Design and Performance Evaluation of a Service-Oriented HLA RTI on the Grid - Grid Computing: Infrastructure, Service, and Applications - page 452

Federate1
LRC

Federate2
LRC

LRC : Local RTI
Component

Invoke services or
get callbacks

Invoke services or
get callbacks

LS1 LS2

LFS LIS LFS LIS

LRIs Interactions

or attribute updates

LRIs

RTI index service

FMS
FMFS

DMS OOMS

OOMFS FMIS DMFS DMIS OOMIS

FMRIs DMRIs

OOMRIs

TMS DDMS

DDMIS

TMFS TMIS DDMFS

TMRIs DDMRIs

LS: Local Service FMS: Federation

Management Service
TMS: Time
Management Service

DMS: Declaration
Management Service
DDMS: Data
Distribution
Management Service
(X)RIs: (X) Resource
Instances

OOMS: Object and
Ownership
Management Service
(X)FS: (X) Factory
Service

(X)IS: (X) Instance
Service

FIGURE 18.1 Framework overview.

The LS is used as a messaging broker of federates and consists of the
factory service LFS, the instance service LIS, and multiple resource
instances LRIs, with one LRI for each federate. A federate communicates
with the outside world through its LRI by invoking services and getting
callbacks. As shown in Figure 18.2 , the LRI is structured into six modules.
Each of the modules except the callback module corresponds to one of the
management services. The callback module is used to buffer callbacks for
a federate.
The objective of separating the HLA service groups into different grid
services and using a modular structure for the LRI is to create a plug-and-
play paradigm for an HLA RTI implementation so as to build an extensible