November 29, 2012

IBM Content Collector for SAP Applications

– Sizing, Configuration, and High Availability –

White Paper
Version 1.0

Before reading this paper

check for the latest version at
http://www.ibm.com/support/docview.wss?uid=swg27036773

Raiko Nitzsche
Software Engineer
Martin Russold
Quality Assurance Lead

IBM ECM Development

IBM Germany Research and Development Lab

Abstract
IBM Content Collector for SAP provides advanced archiving capabilities
that help SAP users run their businesses more efficiently. It provides ac-
cess to business information of all types, supports the transfer of older
information to lower-cost disks and tapes, and improves system perfor-
mance.
This white paper gives information about advanced topics as sizing and
high availability setup of IBM Content Collector for SAP Applications.

c Copyright IBM Corporation 2012. All rights reserved.

November 29, 2012

Contents
1 Introduction 3

2 Content Collector for SAP environment 4

2.1 Physical setups . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 SAP archiving scenarios . . . . . . . . . . . . . . . . . . . . . . . 6

3 Hardware configurations 8
3.1 Gathering input for system sizing . . . . . . . . . . . . . . . . . . 8
3.2 Small environment . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.3 Medium environment . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.4 Large environment . . . . . . . . . . . . . . . . . . . . . . . . . . 9

4 Collector Server configuration 11

4.1 Collector Server instance . . . . . . . . . . . . . . . . . . . . . . . 11
4.2 Small configuration . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.3 Medium configuration . . . . . . . . . . . . . . . . . . . . . . . . 13
4.4 Large configuration . . . . . . . . . . . . . . . . . . . . . . . . . . 13

5 High-availability scenarios 14
5.1 High-availability capabilities and limitations . . . . . . . . . . . . 14

6 For more information 16

c Copyright IBM Corporation 2012. All rights reserved.

November 29, 2012

1 Introduction
Before reading this white paper, get familiar with the product IBM Content
Collector for SAP Applications. Use the following web page as a starting point:
http://www.ibm.com/support/docview.wss?uid=swg27036331
This web page gives you access to the product documentation and all other
important information that is available for IBM Content Collector for SAP
Applications. We strongly encourage you to use the provided information to
get a good understanding of the product.
The goal of this white paper is to give you some guidance regarding sizing,
configuration, high availability of Content Collector for SAP, and the usage of a
load balancer in the context of Content Collector for SAP. As the white paper
will be updated from time to time, check the following web page for the latest
version of this white paper:
http://www.ibm.com/support/docview.wss?uid=swg27036773
The white paper is organized as follows:
Chapter 2 describes the environment where Content Collector for SAP runs.
In addition, it covers all possible physical setups and explains the common
SAP archiving scenarios. Read this chapter before you continue with any other
chapter.
Chapter 3 introduces the standard configurations for different performance re-
quirements. To choose the right configuration for you it is necessary to deter-
mine the performance requirements first. A list of items guides you through this
activity.
Chapter 4 describes how to configure a Collector Server instance based on the
hardware configuration that you selected in Chapter 3.
Chapter 5 shows how to set up high availability and how to use a load bal-
ancer.

Page 3 of 17
November 29, 2012

2 Content Collector for SAP environment

This chapter provides a short overview of the environment where Content Col-
lector for SAP runs so that you understand the role of Content Collector for
SAP in an SAP environment. The paper concentrates on sizing information and
the high-availability setup of Content Collector for SAP. However, it is necessary
to understand the complete SAP archiving environment before you can size it
and set up high availability.
SAP provides an interface to external storage systems for archiving documents
or data. This interface is called SAP ArchiveLink. Each storage system that
wants to connect to an SAP system must support SAP ArchiveLink. The exter-
nal storage system can be a simple file-based archiving tool or a sophisticated
enterprise content management system. Because most storage systems have
their own API - mainly because they provide more functionality than SAP
ArchiveLink - a translator is needed between SAP ArchiveLink and the API
of the storage system. Content Collector for SAP has been designed to trans-
late the protocols of SAP ArchiveLink and several IBM storage and content
management systems, as shown in figure 1.

SAP Content Collector for SAP Backend

Figure 1: Content Collector for SAP integration in SAP archiving environment

In such an environment, the overall performance is determined by SAP and the

storage or content management system. Therefore, it is not sufficient to size and
tune Content Collector for SAP only. Of course, you must ensure that Content
Collector for SAP does not become the bottleneck. It must be able to handle the
data throughput between SAP and the storage or content management system.
But, although it is always good to have some reserves, an oversized environment
is expensive and does not improve the performance.

2.1 Physical setups

You can deploy Content Collector for SAP in three different ways:
• On the SAP servers (figure 2)
• On a separate physical server (figure 3)
• On the enterprise content management system (figure 4)
Technically, there is no preference for any of the three deployment methods.
Nevertheless, only the methods that are shown in figure 3 and figure 4 are
normally used. By installing Content Collector for SAP on its own server, you

Page 4 of 17
November 29, 2012

have full control over the sytem: from choosing the preferred operating system
to fine-tuning the server for maximum document throughput without affecting
the storage or enterprise content management system. On the other hand, by
installing Content Collector for SAP on the enterprise content management
system, you do not need an additional server that requires maintenance. In
addition, the network connection between Content Collector for SAP and the
storage or enterprise content management system is only virtual and hence faster
and with lower latency.

SAP Server Backend Server

Collector
SAP Server Backend
Instance

Figure 2: Collector Server on the SAP server

SAP Server Backend Server

Collector Server

Collector
SAP Server Backend
Instance

Figure 3: Collector Server on a separate server

SAP Server Backend Server

Collector
SAP Server Backend
Instance

Figure 4: Collector Server on the enterprise content management system

Which installation scenario to choose depends on the enterprise content man-

agement system used. If only a single system is used, you can install Collector
Server as a protocol translator directly on the server. In a more complex en-
terprise content management system environment, for example, where multiple
enterprise content management systems use the same Collector Server, the in-
stallation on a separate server is preferable.

Page 5 of 17
November 29, 2012

2.2 SAP archiving scenarios

In a typical SAP archiving environment, the following types of documents and

data are archived.
• Outgoing documents
• Print lists
• Incoming documents
• Data

Outgoing documents are documents that are generated by the SAP ap-
plications, such as invoices or reports. SAP archives such documents in the
external storage system by using ArchiveLink. The document type is normally
PDF and the document size typically ranges from 10 to 100 KB, although, in
some business areas, sizes of several megabytes are common.
The amount of outgoing documents mainly depends on the size and the use
of the SAP system itself, for example, what kind of tasks in the company are
executed inside SAP and how many users work with the system. The number
of generated documents per day and at peak times is a key input for the sizing
of Content Collector for SAP and mainly determines the required throughput
of the system.

Print lists are documents that are generated by SAP and are formatted for
printing. Print lists usually have a size of up to 500 KB. The frequency for
generating such print lists is normally much lower than for the outgoing docu-
ments. When you size your system, add the print lists to the total number of
documents generated per day.

Incoming documents are documents that are generated outside of the SAP
system and are available in electronic form, for example, by scanning documents.
Those documents are typically archived before they are linked to an SAP busi-
ness object. The document type is typically TIFF, but newer documents are
also available as PDFs. Because they are mostly scanned documents, their size
is rather large depending on the number of pages in a document.
Incoming documents are already stored in the archive system and only a short
identifier is transferred to the SAP system. Thus, this scenario does not require
large system resources - at least compared to the outgoing documents scenario.
The actual document only needs to be transferred from the archive system to
SAP when the user accesses it with the attachment link in the business object.
This is a manual interaction compared to the automatic document generation
and the archiving of the outgoing documents. Thus, the required bandwidth
and performance depends on the number of concurrent users in the SAP system
that are actively looking at archived documents.
If outgoing and incoming documents are used with the same Collector Server
instance, you can simplify the sizing by adding all documents that are handled

Page 6 of 17
November 29, 2012

per day and in peak hours. From the perspective of the ArchiveLink servers, all
those documents need to be transferred from one side to the other.

Data is extracted by SAP from its internal databases to a special data file.
SAP then archives the file in the archive system. This is an administrative
task that normally happens rarely and at scheduled times. The generated data
files typically reach sizes of several hundred megabytes, but the time window
allowed to transfer the files to and back from the archiving system is larger than
for the other documents, since this is not an adhoc interactive usage scenario.
Normally, data archiving is automatically covered by the sizing process. But for
the capacity sizing of the content management system, this data must be taken
into consideration.

Page 7 of 17
November 29, 2012

3 Hardware configurations
This chapter describes three standard hardware configurations for different per-
formance requirements.
Each configuration is based on the throughput of documents of the following
sizes:
• Small documents with an average size of 20 KB
• Medium documents with an average size of 500 KB
• Large documents with an average size of 5 MB
Depending on the gathered input data for the sizing, these standard configura-
tions can be used as a starting point for further adjustments to the actual usage
scenario.

3.1 Gathering input for system sizing

Before starting with the sizing of the Content Collector for SAP system, the
following information needs to be collected.
• The dominating document size, that is the average size of the such docu-
ments that are archived or retrieved the most
• The total number of documents archived and retrieved to/from the Col-
lector Server per day (docs/day) and at peak hours (docs/hour)
• How many users are expected?
• What operating system is planed for use?
• How many parallel requests are expected?
• What is the expected workload characteristics?

3.2 Small environment

Document throughput

Document size Document throughput docs/h

20 KB 180.000
500 KB 45.000
5 MB 4.500

Hardware requirements

• Dual-core CPU
• 2 GB RAM

Page 8 of 17
November 29, 2012

• One hard disk (min. 7200RPM and min. 100 IOPS) or

10 MB RAM file system (2 x size of largest archive file - REO or ALF)
• 1 Gbit network card
When running Content Collector for SAP on a separate server that could be
an IBM System x3250 or IBM Power 710, for example. Otherwise, the
described hardware needs to be available on the server.

3.3 Medium environment

Document throughput

Document size Document throughput docs/h

20 KB 750.000
500 KB 200.000
5 MB 20.000

Hardware requirements

• 4 cores
• 4 GB RAM
• 3 hard disks (10K-15K RPM and min. 150 IOPS) as RAID 0 or
1 SSD with more than 1000 IOPS or
50 MB RAM file system (2 x size of largest archive file - REO or ALF)
• 1 Gbit network card
When running Content Collector for SAP on a separate server that could be
an IBM System x3250 or IBM Power 710, for example. Otherwise, the
described hardware needs to be available on the server.

3.4 Large environment

Document throughput

Document size Document throughput docs/h

20 KB 1.500.000
500 KB 400.000
5 MB 40.000

Hardware requirements

• 8 cores
• 8 GB RAM

Page 9 of 17
November 29, 2012

• 6 hard disks (15K RPM and min. 150 IOPS) as RAID 0 or

1 SSD with more than 2000 IOPS
100 MB RAM file system (2 x size of largest archive file - REO or ALF)
• 2 Gbit network cards (one for incoming connections from SAP and one
for outgoing connections to storage or content system)
When running Content Collector for SAP on a separate server that could be
an IBM System x3550 or IBM Power 710, for example. Otherwise, the
described hardware needs to be available on the server.

Page 10 of 17
November 29, 2012

4 Collector Server configuration

4.1 Collector Server instance

Content Collector for SAP supports the concept of instances. An instance is the
smallest work unit of Content Collector for SAP. The number of instances is not
limited by Content Collector for SAP. You can configure an arbitrary number
of instances. However, as server resources are limited, only a certain number of
instances are feasible.
Depending on your usage scenario, it might be necessary or desirable to run
multiple instances. One Collector Server instance can only connect to one SAP
system. However, this connection is only necessary for certain usage scenarios
(for example, incoming documents). In this chapter we assume that only one
Collector Server instance is used.
A Collector Server instance consists of the following components: dispatcher,
engine, and agent (Figure 5).

Dispatcher Engine Agent

SAP Backend
Collector Server Instance

Figure 5: SAP environment with a single Content Collector instance

The dispatcher communicates with the SAP system and the agent communicates
with a storage system or an enterprise content management system. The engine
interprets and translates the messages that it receives through a dispatcher or
an agent, and then forwards the translated messages to an agent or a dispatcher.
The engine starts the dispatchers and agents during its startup phase.
The number and the type of dispatchers and agents that are initialized during
the startup phase are configured in a server configuration profile, which is by
default named archint.ini. Each Collector Server instance has its own server
configuration profile.
Different types of dispatchers and agents exists.
A Collector Server instance can have one or more HTTP dispatchers, RFC
dispatchers and client dispatchers. In addition, it can have one or more TSM
agents, OD agents, CM agents, and P8 agents (Figure 6).
It is also possible to set the number of dispatchers or agents to zero. In this
case, the dispatcher or agent is not started by the engine.
The type of dispatchers and agents depends on the implemented SAP archiv-
ing scenario. The following sections refer to dispatchers and agents in gen-
eral.
The server configuration profile specifies the port number where the configured
dispatchers accepts incoming requests from the SAP system.

Page 11 of 17
November 29, 2012

Client dispatcher P8 agent

RFC dispatcher CM agent

Engine
SAP Backend
Java RFC dispat. OD agent

HTTP dispatcher TSM agent

Collector Server Instance

Figure 6: Overview of the Content Collector for SAP integration in an SAP

environment

Each server configuration profile contains one or more archive sections. The
parameters of such an archive section identify a particular storage system or
enterprise content management system. The name of the archive section corre-
sponds to the name of a content repository defined in SAP.
For the purpose of this white paper, it is sufficient to remember that the server
configuration profile defines the following:
1. The port where the Collector Server instance runs
2. The dispatchers and agents that are started by the engine during its
startup phase
All other settings in a server configuration profile have a minor impact on the
sizing or the high-availability setup of Content Collector for SAP.
You must configure a Collector Server instance that corresponds to the stan-
dard hardware configuration that you identified in Chapter 3. Therefore, the
following sections will present a small, medium, and large instance configura-
tion.

4.2 Small configuration

A small configuration consists of two dispatchers and two agents. Use this con-
figuration if you have a small environment (see Chapter 3.2). The following
shows a sample minimum configuration for a Content Manager backend sys-
tem:
WEBDPS 2
DISPATCHERS 0
ARCHWINS 0

ADSMAGENTS 0
ODAGENTS 0
CMAGENTS 2
P8AGENTS 0

Page 12 of 17
November 29, 2012

4.3 Medium configuration

A medium configuration consists of four dispatchers and four agents. Use this
configuration if you have a medium environment (see Chapter 3.3). The follow-
ing shows a sample standard configuration for a P8 backend system:
WEBDPS 4
DISPATCHERS 0
ARCHWINS 0

ADSMAGENTS 0
ODAGENTS 0
CMAGENTS 0
P8AGENTS 4

4.4 Large configuration

A large configuration consists of 16 dispatcher dispatchers and 16 agents. Use

this configuration if you have a large environment (see Chapter 3.4). The fol-
lowing shows a sample large configuration for a TSM backend system:
WEBDPS 16
DISPATCHERS 0
ARCHWINS 0

ADSMAGENTS 16
ODAGENTS 0
CMAGENTS 0
P8AGENTS 0

You can configure an arbitrary number of dispatchers and agents for an instance.
Remember, however, that the other systems involved (for example, SAP system,
enterprise content management system) then also have to be able to process this
amount of data.
A large configuration is usually sufficient to fulfill most requirements. Neverthe-
less, if you have identified that your requirements exceed a large configuration,
please feel free to ask your Content Collector for SAP contact for help.

Page 13 of 17
November 29, 2012

5 High-availability scenarios
Before setting up Content Collector for SAP for high availability, you must be
familiar with the Content Collector for SAP environment (Chapter 2). You
also have to understand the concept of a Collector Server instance (Chapter 4).
It is assumed that the SAP Content Server HTTP Interface is used for the
communication between the SAP system and Content Collector for SAP. This
interface is based on HTTP and therefore stateless. This means that there is
no relationship between, or dependency on, any requests.
There are many possibilities and products that can be used to set up high
availability. Therefore the following sections focus on the specifics of Content
Collector for SAP.

5.1 High-availability capabilities and limitations

Content Collector for SAP does not offer built-in high availability. The product
has only few dependencies and can, therefore, work self-contained. You can
integrate the product into any high-availability solution, which is transparent
to the applications running on each server.
Chapter 2.2 introduced the following SAP archiving scenarios: outgoing docu-
ments, print lists, incoming documents, and data. Depending on the scenarios
used, different type of high-availability solutions (for example a load balancer)
can be used.
The SAP archiving scenarios can be divided into the following groups:
1. Scenarios driven by SAP system
2. Scenarios driven by Content Collector for SAP

Scenarios driven by SAP system

Outgoing documents, print lists, and data archiving belong to the scenarios that
are driven by the SAP system. Content Collector for SAP does not ensure the
integrity of the data provided by SAP. The SAP system is always the master
system that controls the data flow and maintains data integrity. Due to the
protocol that is used by SAP, the SAP system knows whether a document or
data is correctly stored in the archive system. If an error occurs during the
archiving process, SAP receives an error code or a timeout. SAP then considers
the action failed and the data to be still inside of SAP. It will then retry to store
the document or data later.
The following paragraphs describe, from a high-level point of view, which steps
are necessary to set up high availability for a Collector Server instance.
Step 1: Create two Collector Server instances, each on a different servers. Both
instances must use the same server configuration profile (archint.ini).
Step 2: Determine whether a Collector Server instance is operational or not.
Content Collector for SAP offers a monitoring tool called archcheck for this

Page 14 of 17
November 29, 2012

purpose. With this tool, you can query the operational status of a Collector
Server instance. It checks whether requests can be routed through the dis-
patcher, the engine, the agent and the storage system or enterprise content
management system. Depending on the return code you can identify if a Col-
lector Server instances is operational. A detailed description of the archcheck
tool can be found in the product documentation.1
Step 3: Set up a virtual IP address. The setup of a virtual IP address depends
on the high-availability solution. In addition, change the configuration of the
SAP system so that SAP sends its request to the configured virtual IP address.
The high-availability solution used forwards the request from the virtual IP
address to one of the Collector Server instances.
Thus, the scenarios that fall into this category can be set up for high availability
through a HTTP/HTTPS load balancer, for example. Depending on your needs,
you can set up the system as active-passive or active-active.

Scenarios driven by Content Collector for SAP

Archiving incoming documents is a scenario that is driven by Content Collec-

tor for SAP. In such a scenario, the communication path is reversed and the
high-availability model changes. Because the communication is not based on
HTTP/HTTPS, high-availability solutions based on this protocol cannot be
used. Low level (for instance a TCP load balancer) high-availability solutions,
however, can be used.

1 The monitoring tool archcheck is only available from IBM Content Collector for SAP

Applications Version 3.0. If you use IBM Content Collector for SAP Applications Version
2.2, the administration tool archadmin can be used instead. However, this tool has limited
capabilities. Compared to archcheck it can only verify whether the engine is operational.

Page 15 of 17
November 29, 2012

6 For more information

Use the following web pages to get the latest version of this white paper and to
get more information about IBM Content Collector for SAP Applications:
Sizing, configuration, and high availability for Content Collector for SAP
http://www.ibm.com/support/docview.wss?uid=swg27036773
Content Collector for SAP Applications Version 3.0 publication library
http://www.ibm.com/support/docview.wss?uid=swg27036331
Content Collector for SAP Applications Version 2.2 publication library
http://www.ibm.com/support/docview.wss?uid=swg27022108

Page 16 of 17
November 29, 2012

c Copyright IBM Corporation 2012

IBM Deutschland Research & Development GmbH
ATTN: Dept. 5248
Schoenaicher Strasse 220
71032 Boeblingen
Germany

Produced in Germany
November 2012

IBM, the IBM logo, and ibm.com are trademarks of International Business
Machines Corp., registered in many jurisdictions worldwide. Other product
and service names might be trademarks of IBM or other companies. A current
list of IBM trademarks is available on the Web at ’Copyright and trademark
information’ at www.ibm.com/legal/copytrade.shtml.
References in this publication to IBM products and services do not imply
that IBM intends to make them available in all countries in which IBM op-
erates.
No part of this document may be reproduced or transmitted in any form without
written permission from IBM Corporation.
Product data has been reviewed for accuracy as of the date of initial publication.
Product data is subject to change without notice. Any statements regarding
IBMs future direction and intent are subject to change or withdrawal without
notice, and represent goals and objectives only.
THE INFORMATION PROVIDED IN THIS DOCUMENT IS DISTRIBUTED
AS IS WITHOUT ANY WARRANTY, EITHER EXPRESS OR IMPLIED.
IBM EXPRESSLY DISCLAIMS ANY WARRANTIES OF MERCHANTABIL-
ITY, FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT.
IBM products are warranted according to the terms and conditions of the agree-
ments (e.g. IBM Customer Agreement, Statement of Limited Warranty, Interna-
tional Program License Agreement, etc.) under which they are provided.

Page 17 of 17