Cloud-Based Infrastructure for Data-Intensive e-Science Applications: Requirements and Architecture - Cloud Computing with e-Science Applications - page 33

Although the dilemma of moving data to computing facilities or moving
computing to the data location can be solved in some particular cases, process-
ing highly distributed data on MPP (massively parallel processing) infrastruc-
tures will require a special design of the internal MPP network infrastructure.

2.6 Cloud-Based Infrastructure Services for SDI
Figure 2.5 illustrates the typical e-science or enterprise collaborative infra-
structure that is created on demand and includes enterprise proprietary
and cloud-based computing and storage resources, instruments, control
and monitoring system, visualization system, and users represented by user
clients and typically residing in real or virtual campuses. The main goal
of the enterprise or scientific infrastructure is to support the enterprise or
scientific workflow and operational procedures related to process monitor-
ing and data processing. Cloud technologies simplify the building of such
infrastructure and provision it on demand. Figure 2.5 illustrates how an
example enterprise or scientific workflow can be mapped to cloud-based ser-
vices and later deployed and operated as an instant intercloud infrastructure.
It contains cloud infrastructure segments IaaS (infrastructure as a service)

Enterprise/Scientific workflow

Storage
Data

Special
Proc 1

Data
Archive

Input
Data

Data
Filtering

Visual
Present

Special
Proc 2

Instrum.
Data

Campus A

Campus B

Visuali-
zation

Visuali-
zation

CE

CE

User User

User
Group A

User
Group B

Cloud 2 PaaS

VR6

VR2

VR7

VR4

VR1

Enterprise/Project-based
Intercloud Infrastructure

VR5

Resource/
Service
Provider

VR3

Cloud 1 laaS

Resource/
Service
Provider

Cloud PaaS Provider

Cloud laaS Provider

FIGURE 2.5
From scientific workflow to cloud-based infrastructure.