instance. This is
sufficient for basic evaluation purposes and for estimating the size of your final architecture, but should never
be used for production.
3 Planning Your Oracle ATG Web Commerce Search Installation 9
A Self-Contained Search Installation
You may want to use this setup to estimate the size of your index using Search Administration, and then add
resources to support that index. The original installation can then be used solely for Search Administration.
Architecture for Production Sites
In a production environment, Search Administration and the client application run in separate Oracle ATG
Web Commerce instances, on separate machines, with two instances of the latter to provide redundancy.
Communication between instances is handled by the DAF.Search.Routing module. Note that this module
must be able to communicate with the Search database.
This recommended architecture uses the Search Administration instance for indexing content, with several
standalone engines to serve answers to end-user queries. The dark gray box in the lower right represents any
number of hosts you might assign to one or more production environments, running as many search engines as
your hardware will support.
The diagram also includes Oracle ATG Web Commerce Business Intelligence. Note that your data loaders must
run in a dedicated instance, separate from your application and from Search.
10 3 Planning Your Oracle ATG Web Commerce Search Installation
Application Redundancy and Dedicated Engine Resources
If your content set is large enough, serving answers requires more than one search engine, each of which should
have a dedicated CPU and at least two cores per engine. An index is composed of one or more logical partitions,
each of which is associated with a content set configured in Search Administration. Each logical partition is
composed of one or more physical partitions. Each physical partition is served by a search engine. Each engine
has its own copy of the physical partition it serves.
Oracle ATG Web Commerce Search Installation
Prerequisites
This section addresses prerequisites and preliminary configuration procedures that must be followed before
installing Oracle ATG Web Commerce Search.
• If you are installing on AIX and must run Search using sudo, note that for security reasons, AIX removes
the LIBPATH environment variable when using sudo. To resolve this potential problem you can either set
LIBPATH as root before running Search, or set the LIBPATH in the application server’s startup script.
Note that running any Oracle ATG Web Commerce application as root is not recommended. If at any time you
run Search as root or using sudo, the file permissions in the installation change to root. Therefore, you must
either continue running Search as root or using sudo, or reset the file permissions.
3 Planning Your Oracle ATG Web Commerce Search Installation 11
• If you are installing on a UNIX or Linux variant, you must increase the number of open files permitted. The
default setting is too low to support high network traffic, and can result in a “too many open files” error under
load. Symptoms include Routing having problems communicating with the engine or logging an error.
Set the number of open files permitted to 10,000 or higher on all machines running Search, including the
Routing instances and remote search engines.
• If you are installing on a UNIX variant (such as Solaris or AIX) and there is any chance that your content
includes filenames with non-English characters, make sure that your system locale is set to use UTF-8
encoding.
Note: If you are using multi-byte languages, Search Administration may incorrectly calculate length limits for
fields such as content set names, based on your database capabilities. If your database limits the number of
bytes stored in string columns, set the maxLengthEncoding property of the /atg/searchadmin/adminui/
validator/Validator component to your database encoding. The default setting is UTF-8. If your database
correctly limits string column length based on character encoding, you can set the property to null or leave it
empty.
• For any installation in which you are planning to run Search Administration, before you install Search, install
your application server and the Oracle ATG Web Commerce platform. If you are only planning to run a
standalone search engine, you do not need the platform or an application server.
• You will need a scalable, shared network drive where indexes are initially created and from which they can be
retrieved by the search engines. This directory is called the deployment share. All search engines and Search
Administration instances must have access to this directory. Indexes can take up a great deal of disk space, so
be sure the directory selected has sufficient space (at least 100 GB, more if your index is large or you plan to
archive old indexes). This directory must be writable. In a production environment, this directory should be
located on a dedicated, high-performance machine.
Note: The deployment share must be created and configured as a shared folder prior to installing Search.
Oracle ATG Web Commerce Search Installation Roadmaps
The sections that follow outline the tasks involved in setting up Oracle ATG Web Commerce Search in a variety
of configurations. Each roadmap consists of a step and cross-references to the detailed information required to
perform it.
In many cases, configuration tasks can be performed using the CIM configuration tool. Using CIM ensures that
you have not omitted steps and that you have configured your databases and components correctly.
Oracle ATG Web Commerce Search Evaluation Roadmap
The goal of this roadmap is to install and configure an Oracle ATG Web Commerce Search installation that does
not integrate with any other applications, but allows you to run Search Administration and index content for
evaluation purposes.
Before you begin:
• Make sure you have your database information at hand, and that any necessary drivers are installed.
• Select a Search Deployment Share directory. See Configuring the DeployShare Directory (page 30) in this
guide. In a self-contained installation, this can be any local directory with enough room for your index.
12 3 Planning Your Oracle ATG Web Commerce Search Installation
To set up a search environment for evaluation:
1. Install your application server. See the ATG Installation and Configuration Guide for information on configuring
your application server.
2. Download and install Oracle ATG Web Commerce products:
• Oracle ATG Web Commerce platform; select all components when installing. See the ATG Installation and
Configuration Guide.
• Oracle ATG Web Commerce Search; select the Full Search Install option. See Installing Oracle ATG Web
Commerce Search (page 23).
• Optional HTML and PDF indexing components. See Installing HTMLFilter (page 24) and Installing PDF
Extract (page 24).
3. Create the database tables. See Configuring Search with CIM (page 27), or see Creating the Oracle ATG
Web Commerce Search Database Tables (page 41) in this guide for manual table creation steps.
4. Assemble and deploy an Oracle ATG Web Commerce application EAR file. See Configuring Search with
CIM (page 27). If using a manual assembly process, include the SearchAdmin.AdminUI module in the
EAR. See the Assembling Applications section of the Developing and Assembling Nucleus-Based Applications
chapter in the ATG Platform Programming Guide for details on application assembly.
5. Start your Oracle ATG Web Commerce application, log in, and navigate to Search Administration. See
Accessing Search Administration (page 43) in this chapter.
You can now create a search project, index content, and create customization data as described in the ATG
Search Administration Guide. Note that without a client application, you will not be able to search your index.
Oracle ATG Web Commerce Search with Oracle ATG Web Commerce, ATG Content
Administration, and Oracle ATG Web Commerce Merchandising
The goal of this roadmap is to install and configure Oracle ATG Web Commerce Search in an integrated
environment with Oracle ATG Web Commerce, ATG Content Administration, and Oracle ATG Web Commerce
Merchandising. It also includes the use of standalone search engines.
You will need three server instances for this product combination:
• One runs the administrative interfaces, including Oracle ATG Web Commerce Merchandising and Search
Administration; this is referred to as the asset management server in this roadmap.
• The second runs your Oracle ATG Web Commerce sites for testing, and is referred to as the staging server.
• The final instance runs your live sites, and is referred to as the production server.
The servers can be set up on separate machines, or can run as separate instances on one machine if it has plenty
of memory. Do not use the latter option for live sites, but you may use such a configuration for evaluation or
testing purposes. You may also want to dedicate resources for additional search engines.
Your configuration may also include a staging and preview server.
Some of the steps in this procedure can be performed using CIM. These steps are noted in the roadmap. See
Configuring Search with CIM (page 27) and the CIM online help for additional information.
3 Planning Your Oracle ATG Web Commerce Search Installation 13
Architecture Diagram
This section provides a visual explanation of the pieces involved in an installation. For each server, the diagram
shows key modules and other components.
The diagram includes the following databases, each of which includes data for several repositories:
• Management—Versioned data for ATG Content Administration and other non-catalog information.
• Production—Catalog and other deployed asset information for your sites. Use the switching database
feature described in the ATG Content Administration Programming Guide. If you use switching, the production
database becomes in actuality three databases, two containing switched data and a third containing
information that does not benefit from switching, such as user profiles and ATG Search.
• Staging—Duplicates the part of your production database that is necessary for staging and testing. In this
configuration, that includes at least your product catalog and refinements.
If you use Oracle ATG Web Commerce Business Intelligence, you also have a data warehouse, not included in this
diagram. See the ATG Business Intelligence Installation and Configuration Guide for reporting information.
14 3 Planning Your Oracle ATG Web Commerce Search Installation
Servers, Products, and Databases for Search and Commerce
Basic Installation Steps
This section covers the actual installation of the components and database configuration.
1. Install your application server on all machines that will be used for Oracle ATG Web Commerce instances. See
the ATG Installation and Configuration Guide for information on configuring your application server.
Note: If you are using standalone search engines, those do not need to have an application server.
2. Download and install Oracle ATG Web Commerce products. Some applications will be included in more than
one EAR file during deployment.
• Oracle ATG Web Commerce platform; select all components when installing
3 Planning Your Oracle ATG Web Commerce Search Installation 15
The platform includes the DAF.Search.Routing module, which will run on the production and staging
servers using local Routing (see the Oracle ATG Web Commerce Search Architecture (page 8) section).
• Oracle ATG Web Commerce Merchandising
• Oracle ATG Web Commerce Search; select the Full Search Install option. Install the HTML and PDF
components if needed (see Installing HTMLFilter (page 24) and Installing PDF Extract (page 24) in
this guide).
If using standalone search engines, see Adding Search Engines (page 25).
3. Configure databases and data sources for all applications. See Configuring Search with CIM (page 27), or
see the following documents for manual database configuration:
• Oracle ATG Web Commerce Platform. See Configuring Databases and Database Access in the ATG Installation
and Configuration Guide.
• Oracle ATG Web Commerce Search. See Creating the Oracle ATG Web Commerce Search Database
Tables (page 41) and Configuring Data Sources for Oracle ATG Web Commerce Search (page 40) in
this guide.
• Oracle ATG Web Commerce. See the Configuring and Populating a Production Database chapter of the ATG
Commerce Programming Guide. This includes creating tables, importing data, and configuring data sources
for both the publishing and target databases, and configuring a SwitchingDataSource if using one.
• ATG Content Administration. See Setting up the ATG Content Administration Database in the ATG Content
Administration Programming Guide.
• Oracle ATG Web Commerce Merchandising. See Configuring Database Tables in the ATG Merchandising
Administration Guide.
4. Create your asset management, staging, and production ATG servers. See Configuring Search with CIM (page
27), or see the ATG Platform Programming Guide for manual server creation information.
5. Configure lock management. See Configuring Search with CIM (page 27), or see Configuring the Lock
Manager (page 30) in this guide for manual lock management configuration.
Asset Management Server Configuration Steps
Configure the asset management server. This server runs ATG Content Administration, Oracle ATG Web
Commerce Merchandising, and Search Administration.
1. Configure the /atg/search/service properties for the administration server. See Configuring
SearchSQLRepository Components (page 35) in this guide.
2. Configure Oracle ATG Web Commerce components. See Configuring the IndexingOutputConfig
Component (page 31) and Configuring Customization Adapters for Search Merchandising (page 32) in
this guide.
3. Set up your asset management server. See the Setting Up an Asset Management Server chapter of the ATG
Content Administration Programming Guide.
Staging and Production Server Configuration Steps
Configure the production and staging servers.
1. Create a client application. See the ATG Commerce Programming Guide.
16 3 Planning Your Oracle ATG Web Commerce Search Installation
2. Configure your Oracle ATG Web Commerce Commerce client. See the ATG Search Query Guide.
Build and Deploy Applications
See Configuring Search with CIM (page 27), or see the Developing and Assembling Nucleus Applications
chapter of the ATG Platform Programming Guide for manual application assembly information. If you are
assembling manually, include the following modules:
• Asset management server EAR file
• DCS.Search.Versioned
• SearchAdmin.AdminUI
• DCS-UI.Search
• Publishing
• DCS-UI.Versioned
Note: If you are going to run Search Administration as part of a standalone EAR, you must do the following:
• Run the Oracle ATG Web Commerce Search installer and install a standalone search engine on the machine
to which you are going to deploy your EAR. See Adding Search Engines (page 25) in this guide.
• Set the engineDir property of the /atg/search/routing/Configuration component to the new
search engine’s <Searchdir>\SearchEngine directory. For example:
engineDir=c:\\ATG\\ATG10.1.2\\Search10.1.2\\SearchEngine
• Production and staging server EAR files
• DCS.Search
• DCS.PublishingAgent
• DAF.Search.Routing
• DafEar.Admin
Post-Deployment Configuration
The following tasks are performed in your running Oracle ATG Web Commerce applications.
1. Use ATG Content Administration to create your deployment topology, and initialize your deployment targets
by doing a full deployment. See the ATG Content Administration Programming Guide.
Note: If you are going to use refinement configurations (facets), include the RefinementRepository in your
repository mappings for your production and staging deployment targets.
2. After deploying your catalog from ATG Content Administration to the production server, make sure the
Product Catalog update and Catalog Maintenance services are run (this step is necessary in order to index
your catalog). See the ATG Commerce Programming Guide.
3. Use the Oracle ATG Web Commerce Business Control Center to access Search Administration. See Accessing
Search Administration (page 43) in this guide.
4. In Search Administration, create a search project and add content and any customizations, such as languages
and custom dictionaries. See the ATG Search Administration Guide. Use the following settings when you add
the content:
3 Planning Your Oracle ATG Web Commerce Search Installation 17
• Content Type—ATG Repository
• IndexingOutputConfigPath—/atg/commerce/search/ProductCatalogOutputConfig
• Location—Remote
Set the Host Machine and Port as appropriate for the content location.
At this point, you can use the default environments to estimate the size of your index and evaluate whether
you have dedicated sufficient resources for indexing and answer serving.
Note: A search project bears no relationship to an ATG Content Administration project. The search project is
a persistent item that defines the content you want to include in your index, what type of index you want to
create and any customization data used during indexing. Search does not use workflows.
5. If using standalone search engines, add those hosts to your search environment. See the ATG Search
Administration Guide. Search environments associate a project with the physical resources that project uses to
index content and serve answers.
6. Index your catalog, using the “Full” indexing option.
7. Configure Oracle ATG Web Commerce Merchandising as described in Structuring the Search Configuration Tree
in the ATG Merchandising Administration Guide.
Testing Your Installation
To test your installation, you can use the Search Query Console (see the Sample Application section of the ATG
Search Query Guide) or Oracle ATG Web Commerce Merchandising (see the ATG Merchandising Guide for Business
Users). Use the following steps:
1. Deploy your content.
2. Index the content.
3. On your production sites, make sure you get search results
4. Use Merchandising to change the catalog.
5. Deploy the changes.
6. Reindex the content.
7. Verify the change on your production sites.
Oracle ATG Web Commerce Search with Oracle ATG Web Knowledge Manager and Oracle
ATG Web Knowledge Manager Self Service
The goal of this roadmap is to install and configure Oracle ATG Web Commerce Search in an integrated
environment with Oracle ATG Web Knowledge Manager and Oracle ATG Web Knowledge Manager Self Service.
You will need at least three servers for this product combination:
• One runs Search Administration and ATG Content Administration; this is referred to as the asset management
server in this roadmap.
• Another runs the Service Administration application and Oracle ATG Web Knowledge Manager, and is called
the agent server.
18 3 Planning Your Oracle ATG Web Commerce Search Installation
• A third server runs Oracle ATG Web Knowledge Manager Self Service, and is referred to as the production
server.
The servers can be set up on separate machines, or can run as separate instances on one machine if it has
plenty of memory (the latter option is not recommended for live sites, but you may use such a configuration for
evaluation or testing purposes). You may also want to dedicate resources for additional search engines.
Architecture Diagram
This section provides a visual explanation of the pieces involved in an Oracle ATG Web Commerce Search
installation with Oracle ATG Web Knowledge Manager. In addition to the three servers, the diagram includes the
following databases:
• Management—Versioned data for ATG Content Administration, Search Administration, and other non-catalog
information.
• Production—Catalog and other deployed asset information for your sites. Use the switching database feature
described in the ATG Content Administration Programming Guide. If you are using switching, the production
database becomes in actuality three databases, two containing switched data and a third containing
information that does not benefit from switching, such as user profiles.
• Agent—Versioned and non-versioned repositories including user profiles, internal and external user
segments, solutions, content groups, Search Routing, ATG Ticketing, and topics.
If you are using Oracle ATG Web Commerce Business Intelligence, you will also have a data warehouse, not
included in this diagram. See the ATG Business Intelligence Installation and Configuration Guide for an overview.
3 Planning Your Oracle ATG Web Commerce Search Installation 19
Search with Knowledge Manager
Basic Installation
This section covers the actual installation of the components and database configuration.
1. Install your application server on all machines that will be used for Oracle ATG Web Commerce instances. See
the ATG Installation and Configuration Guide for information on configuring your application server for ATG
products.
Note: If you are using standalone search engines, those machines do not need to have an application server.
2. Download and install Oracle ATG Web Commerce products.
• Oracle ATG Web Commerce platform; select all components when installing.
• Oracle ATG Web Commerce Search; select the Full Search Install option. Install the HTML and PDF
components if needed (see Installing HTMLFilter (page 24) and Installing PDF Extract (page 24) in
this guide).
• ATG Service; select ATG Service Administration and Oracle ATG Web Knowledge Manager.
If using standalone search engines, see Adding Search Engines (page 25).
3. Configure databases and data sources for all applications. Note that the documentation referenced is
extensive; review it carefully before proceeding.
• Oracle ATG Web Commerce Platform. See Configuring Databases and Database Access in the ATG Installation
and Configuration Guide.
• Oracle ATG Web Commerce Search. See Creating the Oracle ATG Web Commerce Search Database
Tables (page 41) and Configuring Data Sources for Oracle ATG Web Commerce Search (page 40) in
this guide.
• Oracle ATG Web Knowledge Manager. See Configuring Databases for ATG Service and Configuring the ATG
Service Servers in the ATG Service Installation and Configuration Guide.
4. Create your administration, agent, and production ATG servers. See Configuring Search with CIM (page
27), or see the ATG Platform Programming Guide for manual server creation information.
5. Configure lock management. See Configuring Search with CIM (page 27), or see Configuring the Lock
Manager (page 30) in this guide for manual lock management configuration.
Server Configuration Steps
Configure the administration server. This server runs Search Administration and ATG Content Administration.
1. Configure the /atg/search/service properties for the administration server. See Configuring
SearchSQLRepository Components (page 35) in this guide.
2. Run the create-atgservice-searchproject script against your ATG Search database as described in the
Configuring Search Environment section in the ATG Service Installation and Configuration Guide.
3. Set up your asset management server. See the Setting Up an Asset Management Server chapter of the ATG
Content Administration Programming Guide.
Create your client application, which runs on the production server.
20 3 Planning Your Oracle ATG Web Commerce Search Installation
Configure the agent server. Search Routing runs on this server. See Configuring the Agent Server in the ATG Service
Installation and Configuration Guide.
Build and Deploy Applications
See Configuring Search with CIM (page 27), or see the Developing and Assembling Nucleus Applications
chapter of the ATG Platform Programming Guide for manual application assembly information. If you are
assembling manually, include the following modules:
• Administration server EAR file
• DafEar.Admin
• SearchAdmin.AdminUI
• PubPortletService.ExternalUsers
Note: If you are going to run Search Administration as part of a standalone EAR, you must do the following:
• Run the Oracle ATG Web Commerce Search installer and install a standalone search engine on the machine
to which you are going to deploy your EAR. See Adding Search Engines (page 25) in this guide.
• Set the engineDir property of the /atg/search/routing/Configuration component to the new
search engine’s <Searchdir>\SearchEngine directory. For example:
engineDir=c:\\ATG\\ATG10.1.2\\Search10.1.2\\SearchEngine
• Production server EAR file
• DafEar.Admin
• ARF.Base
• Service.SelfService
• Agent server EAR file
• DafEar.Admin
• ARF.Base
• ARF.WSCP
• Service.Knowledge
• Service.admin
• Service.SampleCRMClient
• Service.CRMIntegration
Post-Deployment Configuration
The following tasks are performed in your running Oracle ATG Web Commerce applications.
1. Use ATG Content Administration to create your deployment topology, and initialize your deployment targets
by doing a full deployment. See the ATG Content Administration Programming Guide.
2. Use the Oracle ATG Web Business Control Center to access Search Administration. See Accessing Search
Administration (page 43) in this guide.
3 Planning Your Oracle ATG Web Commerce Search Installation 21
3. In Search Administration, add content to your project as described in the Configuring Search Environment
section in the ATG Service Installation and Configuration Guide.
4 Installing Oracle ATG Web Commerce Search 23
4 Installing Oracle ATG Web
Commerce Search
This chapter explains how to install Oracle ATG Web Commerce Search and its related components, and how to
configure your installation.
This chapter is intended to be used in parallel with the Search Roadmaps described in the previous chapter.
Before you perform any of the tasks described in this chapter, review the Planning Your Oracle ATG Web
Commerce Search Installation (page 7) chapter to be sure that you understand all of the required steps and
prerequisites.
This chapter includes the following sections:
Installation Process (page 23)
Installing HTMLFilter (page 24)
Installing PDF Extract (page 24)
Adding Search Engines (page 25)
Configuring Search with CIM (page 27)
Configuring a Multi-Server Installation (page 29)
Configuring Data Sources for Oracle ATG Web Commerce Search (page 40)
Creating the Oracle ATG Web Commerce Search Database Tables (page 41)
Accessing Search Administration (page 43)
Installation Process
To install ATG Search:
1. Download the Oracle ATG Web Commerce Search installation executable.
2. Start the installer.
3. On the start page, click Next to begin the installation.
4. Read the license agreement. Select “I accept the terms of the license agreement” and click Next.
24 4 Installing Oracle ATG Web Commerce Search
5. Select Search Administration + Search Engine for the installation type. Click Next.
Enter an installation location, or accept the default. Click Next.
6. Select locations for product icons and shortcuts.
7. Review your selections, and click Install to begin the installation.
8. Enter a Deployment Share folder. This is the scalable, shared directory you created before beginning the
installation.
9. When the installation is finished, click Done.
Installing HTMLFilter
If the content you want to make searchable includes Word files and other rich-text formats, download and install
the HTML Filter component for Oracle ATG Web Commerce Search.
1. Download the ATGSearchHtmlFilter executable.
2. Double-click the installer icon to start installing.
3. On the start page, click Next to begin the installation.
4. Accept the license agreement, and click Next.
5. Select an installation directory, or accept the default, and click Next. The directory must include a valid Search
installation.
6. Click Install to install the component.
No additional configuration is necessary to use the component.
Installing PDF Extract
If the content you want to make searchable includes documents in PDF format, download and install the PDF
Extract component for Oracle ATG Web Commerce Search.
1. Download the ATGSearchPDFExtract executable.
2. Double-click the installer icon to start installing.
3. On the start page, click Next to begin the installation.
4. Accept the license agreement, and click Next.
5. Select an installation directory, or accept the default, and click Next. The directory must include a valid Search
installation.
6. Click Install to install the component.
No additional configuration is necessary to use the component.
4 Installing Oracle ATG Web Commerce Search 25
Adding Search Engines
A search engine is a piece of software that performs the work of either indexing content or answering queries.
Engines are managed through Search Administration and are started and shut down automatically through a
service running on the engine location.
Note: A given partition on a given host should only have one search engine associated with it. In order to add
engines to support that partition, additional hardware is required.
The process of adding engines is most easily understood through an example.
Say that you have installed Oracle ATG Web Commerce Search and used the Search Administration’s estimation
tool to determine the size of your index. It is going to be large, with four physical partitions. You have two
options:
• Run Search Administration on hardware robust enough to support four engines for answering queries in
addition to performing indexing tasks.
• Add hardware resources that will be dedicated to answering queries.
The first option does not require any additional installation work to add the engines. In Search Administration,
you create an indexing environment and specify the machine on which you are running Search Administration
as the host. Engines will be created as needed on this machine, up to the number specified in the environment
configuration. This is not a recommended solution, however, since indexing and answer serving functions could
compete for resources.
Instead, you add two machines to your Search installation. Each has four cores, providing for a robust
answering configuration. The machine on which you are running Search Administration will continue to
handle indexing. You run the Search installer once on each machine, following the steps described below. In
Search Administration, you create a new environment, give it the type “production” and associate your two
new machines with that environment (see the ATG Search Administration Guide for detailed information on
environment configuration). When your sites are up and answering user queries, engine instances will be started
as needed, up to the limit configured on each machine.
The following diagram summarizes the procedures involved. This diagram uses indexing engines as an example;
to add engines dedicated to server query answers instead, you would create a new environment in Search
Administration (see Step 2 in the diagram).
Note: All search engines in your installation, including the indexing engine, must run on the same operating
system with the same bit size. Load is balanced automatically across the engines, based on their capacity.
26 4 Installing Oracle ATG Web Commerce Search
The Search Administration installation includes the search engine and a default environment, which can be
used for indexing. For anything other than a basic evaluation environment with a very small index, you will
need additional, standalone search engines on separate hardware. If you have multiple physical partitions in
your index, you need one search engine per partition. See the Estimating Index Size section of the ATG Search
Administration Guide for information on sizing your index.
These standalone search engines run on dedicated hardware; you must add hardware and install the search
engine on them. Engines are in turn associated with search environments. The environments are configured in
Search Administration.
Standalone search engines run without an Oracle ATG Web Commerce Search installation or application server.
To add a standalone search engine to your installation:
1. Download the Oracle ATG Web Commerce Search installation executable.
2. Double-click the installer icon to start installing.
3. On the start page, click Next to begin the installation.
4. Read the license agreement. Select “I accept the terms of the license agreement” and click Next.
5. Select Search Engine Only for the installation type.
6. Select an installation location.
7. Review your selections, and click Install to begin the installation.
8. On the machine where you have installed the standalone engine, navigate to the search engine installation
and execute the <SearchEngineDir>/SearchAdmin/bin/startRemoteLauncher.sh or .bat script.
The startRemoteLauncher script starts up a launcher service, which is then used by Routing to start search
engines on that machine. Under normal use, you do not need to manually start or stop search engines.
Note: This step is required only for standalone search engines. Engines running local to your Search
Administration installation do not need this script.
The startRemoteLauncher script includes two optional parameters –p and –o, which allow you to specify
the ports used for remote method invocation (RMI). For example:
startRemoteLauncher.bat -p 8860 -o 8861
If you do not specify a port, the second port is chosen randomly. The default for the first port is always 8860.
9. If you are adding engines for indexing purposes, see the Configuring Remote Indexing Engines (page 26)
section of this guide for additional steps.
In Search Administration, you can now create an environment that includes this machine. When Search indexes
content (if an indexing environment) or serves answers (if a staging or production environment), the Routing
module starts engine instances on that machine as needed. See the Managing Search Environments section of
the ATG Search Administration Guide for information on creating search environments.
When you deploy a new index, the search engines serving the old index continue to answer query requests until
new engines are started for the new index. Once the new engines are started, queries are directed to them by
the Routing module, and the old ones are shut down.
Configuring Remote Indexing Engines
If you are running search engine instances on machines other than the Search Administration machine, and you
want to use those instances in an indexing environment, additional configuration is required.
4 Installing Oracle ATG Web Commerce Search 27
In order to create a full index (see Note below) the indexing engine needs a clean partition, a file from which
all indexes are created. If the engine is not local to the Search Administration installation, you must make it
available. The clean partition is located in the <Searchdir>\SearchEngine\operatingsystem\data\
directory.
To make a clean partition available:
1. Either move the clean partition into your deployment share folder, or create a new share with read access for
the search engine (for example, \\mymachine\sharedData).
2. Create a Configuration.properties file in your <ATG10dir>/atg/search/routing directory
to point to the partition in the shared directory, using the full network path. The file should contain a
cleanPhysicalPartitionPath property as shown in the following examples.
Windows:
cleanPhysicalPartitionPath=\\\\server\\deploymentshared\\initial.index
UNIX:
/mount_location/deploymentshared/initial.index
The search engines must also be able to access all directories that contain content to be indexed. When indexing
a file system, the search engines receive paths to the files, for example, D:\CustomerDocs\whitepaper.doc.
If the search engine is not on the same machine as Search Administration, then the D drive is not accessible, and
the engine cannot open the file to index it. In that case, you should share the content location, for example as \
\some_machine\CustomerDocs.
Note: Even if you use incremental indexing, you should recreate the full index every so often; see the ATG Search
Administration Guide for indexing suggestions.
Configuring Search with CIM
This section provides step by step instructions for using CIM to configure your ATG Search installation.
CIM greatly reduces the chance for configuration errors. It configures datasource components, calculates which
database scripts to run in what location and what order, determines which data imports are required, configures
lock managers, and creates the required server types for the selection of products have you have installed.
For general instructions on using CIM, see the ATG Installation and Configuration Guide. CIM dynamically
generates menus based on which configuration steps you have performed and which products you have
installed.
Note that you will need connection information for your database (usernames, passwords, driver location, etc.)
before performing the database configuration steps.
1. Start CIM.
2. From the Main Menu, select Product Selection.
3. From the Product Selection menu, choose Search and any other products you want to configure. Only
products you have installed can be selected.
4. From the Commerce Addons menu, select any additional features to configure, such as dedicated lock servers
or staging servers.
28 4 Installing Oracle ATG Web Commerce Search
Additional steps not related to Search may follow, depending on your selected addons.
5. Specify which server should include the optional QueryConsole tool, if you plan to use it. See the Sample
Application section of the ATG Search Query Guide for information.
6. Select Index by Product or Index by SKU. See the ATG Search Administration Guide.
Additional steps not related to Search may follow, depending on your selected addons.
7. From the Main Menu, select Datasource Configuration. CIM guides you through the process of configuring
the datasources needed to communicate with the schemas required by your product combination. Follow
the prompts to configure your datasources.
• Select database type
• Enter the use name for the database connection.
• Enter the password for the database connection.
• Enter the hostname of the machine on which the database is installed.
• Enter the port number used to connect to the database.
• Enter the database name.
• CIM automatically creates the database URL. Check it for errors before proceeding.
• Enter the path to your database driver.
• CIM automatically assigns a JNDI name to the datasource. You can accept the default or provide your own
JNDI name.
8. From the Configure Datasource menu, select Test Connection to ensure that your datasource is correctly
configured.
9. From the Configure Datasource menu, select Create Schema.
10.From the Configure Datasource menu, select Import Initial Data.
11.From the Main Menu, select Server Instance Configuration. CIM automatically determines which server types
are required. Perform the steps required under general configuration.
For a publishing server, specify the hostname for the Search instance, and the RMI port it will use to connect
to either the staging server or production server.
Note: Server instance types represent publishing, asset management, etc. server types. Server instances
represent individual servers of that type.
12.For each server instance type, configure at least one server instance. For many installations, you only need
one instance of each type.
13.Perform any non-CIM configuration tasks as specified by the relevant roadmap.
14.Use CIM to assemble and deploy your applications. From the Main Menu, select Application Assembly and
Deployment. CIM walks you through creating EAR files based on your server instances and deploying them to
the application server.
Note that CIM does not configure the following:
4 Installing Oracle ATG Web Commerce Search 29
• Standalone search engines.
• Any components for a staging environment to be used with Search.
• Your ATG Content Administration topology. See the Setting Up an Asset Management Server chapter of the ATG
Content Administration Programming Guide.
• Your IndexingOutputConfig component. CIM does handle some configuration options, such as whether
you plan to index by product or by SKU, but you will most likely have to do additional configuration.
• Automatic startup for the ProductCatalogOutputConfig components on either production or staging
servers. You must manually configure these components to start automatically.
• Configuration for deployment-triggered indexing (see Synchronizing Versioned Repository Deployment and
Indexing (page 39)). You must manually configure these components.
• Your search project, indexing rules, and customizations. See the ATG Search Administration Guide.
• Additional search environments. CIM automatically adds a /atg/commerce/search/catalog/
QueryRequest.properties file with the following configuration:
targetName=Staging
However, in order to use the target, you must manually add the appropriate search environments.
Configuring a Multi-Server Installation
The Oracle ATG Web Commerce Search installer assumes that all Search modules are running locally, as with the
evaluation example in the section; in real use, this will never be the case. This section of the guide addresses
configuration that must be done in order for Search to communicate with remote search components, or
with other Oracle ATG Web Commerce products. See the Oracle ATG Web Commerce Search Installation
Roadmaps (page 11) section for some common multi-server arrangements.
Note: Before you begin, review the documentation for each of the products you plan to configure.
The sections that follow explain the configuration steps involved. Configure all components in the
<ATG10dir>/home/servers/servername/localconfig/ configuration layer. See Creating Additional ATG
Server Instances in the ATG Installation and Configuration Guide for information on creating servers.
Configuring Search Routing
Routing refers to a group of services responsible for sending indexing and searching requests to search engines
and processing the responses.
The default Oracle ATG Web Commerce Search configuration assumes that all Search modules are running
locally. In production, you will have Routing modules running in other locations (for example, in your client
application). You must perform the following configuration steps:
1. On each server on which you plan to run DAF.Search.Routing, configure the shared deployment directory.
See Configuring the DeployShare Directory (page 30).
2. Configure server and client lock managers between your servers. See the Locked Caching section of the ATG
Repository Guide for information.
30 4 Installing Oracle ATG Web Commerce Search
Configuring the DeployShare Directory
Search Administration and all search engines must have access to the deployment share directory in order
to deploy and search indexes. This is a single, scalable, shared directory where master copies of indexes are
stored. Your installation can have only one deployment share. This directory should ideally be located on a high-
performance drive separate from the Search Administration machine. For all machines running search engines,
the deployshare directory must be mounted at the same location, and the path must be exactly the same for all
components accessing this directory.
On each server on which you plan to run the DAF.Search.Routing module, configure the shared deployment
directory. To do this, change the deployShare property in the /atg/search/routing/LaunchingService
component on each server. Make sure the deployment directory is visible to all of the servers running search
engines. For example, a Windows path might be:
deployShare=\\\\yourSharedDirectory\\yourDeployShareDirectory
Configuring the Lock Manager
Configure server and client lock managers between your servers. The lock manager is used for locked caching
mode (see the Locked Caching section of the ATG Repository Guide).
Both Search Administration and Oracle ATG Web Knowledge Manager automatically point their
SearchClientLockManager components to the ClientLockManager_production component;
SearchClientLockManager requires no further configuration. If you are using other Oracle ATG Web
Commerce applications with Search, you may need to perform this configuration manually. You must also
configure the following components:
• /atg/dynamo/service/ClientLockManager component on the production server
• /atg/dynamo/service/ClientLockManager_production component on the asset management server
• /atg/dynamo/service/ClientLockManager_production component on the staging server (if using one)
In each of those components, perform the following configuration:
useLockServer=truelockServerPort=production_server_portlockServerAddress=production_host_name_or_IP
The address and port specified for this component on the asset management and staging servers must match
that of the ClientLockManager on the production instance.
Configuring the IDGenerator
An IDGenerator generates IDs for repository items. There should be only one per repository.
On the asset management and agent servers, by default the /atg/search/service/SearchIdGenerator
component points to the /atg/dynamo/service/IdGenerator_production component, which is the ID
generator for the production server. This ensures that Search Administration and all search client applications
use the same ID generator for the ATG Search repositories.
If you are adding a staging server to your configuration, you may need to configure this component as shown:
4 Installing Oracle ATG Web Commerce Search 31
/atg/search/service/SearchIdGenerator:componentPath=/atg/dynamo/service/IdGenerator_production
Configuring the IndexingOutputConfig Component
This guide addresses only certain basic configuration options for IndexingOutputConfig components. See
the ATG Search Administration Guide for detailed information on this component. By default Oracle ATG Web
Commerce Search includes an instance of this component called ProductCatalogOutputConfig, which is
intended for use in indexing Oracle ATG Web Commerce catalogs.
To ensure that this component is started automatically, you must add your ProductCatalogOutputConfig to
the list of exported services in the /atg/dynamo/server/RmiServer.properties file (if you are using CIM,
this step is done automatically). For example:
exportedServices+=/atg/commerce/search/ProductCatalogOutputConfig
Additional configuration depends on your system architecture:
Architecture IndexingOutputConfig Component Configuration Required
Indexing production repository
from the asset management
instance
None
Indexing production repository
from the production instance
On the asset management instance, configure the following:
/atg/commerce/search/ProductCatalogOutputConfig
targetName=Production
The targetName value is the ATG Content Administration deployment
target name for the production instance
Indexing production and
staging repositories from the
asset management instance
None
Indexing production and
staging repositories from the
asset management instance
On the asset management instance, configure the following:
/atg/commerce/search/ProductCatalogOutputConfig
targetName=Production
The targetName value is the ATG Content Administration deployment
target name for the production instance.
On the asset management server, configure the following:
/atg/commerce/search/ProductCatalogOutputConfig_staging
targetName=Staging
The targetName value is the ATG Content Administration deployment
target name for the staging instance.
32 4 Installing Oracle ATG Web Commerce Search
For all servers, make sure that your IndexingOutputConfig component is started automatically (if you are
using CIM, this step is done automatically). You can do this by adding it to the initialServices list as shown:
/atg/commerce/Initial:initialServices+=\ /atg/commerce/search/ProductCatalogOutputConfig_staging
If you are using a staging server, you must also add the ProductCatalogOutputConfig_staging component
to the /atg/search/repository/IndexingDeploymentListener on the asset management server in order
for synchronized indexing (see Synchronizing Versioned Repository Deployment and Indexing (page 39)) to
work on that instance.
indexingOutputConfigs+=\ /atg/commerce/search/ProductCatalogOutputConfig_staging
Configuring Customization Adapters for Search Merchandising
The Oracle ATG Web Commerce platform uses customization adapters to submit post-indexing customization
data to Oracle ATG Web Commerce Search. Each customization adapter is a Nucleus component that is invoked
by Search Administration when it runs an indexing job. When you set up a project in Search Administration for
indexing your Commerce catalog, you can select the customization adapters to include in the project as part of
specifying the indexing content.
There are three customization adapters commonly used to submit post-indexing customization data for
Commerce sites using Oracle ATG Web Commerce Merchandising:
• /atg/commerce/search/refinement/CatalogRefineConfigAdapter – Manages the process of
generating the refinement configuration files used for Faceted Search and submitting them to Search.
• /atg/commerce/search/config/CatalogRankConfigAdapter – Manages the process of generating the
search configuration files used for Search Merchandising and submitting them to Search.
• /atg/commerce/search/config/SearchUpdateAdapter – Manages the process of collecting updated
inventory information used for Search Merchandising and submitting it to Search.
Registering Adapters
To make the customization adapters available in Search Administration, they must be registered with the /atg/
search/adapter/customization/CustomizationAdapterRegistry component. This configuration is
set up by default for production and staging environments; if you add a third environment, however, you must
register adapters for it.
The CustomizationAdapterRegistry component has a nameToNucleusPath property that lists the
registered adapters (if you are using CIM, this step is done automatically). This property is a Map in which each
key is the display name of an adapter and the corresponding value is the actual Nucleus component.
The CustomizationAdapterRegistry component is in the DAF.Search.Index module, but by default
nameToNucleusPath is not set there. The DCS.Search.Index module adds the three Oracle ATG Web
Commerce customization adapters to the Map:
nameToNucleusPath+=\
4 Installing Oracle ATG Web Commerce Search 33
select_custsource_type.repository.rank_config=/atg/commerce/search/config/ CatalogRankConfigAdapter,\select_custsource_type.repository.staging_rank_config=/atg/commerce/search/config/ CatalogRankConfigAdapter_staging,\select_custsource_type.repository.search_update=/atg/commerce/search/config/Search UpdateAdapter,\select_custsource_type.repository.staging_search_update=/atg/commerce/search/ config/SearchUpdateAdapter_staging,\select_custsource_type.repository.refine_config=/atg/commerce/search/refinement/ CatalogRefineConfigAdapter,\select_custsource_type.repository.staging_refine_config=/atg/commerce/search/ refinement/CatalogRefineConfigAdapter_staging
In this configuration, each Map key (such as select_custsource_type.repository.rank_config), is a
resource bundle key that represents the displayName property of the corresponding adapter component,
rather than being the display name itself. Using the resource bundle key allows you to modify or translate the
displayName value without affecting the registration of the component.
Configuring Adapters
The customization adapters are invoked on the Search Administration environment, but the processes they
execute run on the Oracle ATG Web Commerce environment. Therefore, you must have separate sets of adapter
components on each environment, and configure them to communicate over RMI.
The CatalogRefineConfigAdapter, CatalogRankConfigAdapter, and SearchUpdateAdapter
components are actually generic references that you configure to point to other components:
• On the Commerce environment, these components point to local customization adapters that actually run
the processes.
• On the Search Administration environment, these components point to proxies that communicate remotely
with the customization adapters running on the Commerce environment.
Configuring the Adapters on the Production Server
The customization adapters are configured by default, so you typically do not need change the configuration on
the Commerce environment.
If for some reason you need to configure a customization adapter on the Commerce environment, set its
componentPath property to the Nucleus pathname of the corresponding local adapter:
• For the CatalogRefineConfigAdapter, set the componentPath property to /atg/commerce/search/
refinement/LocalCatalogRefineConfigAdapter.
• For the CatalogRankConfigAdapter, set the componentPath property to /atg/commerce/search/
config/LocalCatalogRankConfigAdapter.
• For the SearchUpdateAdapter, set the componentPath property to /atg/commerce/search/config/
LocalSearchUpdateAdapter.
Configuring the Adapters on the Asset Management Server
In the asset management environment, set up the proxies that communicate with the customization
adapters on the Commerce environment. First, configure the CatalogRefineConfigAdapter,
CatalogRankConfigAdapter, and SearchUpdateAdapter components to point to their corresponding
remote proxies:
34 4 Installing Oracle ATG Web Commerce Search
• For the CatalogRefineConfigAdapter, set the componentPath property to /atg/commerce/search/
refinement/RemoteCatalogRefineConfigAdapter.
• For the CatalogRankConfigAdapter, set the componentPath property to /atg/commerce/search/
config/RemoteCatalogRankConfigAdapter.
• For the SearchUpdateAdapter, set the componentPath property to /atg/commerce/search/config/
RemoteSearchUpdateAdapter.
Now configure the RemoteCatalogRefineConfigAdapter, RemoteCatalogRankConfigAdapter, and
RemoteSearchUpdateAdapter components to communicate with the adapters running in the production
server. Set the enabled property of these components to true, and set the remoteHost and remotePort
properties to the machine name and RMI port number of the server instance on which the Commerce
application is running (if you are using CIM, this step is done automatically). For example:
enabled=trueremoteHost=production.host.comremotePort=28860
If your installation includes a staging server, you will need to give the asset management server access to those
configuration adapters as well. Configure the following components as indicated:
/atg/commerce/search/config/CatalogRankConfigAdapater_staging:componentPath=RemoteCatalogRankConfigAdapter_staging
/atg/commerce/search/config/RemoteCatalogRankConfigAdapter_staging:enabled=trueremoteHost=staging.host.comremotePort=18860
/atg/commerce/search/config/SearchUpdateAdapter_staging:componentPath=RemoteSearchUpdateAdapter_staging
/atg/commerce/search/config/RemoteSearchUpdateAdapter_staging:enabled=trueremoteHost=staging.host.comremotePort=18860
Invoking Adapters
Each time a catalog indexing job runs, Search Administration invokes the specified post-indexing customization
adapters after the indexing itself is complete. Doing this ensures that Search has up-to-date search refinements,
search configurations, and inventory information (assuming you specify the appropriate adapters).
Depending on the needs of your sites, you may want to run the customization adapters more frequently than
you re-index the product catalog. This is particularly true if your inventory information changes rapidly, and your
search configurations include rules based on inventory status. For example, for rules that exclude out-of-stock
items, you want to make sure the data that triggers these rules is up to date.
When you set up scheduled indexing jobs for your project in Search Administration, you can also set up post-
indexing customizations jobs and schedule them to run more frequently than the indexing jobs. For more
4 Installing Oracle ATG Web Commerce Search 35
information about creating and scheduling post-indexing customizations jobs, see the ATG Search Administration
Guide.
Configuring SearchSQLRepository Components
The SearchSQLQLRepository and SearchSQLRepositoryEventServer components are used by the
subscriber table needed for GSA distributed cache invalidation.
These components are located in the DAF.Search.Base module, in the /atg/search/service component
directory. By default, these components point to the local, default JDBC connection and the related services. If
your environment has both production and internal-facing ATG instances, you must change these components
on the internally-facing machines to point to the _production versions of the target components.
Note: Oracle ATG Web Knowledge Manager sets these components to point to their _production equivalents
automatically, and you do not need to configure them further.
Component Default Setting Production Setting
/atg/search/service/
SearchSQLRepository
/atg/dynamo/service/
jdbc/
SQLRepository
/atg/dynamo/service/jdbc/
SQLRepository_production
/atg/search/service/
SearchSQLRepositoryEventServer
The SQLRepositoryEventServer
used by ATG Search. Should always
match the data source configuration
of SearchJTDataSource.
/atg/dynamo/server/
SQLRepositoryEventServer
/atg/dynamo/server/
SQLRepositoryEventServer_production
Note: There should only be one SQLRepositoryEventServer per data source. It’s important that /atg/
dynamo/server/SQLRepositoryEventServer_production on the asset management server and /atg/
dynamo/server/SQLRepositoryEventServer on the production server are connected to the same database
account.
If you want to use something other than the production server as the target, clone the _production
components and configure them to point to components on your alternative server. Most installations should
not require this.
Indexing Repositories Managed by ATG Content Administration
This section addresses configuration options for indexing versioned data managed by ATG Content
Administration, such as a product catalog. Repository indexing relies on the IndexingOutputConfig
component. This installation guide discusses this component in terms of configuring your installation
architecture; for other information on this component, see the ATG Search Administration Guide.
Selecting a Data Loading Location
You will have to decide whether you want data loading to run on your asset management server or your
production server (see the ATG Search Administration Guide for information on this process). Consider the
following factors when making your decision:
36 4 Installing Oracle ATG Web Commerce Search
• Size of your sites. The main benefit to loading data on the asset management server is that it relieves load
on the production server; asset management-side loading may be appropriate if you have relatively small
sites. Alternatively, data loading on production is more appropriate if you have large sites and can dedicate
additional non-DRP servers to handle the load.
• Data availability. You may have data that is available to production, but not to the asset management server,
such as inventory information.
When you add repository content to a search project in Search Administration, you specify local or remote
location. If Search Administration is running on your asset management server, select Local to load data on the
asset management server. To load data on a production server, select Remote, and specify a production server as
the host. See the ATG Search Administration Guide for details on search projects.
The two diagrams that follow show the arrangement of components in both configurations. For simplicity, these
diagrams omit the staging server and database.
Data Loading on the Asset Management Instance
4 Installing Oracle ATG Web Commerce Search 37
Notice the connector in the second diagram that runs between the Asset Management Instance and the
IndexingOutputConfig component on the Production Instance:
Data Loading on the Production Instance
Using Incremental Loading with ATG Content Administration
If you are using incremental indexing of your versioned content, whenever you modify and redeploy your
content, the incremental loading system needs to be notified about the changes in order to reflect them in
the next incremental index. Therefore, an IndexingOutputConfig component must run in the ATG Content
Administration environment (asset management server) and listen for change events on the deployment
repository (production server).
You can actually load the content using an IndexingOutputConfig component running on either server,
as explained in the previous section. If you want to load content on the production server, configure the
targetName property of the IndexingOutputConfig component on the management server to the name
38 4 Installing Oracle ATG Web Commerce Search
of the ATG Content Administration deployment target. However, if you are loading content on the asset
management server, you must not configure this property, or loading will not work.
If you are using a secure product catalog (for instance, as part of a B2B commerce installation), you must change
the indexing configuration on the asset management server, so that it can listen for change events on the on the
deployment target repository. By default, this is configured to point to the non-secure catalog repository. To do
this, in the /atg/commerce/search/IndexedItemsGroup.properties file, change the following property:
repository=/atg/commerce/catalog/ProductCatalog_production
The new value should be as shown:
repository=/atg/commerce/catalog/SecureProductCatalog
Configuring the IndexingOutputConfig Component for Versioned Repositories
If you are indexing a versioned repository and are running your indexing process on the asset management
server, make sure that the IndexingOutputConfig component and the related /atg/commerce/search/
IndexedItemsGroup are configured to point to the production repository, rather than the versioned repository.
If this configuration is incorrect, your index may not accurately reflect your published repository. This example
describes how things could go wrong:
1. A project is checked in to ATG Content Administration, and the IndexingOutputConfig component on the
asset management server records the change in the incremental item queue.
2. The project is deployed to production.
3. Another project is checked in, and these changes are also recorded in the incremental item queue.
4. When incremental indexing runs, both sets of changes are processed, although the second set has not
actually been deployed to production.
The result is an index that does not accurately reflect your repository, in ways for which it is difficult to test.
By default, the IndexingOutputConfig component is configured to load product catalog data, and to be
used on either the asset management or the production server. The IndexingOutputConfig component’s
repository property refers to the catalog repository as ProductCatalog:
repository=/atg/commerce/catalog/ProductCatalog
In DCS.Search.Versioned, this value is overridden to point to ProductCatalog_production:
repository=/atg/commerce/catalog/ProductCatalog_production
The IndexingOutputConfig component may also use a repositoryItemsGroup property:
repositoryItemsGroup=/atg/commerce/search/IndexedItemsGroup
By default, IndexedItemsGroup also points to /atg/commerce/catalog/ProductCatalog, and is
overridden by DCS.Search.Versioned to ProductCatalog_production.
4 Installing Oracle ATG Web Commerce Search 39
If you are using similar components to index non-catalog repository data, be sure they are configured to point to
the correct repository.
Note: You will see a warning during indexing if the IndexingOutputConfig and IndexedItemsGroup
components’ repositories don’t match.
Synchronizing Versioned Repository Deployment and Indexing
You can configure search components to synchronize your repository deployments with the indexing process,
ensuring that changes deployed from ATG Content Administration are reflected in your search index as soon
as possible. This means that you do not have to schedule indexing for your versioned repositories in Search
Administration. You will still have to schedule indexing for any other data, such as files or other repositories that
are not part of synchronized deployment.
To use synchronized deployment indexing:
1. If indexing a repository other than the product catalog, append your IndexingOutputConfig
component to the list in the IndexingOutputConfigs property of the /atg/search/repository/
IndexingDeploymentListener.properties file.
Note: The product catalog IndexingOutputConfig is in this list by default.
2. In Search Administration, create a search project.
3. Add the appropriate IndexingOutputConfig component as a Content Set to the search project.
4. Create automatic indexing rules for your search project. See the ATG Search Administration Guide for
information on automatic indexing rules.
5. When you deploy your catalog, indexing starts.
Note: While indexing takes place, the repository and search index are in an inconsistent state.
The IndexingOutputConfig component includes a list of repositories to track for synchronization purposes in
the synchronizedRepositories property. For each repository, you can specify item descriptors to watch for.
In the default configuration, the DCS.Search.Versioned configuration layer adds the following configuration
to the ProductCatalogOutputConfig:
synchronizedRepositories+=\ /atg/search/repository/RefinementRepository_production=*
The asterisk tells the deployment synchronization listener that when any item type in the production refinement
repository is deployed, with no catalog repository changes, run the CA_deployment_target_name-
Customizations indexing rule.
If you want to synchronize only on certain item types, you can specify the types to watch by listing them as
shown:
synchronizedRepositories+=\ /atg/search/repository/RefinementRepository_production=\ itemTypeA,\ itemTypeB,\ itemTypeC
In this case, the CA_deployment_target_name-Customizations indexing rule runs only when changes to the
specified item types are deployed.
40 4 Installing Oracle ATG Web Commerce Search
Configuring Remote Routing
Under normal circumstances, Search Routing runs locally, as part of your Oracle ATG Web Commerce Search
installation. It is possible to direct queries to Routing instances running as part of other Search installations,
however, and you may want to do this as part of your development environment, allowing multiple clients to
access a single Search Administration instance, or to conserve processing resources.
The DAF.Search.Base module is a lightweight Routing module that can be run in a separate Oracle ATG Web
Commerce instance, independent of Search Administration. You can run multiple instances of this module. An
example of this architecture follows:
Two-Machine Setup with Remote Routing
To connect a form handler to a remote Routing instance, configure the /atg/search/formhandlers/
SearchContext component as shown:
primaryConnection=rmi://remotehostA:8860/atg/search/routing/SearchServicefailoverConnections=\ rmi://remotehostB:port/atg/search/routing/SearchService,\ rmi://remotehostC:port/atg/search/routing/SearchService,\ rmi://remotehostD:port/atg/search/routing/SearchService
The remote Routing component uses the primaryConnection unless it fails, in which case it uses a connection
from the failoverConnection. If the first failoverConnection fails, it moves to the next, etc.
Configuring Data Sources for Oracle ATG Web Commerce
Search
ATG data source components are wrappers that provide the connection information that applications use to
connect to the repositories that represent your databases. You can configure datasources very easily through
CIM. If you must configure a datasource manually, see the information in this section.
You must configure the componentPath property of the /atg/search/service/SearchJTDataSource
component on each of your servers:
4 Installing Oracle ATG Web Commerce Search 41
componentPath=/atg/dynamo/service/jdbc/Your_Production_DataSource_Component
By default the SearchJTDataSource component is configured to point to JTDataSource_Production, which
is typically a switching data source. You must change this configuration to use a non-switching datasource (such
as JTDataSource_Production_core) that points to the production database. (If you are using switching, you
can configure the componentPath to point to either datasource that makes up your switch configuration, but
do not use the switching datasource itself.)
Note: The startSQLRepository script, which is used to import initial data into Oracle ATG Web Commerce
application databases, uses a FakeXADataSource, while JBoss uses a JNDIReference JTDataSource. To
work around this, you define a FakeXADataSource in <ATG10dir>/home/localconfig/atg/dynamo/
service/jdbc (which startSQLRepository uses) and the JNDIReference in <ATG10dir>/home/
servers/server_name/localconfig/atg/dynamo/service/jdbc (which JBoss uses). The JNDI name must
match the JBoss name for the management schema.
Note: All servers that share a data source need to use a common server lock manager.
Creating the Oracle ATG Web Commerce Search Database
Tables
Your Search Administration installation and all Routing installations must have access to the Oracle ATG Web
Commerce Search databases. You can create and populate databases very easily through CIM. If you must create
database tables manually, refer to this section.
Search components rely on two database schemas, one internal, which is used by management applications
such as Search Administration and Oracle ATG Web Commerce Merchandising (referred to in the sections that
follow as the management schema), and one external, which is used by search client applications such as the
Routing module, Oracle ATG Web Commerce and Oracle ATG Web Knowledge Manager (referred to as the
production schema). Most of the search tables are installed in the production schema so they are available at
runtime to other applications.
Creating Base Oracle ATG Web Commerce Search Database Tables
After installing ATG Search, create the database tables.
1. Run SQL scripts to create the Search Routing tables in your production database schema:
• <ATG10dir>\DAF\Search\Routing\sql\db_components\db_type\
routing_ddl.sql
• <ATG10dir>\DAF\Search\Index\sql\db_components\db_type\
search_ddl.sql
• <ATG10dir>\DAF\Search\common\sql\db_components\db_type\
refinement_ddl.sql
• <ATG10dir>\DAF\Search\Topics\sql\db_components\db_type\
42 4 Installing Oracle ATG Web Commerce Search
search_topics_ddl.sql
The search_ddl.sql script creates the tables used by the IncrementalLoader to track changes
to the indexed repository. The refinement_ddl.sql script creates the tables used by the
RefinementRepository.
2. Run the scripts to create the Search Administration tables in the production database schema. The following
script is located in the <ATG10dir>\Search10.1.2\SearchAdmin\common\sql\db_components
\db_type\ directory:
• searchadmin_admin_ddl.sql
The following scripts are located in the <ATG10dir>\Search10.1.2\SearchAdmin\customizations\sql
\db_components\db_type\ directory:
• searchadmin_dictionary_ddl.sql
• searchadmin_termweights_ddl.sql
• searchadmin_queryrules_ddl.sql
• searchadmin_tpo_ddl.sql
• searchadmin_facet_ddl.sql
3. If you have not already done so for any of your other applications, run the following script to create Oracle
ATG Web Commerce Business Control Center tables:
• <ATG10dir>\BIZUI\sql\install\db_type\bizui_all_ddl.sql
Importing Initial ATG Search Data
Having created your database tables, they must be populated with initial data before using Oracle ATG Web
Commerce Search.
Note: The import scripts import data using startSQLRepository. Make sure your JTDataSource and
FakeXADatasource are configured to permit this in home/localconfig. JTDataSource should automatically
point to FakeXADataSource. FakeXADataSource should point to the administration database, and
FaxeXADataSource_production should point to the production database. Note that some product
installation processes, including Oracle ATG Web Knowledge Manager Self Service, need FakeXADataSource
pointed at the production database, so you may need to change the FakeXADataSource between installations.
If you haven’t already done so for other applications, run the following scripts to import initial data for the Oracle
ATG Web Commerce Business Control Center:
<ATG10dir>\Publishing\base\install\importPublishing.bat
<ATG10dir>\BIZUI\install\importBizui.bat
To import initial data into your Oracle ATG Web Commerce Search schema, run the following commands:
<ATG10dir>\home\bin\startSQLRepository.bat|sh -m DAF.Search.Topics-repository /atg/searchadmin/TopicRepository -import
4 Installing Oracle ATG Web Commerce Search 43
<ATG10dir>\DAF\Search\Topics\data\initial\topics-data.xml
<ATG10dir>\home\bin\startSQLRepository.bat|sh -m SearchAdmin.AdminUI-m DPS.InternalUsers -repository/atg/userprofiling/InternalProfileRepository-import<ATG10dir>\Search10.1.2\SearchAdmin\common\data\initial\searchadmin-profile.xml
<ATG10dir>\home\bin\startSQLRepository.bat|sh -m SearchAdmin.AdminUI-repository /atg/searchadmin/SearchAdminRepository -import<ATG10dir>\Search10.1.2\SearchAdmin\common\data\initial\searchadmin-data.xml
Creating Database Tables for Oracle ATG Web Commerce Search with Multisite
If you are using the Oracle ATG Web Commerce multisite feature, see the ATG Multisite Administration Guide
for configuration information. You will need to run the following additional database script against your
administration, production, and versioned database schemas:
<ATG10dir>\Search10.1.2\SearchAdmin\common\sql\db_components\db_type\search_site_ddl.sql
Creating Database Tables for Oracle ATG Web Commerce Search with Oracle ATG Web
Commerce Merchandising
If you are running Oracle ATG Web Commerce Search with Oracle ATG Web Commerce Merchandising, see
Configuring Database Tables in the ATG Merchandising Administration Guide, skipping any steps that you have
already performed.
Deleting Oracle ATG Web Commerce Search Database Tables
Each table creation script noted in the previous sections has a corresponding drop script located in the
component_path\sql\uninstall\db_type directory. For example, the routing_ddl.sql table creation
script for Oracle has a corresponding drop script at:
<ATG10dir>\DAF\Search\Routing\sql\uninstall\oracle\drop_routing_ddl.sql
Accessing Search Administration
To access Search Administration:
44 4 Installing Oracle ATG Web Commerce Search
1. Make sure your database is running.
2. Deploy the EAR to your application server. See the Oracle ATG Web Commerce Search Installation
Roadmaps (page 11) section of the Planning Your Oracle ATG Web Commerce Search Installation (page 7)
chapter for details on how to build EAR files for different search applications.
3. In your Web browser, access the Oracle ATG Web Commerce Business Control Center on your administration
server. The default location is http://hostname:port/atg/bcc.
4. Enter your login and password.
5. In the left navigation area, click Search Administration.
6. Click a menu option to begin working with Search Administration.
Note: If you are familiar with ATG Content Administration, bear in mind that search projects are not related to
ATG Content Administration projects.
The Search Administration interface is loaded into the current browser window. See the ATG Search
Administration Guide for administration tasks.
Appendix A. Oracle ATG Web Commerce Search Component Ports 45
Appendix A. Oracle ATG Web Commerce
Search Component Ports
The default ports used by Oracle ATG Web Commerce Search components may conflict with your organization’s
current port assignments. This appendix describes the ports used.
If you are using Search with Oracle ATG Web Knowledge Manager or Oracle ATG Web Knowledge Manager Self
Service, the RMI IIOP port must be opened so that Search can extract content from the Knowledge Manager/Self
Service installation. By default, the RMI port is 8860.
Knowledge Manager, Self Service, and Search Administration communicate with the search engines using HTTP.
One port per search engine is by default allocated from the 6070—6100 range.
The following table summarizes the locations where ports are set and used:
Component Default
Port(s)
Locations
Search engines 6072-6100 Search Administration: Host Machine Details for each machine.
5 Glossary 47
5 Glossary
The glossary provides a description of the terms used in ATG Search.
Term Definition Replaces prior term
content Raw data for indexing, such as html files, Word
documents, or repositories.
content label The content label identifies the purpose of a
content set, such as a catalog or a collection
of articles. If your search project includes
different types of content that you want to
search individually, you can use content labels to
differentiate the content types.
content set Bundles content into a single organizational unit
that can be indexed as one.
document repositories/sets
customizations Blanket term for topic sets, query rules,
dictionaries, facet sets, languages, parsing
options, query rules, and preferred answers.
deploying The action of moving indexes and/or
customization data from a staging to a
production environment
dictionary Contains definitions of special words and the
relationships between them. Users can add terms
to improve the search responses.
expansion A synonym for a term, either defined in a custom
dictionary or in the default language files
facet set Allows users to search within an existing result
set.
index The end result of indexing content, and what is
used to answer queries submitted by end-users.
IdeaMap
indexing environment Search environment that indexes content. indexing server
48 5 Glossary
Term Definition Replaces prior term
partition An index is divided into one or more Logical
Partitions, depending on the number of Content
Sets. Each Logical Partition is composed of one or
more Physical Partitions, depending on the size of
the content.
learning files Rules that automatically apply topics to a content
collection.
preferred answer Custom response to frequently-asked end-user
questions. Can be linked to source text within
content collections.
project Projects represent the top-level container for
managing and indexing content. The project
associates customization data and search
environments with an index.
Job script (also included
actions, which are now part of
synchronization)
query rules Link actions in Oracle ATG Web Commerce Search
to certain types of user queries.
Search Administration The Search user interface, including both project
administration and the workbench.
Management Console
search engine Part of Search that creates the index, receives
an end-user query and returns the results of the
search.
Answer Server
search environment A combination of a single index and one or more
machines and search engines used for indexing or
searching.
Content clusters, also called
clusters
stem In linguistics, the stem is the part of a word that is
common to all its inflected variants. For example,
the stem of banking, banker, and banks is bank.
structured data Content that adheres to a well-defined structure;
typically repository data that has been exported
to XHTML.
target type Identifies the general purpose of the search
project. If you are using ATG Content
Administration, the target types should
correspond to your ATG Content Administration
targets.
topic set A hierarchy of topics. topic taxonomy
term Individual entry within a dictionary.
5 Glossary 49
Term Definition Replaces prior term
term weight set/stop
words
A set of term/weight pairs, in which the weight
describes the amount of influence a term has in a
search.
topics User-defined classification system that can be
applied to indexes, allowing end-users to refine
their search by finding related information.
type-ahead Search feature that allows you to provide a list
of terms that can be used in your UI to auto-
complete search terms as they are entered.
unstructured data Content that has no governing structure, such
as that typically found in file systems, which
might include a variety of document formats and
information types.
Workbench Part of the user interface where users create and
edit customization data.
Management Console (partial)
Index 51
Index
Aarchitecture, 8
ATG Content Administration, 12
ATG Merchandising, 43
audience, 1
Bback-of-the-book index, 4
Business Control Center, 6
CCatalogRankConfigAdapter, 33
clean partition, 26
client application, 5
components, 4
Content Administration, 4
conventions, 1
customization adapters, 32
configuring, 33
invoking, 34
local access, 33
registering, 32
customization data, 4
Ddata loading location, selecting, 35
data sources
configuring, 40
database, 6
configuring, 41
importing intial data, 42
database tables
deleting, 43
deployment share directory, 11
deployment, synchronized with indexing, 39
deployShare property, 30
dictionaries, 4
documents, 3
Ffeatures, 3
HHTMLFilter, 24
IIDGenerator, configuring, 30
index, 3, 4
customizing, 5
index item, 3
indexing
synchronizing with deployment, 39
IndexingOutputConfig component, 31
for versioned repositories, 38
installing, 23
Llock managers, 30
logical partition, 5
Mmulti-server installations, configuring, 29
Nnatural language components, 4
OOracle ATG Web Commece Merchandising, 12
Oracle ATG Web Commerce, 12
Oracle ATG Web Commerce Search, installing, 23
Oracle ATG Web Knowledge Manager, 17
Oracle ATG Web Knowledge Manager Self Service, 17
PPDF Extract, 24
physical partition, 5
prerequisites, 10
Project Administration, 4
Rremote Routing, 40
requests, 4
rich text, indexing, 24
Routing, 5, 9
configuring, 29
SSearch Administration, 4
accessing, 43
52 Index
search engine, 5
search engines
adding, 25
configuring remote, 26
Search Merchandising
customization adapters, 32
search projects, 4
Search Query request, 4
Search Workbench (see Workbench)
SearchAdmin role, 6
searching, definition, 3
SearchSQLRepository component, 35
SearchSQLRepositoryEventServer component, 35
SearchUpdateAdapter, 33
security, 6
server, 8
Vversioned repositories
indexing incrementally, 37
View Item request, 4
WWorkbench, 5