Reason behind the restriction of having either an inference agent or a process agent in a cluster, but not both.

Reason behind the restriction of having either an inference agent or a process agent in a cluster, but not both.

book

Article ID: KB0090525

calendar_today

Updated On:

Products Versions
TIBCO BusinessEvents Enterprise Edition -
Not Applicable -

Description

Resolution:
Description:
============
TIBCO BusinessEvents - Process Orchestration document says that a cluster can have either an inference agent or a process agent, but not both. Do not mix a stand alone inference agent and process agent in a cluster.

Environment:
==========
TIBCO BusinessEvents 5.1.x
All Operating Systems


Resolution:
=========
You can have only an inference agent or a process agent in a cluster. If needed you can have an Inference Task in your process agent, which is a simple task that is designed to run a set of rules on the Job Context variable data objects until the completion of a single RTC cycle.

If some concepts and/or events were used by different agents in the cluster then there is a strong possibility that the sequence  ID (internal ID) for the newly created objects persisted in cache could be identical, causing undesirable results. Hence the restriction.


References:
==========
TIBCO BusinessEvents® Process Orchestration 1.0.0 - Manual

Issue/Introduction

Reason behind the restriction of having either an inference agent or a process agent in a cluster, but not both.