IBM Information Server Services and ASB Agent

What Is IBM Information Server Services? 

A service receives a request from an application, and this request will be passed to the 
processing Information Server product module (DataStage Jobs or Queries) where the 
request will be processed and the result will be sent back to the application. Services 
follow a “request and respond” technique in a service-oriented architecture. 



What Is ASB Agent? 

ASB stands for “Application  Service  Backbone”. An ASB agent is a Java based 
background process running on all machines that have an IBM Information Server 
engine installed. Whenever there is a request for service that needs processing by the 
IBM Information Server component, ASB agent receives and conveys the request. ASB 
agent acts as a bus over which the services are delivered. ASB agent is a 
communicator between the service tier and engine tier within the IBM information server 
topology. See Figure 1.

As shown in Figure 1, the IBM Information Server includes four tiers—Client Tier, 

Services Tier, Engine Tier, and Metadata Repository Tier.  Each of these tiers are 

described below. 

Client Tier 

Client tier is used to develop, deploy and administrate the information server. The client 

tier includes the IBM Information Server Console, IBM WebSphere DataStage and 

QualityStage and other product client modules. It is important  to note that IBM 

Information Server product client modules are not web based. 

Services Tier 

Services tier handles common services and product specific services. This tier has 

access to the Engine tier and Metadata repository tier. This services tier contains IBM 

WebSphere Application Server, and services for IBM Information Server.

Engine Tier 

Engine tier is where all the jobs and other processing will run. This tier includes IBM 

Information Server Engine and its components. ASB agent is part of IBM WebSphere 

DataStage Engine; when you install the IBM Information Server Engine, the ASB 

communication agent will also be installed. 

Metadata Repository Tier 

Metadata repository is a database that stores the metadata and other data for the IBM 

Information Server and all product modules. The IBM Information Server services are 

hosted by Application Server (Service Tier) and connect to the Metadata Repository 

Tier to store and retrieve data from the database.

All of the above described tiers work closely together to provide services for client 
applications. ASB agent, as a communicator, receives and conveys the services from 
the service tier to the engine tier. (This communication agent is only needed if the 
WebSphere DataStage Server and WebSphere Application Server are on separate 
machines.) 



Service Message Flow and ASB Agent Building Blocks 

When a service is requested, a sequence of steps occurs for processing the request. 
Figure 2 describes two ASB agents with a single Information Services Director server. A 
service request will arrive at WISD (WebSphere Information Services Director) Endpoint 
and is handed over to the ASB adapter. The service  will carry information about the 
requested operation. These operation details are used to determine the providers that will 
process the request. ASB adaptor will pass the request to ASB agent. 
If more than one ASB agent is available then the load balancer in the ASB adaptor is used 
to pass the request to the correct ASB agent. Once the ASB agent receives the request it 
will be placed in the operation queue until a slot is available in the pipeline. When a slot is 
available, the request will be moved from the operation queue to the pipeline. The request will 
then be processed by DataStage jobs or Queries. See Figure 2.

Processing a Service Request using a Datastage Job 

In the DataStage job, (See Figure 3), a service request enters through WISD_Input 

stage (as arguments in columns tab).  Argument values are interpreted as rows, and 

mapped to the other stages for processing. Finally, the processed data rows are 

provided to the client application through the WISD_Output stage. For service-oriented 

architecture implementation, jobs need to be always running (Always-on behavior). 

Jobs can also have multiple instances to handle/process many requests.

Role of ASB Agent to Handle Job Instances 

ASB agent plays a key role in handling a service request, ASB agent starts the job 

instances ahead of the service requests to get rid  of latency, and to deliver a high 

throughput rate in messaging/communicating with the service client. 

The ASB agent takes care that a minimum number of active instances are available for 

service requests. At the same time, ASB agent also  reduces the active job to a 

minimum limit. ASB agent never starts a new instance when a maximum active limit is 

attained. 

Summary 

The ASB agent is part of the ISF – Information Service Framework, that provides a 

means of passing requests from the ASB server (i.e. the location where the IBM 

Information Server is installed) to the remote client (i.e. to where the ASB agent is 

located) without the need of a complete J2EE Application Server at each client location. 

To sum up, the foremost task of the ISF agent (ASB agent) is to route requests and 

return responses. This is accomplished by ASB agent serving as a communicator 

between the service tier and engine tier.  This communication occurs via a background 

process that runs on the IBM Information Server.  Finally, ASB agent is only needed 
when the WebSphere DataStage and Application Servers are on separate machines.