The Computing Architecture - Distributed vs Grid Computing

Enterprise Computing Architecture has seen enormous growth over the years. In the olden days, businesses used multiple boxes to host data and run applications.These systems were mostly driven by mainframes and client-server models. With the advent of advanced networking architectures, the enterprise systems moved away from client-server models towards web-based N-Tier models.

Along the way, there were number of changes to the way the enterprise systems performed the required functions. Distributed Computing, Grid Computing, and Virtualization changed the outlook of how computations are done over the network. The blog-post focuses on highlighting the above concepts.

Distributed Computing:
Distributed Computing is basically "distribution" of a "specific function" over multiple nodes in a network. Distributed computing can be achieved from client-server, 3-tier, or a N-Tier architecture. Whatever be the architecture, the fundamental principle is to break the program/function into pieces and distribute it to multiple nodes for computation. The distributed systems are highly scalable as the system can easily be expanded by adding more machines as needed (Scale out). Further, since all the nodes can perform the same function, even if one of the nodes fail, the system is still functional and reliable.

There are several considerations an architect should review with distributed computing: Network Latency, Bandwidth, Topology, Homogeneous vs Heterogeneous, Security, interoperability. Applications built using JEE framework provides for distributed computing requirements. The  JEE platform allows users to build web services that interoperate with other legacy systems. Moreover, The applications developed for distributed computing should also be open and loosely coupled so as to interact with other applications/systems in the environment. Web Services applications built using WSDL, SOAP, XML provide language and platform independence.

Grid Computing:
Grid Computing is based on concepts such as virtualization, clustering, and dynamic provisioning of IT resources. It involves pooling of IT resources and creating a virtualized environment. Thus, Grid computing allows for flexibility, improved utilization, and lowered operational cost as the underlying infrastructure is shared by the applications.

It is important to understand the key difference between Grid and Distributed computing. While Distributed computing is primarily about application design, Grid computing is about pooling of infrastructure for efficient use. Basically, distributed computing and grid computing can co-exist in IT architecture.

Grid computing can be broadly categorized as Application Grids and Database Grids. Application Grids usually consists of SOA Services and Applications, Data Services (for persistence, events, messaging, analytics), Containers that hold JVM and native platforms, Grid Controls and Management, IDE and Configurations tools. Database grids consists of DBMS applications and storage.

The Modern IT Era:
From accidental architectures of the past, we are moving towards a sophisticated IT architecture that promotes Scalable, Interoperable, and Reliable systems. In the modern era, infrastructures that makes use of CLOUD architecture is becoming the preferred choice for most businesses. From a technical standpoint, CLOUD is nothing more than a sophisticated GRID. However, from a business standpoint public cloud computing helps to reduce the operational cost significantly. While there are threats related to data security and privacy, there is a clear case for businesses to pursue a cloud strategy to stay competitive in its industry.

References:
http://www.oracle.com/technetwork/topics/entarch/itso-165161.html
http://insidehpc.com/2006/07/14/what-is-a-clustered-database/
http://publib.boulder.ibm.com/infocenter/txformp/v6r0m0/index.jsp?topic=%2Fcom.ibm.cics.te.doc%2Ferziaz0015.htm




Comments

Post a Comment

Popular posts from this blog

No Non-Sense Transformation

‘Context’ Many years ago, when we first started setting up shared services in a major technology firm, we embarked on a journey that had many unknowns. Whether it was scope of work, projects, technology, or processes – nothing was clearly outlined. We had few excel sheets to guide us through the process of setting up shared service organization. It was no more than a simple ‘lifting’ of processes from each entity into a regional centre. The initial days were tough – It was about power, control, and other hidden aspects in organization. The lower end resources who managed the day-to-day operations struggled to meet the expectations from the internal clients. We had ‘all fingers pointing at one direction: shared services’, and always for the wrong reasons. ‘No defined parameters’ Now, when I look back to the unwieldy phase – it appears that there were number of interventions that could have been employed to mitigate the organizational resistance. In my experienc...
Read more

Patrick Pichette and Art of Corporate Finance in IT Industry

I was reading through Patrick Pichette's interview with Mckinsey Quarterly last week, and it was perhaps one of the few interview reports that made me sit back, read, read, and read through again. Its quite an old interview, but worth looking given its relevance in today's fast changing IT landscape. The report certainly made me to think about many aspects of corporate finance and strategy. Patrick's  ideas are certainly commendable and thought provoking - From managing Short Term metrics to Start-Up culture. Here are some excerpts and my thoughts... 'Managing Short Term/Tactical metrics' Managing short term metrics is part of corporate life once an organization goes public. There is a general expectation from investing community to meet/beat analyst estimates and that puts a lot of pressure on C-Level executives. Nevertheless, it is required of C-Level executives to meet short-term goals as well as long-term vision. In the report, Patrick suggests that short...
Read more

The RASSP in IT Architecture

The QUALITY Attributes are those that defines how the system should perform over a business cycle/Time Period. Fundamental to good IT design is forecasting the business peaks and troughs and how the system should behave over the business cycle. When there are peak loads, systems should automatically be able to scale and perform with consistent service level agreements. However, there are several cases in the past where such quality attributes were ignored leading to loss of revenue. Quality Attributes can be either run-time or non run-time attributes. Clearly, from a business standpoint, the run-time consideration is of prime importance as no one would want to lose business because of a system downtime or a performance lag. The blog post focuses on the run-time considerations an architect should adhere when making architecture decisions. The following attributes are considered the most important run-time attributes when designing new systems. Reliability (R), Availability (A), Sca...
Read more