1.
Aneka Cloud Application Platform: Framework Overview
The Aneka framework is built on a Service-Oriented Architecture (SOA) and consists of a
middleware that manages and scales distributed applications across heterogeneous resources
(desktops, clusters, data centers, virtual clouds).
Core Components:
o Aneka Container: The basic building block and runtime environment installed
on every node.
o Aneka Services: The core functionality housed within the container, classified
into three types.
o Aneka SDK (Software Development Kit): Provides APIs and tools for
application development.
o Management Tools (Aneka Management Studio): A GUI for administrators to
monitor, manage, and configure the cloud.
The architecture typically involves a layered approach, with the Aneka middleware residing on
top of the physical and virtual resources.
2. Anatomy of the Aneka Container
The Aneka Container is a lightweight software layer and the unit of deployment in Aneka
Clouds. It hosts the various services and interacts with the underlying operating system and
hardware via the Platform Abstraction Layer (PAL).
The services within the container are classified as follows:
Service Category Description Key Services/Functions
The lowest layer, directly Hardware Profiling, Dynamic Resource
Fabric Services
interfacing with the resource. Provisioning, Monitoring.
Membership Management, Resource
Foundation Core middleware functions for
Reservation, Storage, Accounting,
Services managing the distributed system.
Logging, Heartbeat.
Application Responsible for application Scheduling Services (for Task, Thread,
Services (Execution scheduling and execution, and MapReduce), Execution/Allocation
Services) supporting programming models. Services.
3. Building Aneka Clouds
Aneka clouds can be organized based on their Infrastructure or their Logical role, and
deployed using different cloud models.
Infrastructure Organization
The infrastructure model focuses on the physical/virtual components used for deployment. Key
components include:
Aneka Repository: Central storage for software libraries, including the Aneka Platform
installation files (images).
Administrative Console (Management Studio): Central hub for administrative
operations, managing repositories, and deployment.
Aneka Daemon (Node Manager): Remote management service installed on each node
to deploy and control container instances.
Aneka Containers/Nodes: The actual resources hosting the runtime environment.
Logical Organization
The most common logical organization is the Master-Worker configuration, often with
dedicated storage nodes.
Master Node: Hosts services that provide the intelligence and coordination for the cloud,
such as: Scheduling Service (master copy), Resource Provisioning Service,
Accounting Service, and mandatory services (Logging, Heartbeat, Monitoring).
Worker Nodes: Constitute the workforce, hosting services like: Execution Services (for
supported programming models), Allocation Service, and mandatory services.
Storage Nodes: Optimized to provide storage support, hosting the Storage Service.
Cloud Deployment Modes
Aneka supports the three standard cloud deployment models:
Private Cloud Deployment Mode:
o Utilizes local physical resources and infrastructure management software (e.g.,
local virtual machine manager).
o Best for predictable workloads where excess capacity demand is addressed
locally.
o All resources are in-house (within the enterprise).
Public Cloud Deployment Mode:
o Involves installing Aneka Master and Worker nodes over a completely
virtualized infrastructure hosted by an external IaaS provider (e.g., Amazon
EC2, Windows Azure).
o Deployment is typically contained within the boundaries of a single IaaS provider.
Hybrid Cloud Deployment Mode:
o The most common deployment, combining local infrastructure (static) with
public cloud resources (elastic).
o The local infrastructure is leveraged, and the external public cloud is used on-
demand for elastic scaling (cloud bursting) when additional resources are
required.
o Leverages dynamic resource provisioning and reservation.
4. Cloud Programming and Management
Aneka SDK (Software Development Kit)
The SDK provides APIs and libraries for developers to create distributed applications and
implement new services/programming models.
Application Model:
o Represents the minimum set of APIs common to all programming models for
building and representing distributed applications.
o Aneka supports multiple programming models as abstractions for distributed
execution:
Task Programming: For independent tasks that can run in any order
(common in Grid Computing).
Thread Programming: For multithreaded applications with finer control
over components.
MapReduce Programming: For data-intensive applications, abstracting
the processing of large datasets.
Service Model:
o Defines the general requirements and structure for implementing a service that
can be hosted and managed within an Aneka Container.
Management Tools
Administrators use these tools for setting up, monitoring, and maintaining the Aneka Cloud
infrastructure. The primary tool is the Aneka Management Studio (GUI).
Infrastructure Management:
o Deploying and controlling container instances on nodes (via the Aneka Daemon).
o Configuring the resource pools (e.g., integrating private machines, public IaaS
providers).
Platform Management:
o Starting, stopping, and configuring services within the Aneka Containers.
o User and permission management.
o Managing repositories.
Application Management:
o Monitoring the execution and status of applications.
o Viewing logging, accounting, and billing information.
o Managing job queues and scheduling.