|By Srinivasan Sundara Rajan||
|July 13, 2012 11:45 AM EDT||
Over the last couple of years much has changed in the way enterprises map their enterprise IT architecture and realize their business capabilities. Till 2010 or so, enterprise architecture have been purely built on:
- On-Premise Packaged Applications (SAP, Oracle, Siebel)
- On-Premise Custom Applications Frameworks (.NET, Java EE)
- Structured Data (Oracle, SQL Server, DB2)
- Traditional Access methodologies (Web, Client/Server, CITRIX, other reporting tools)
However, now the enterprises have to rethink their reference architecture to fit all the new avenues opened due to the advances in these areas.
The New Enterprise Reference Architecture
The following diagram outlines the New Enterprise Reference Architecture. The following text provides a detailed explanation of the role of each layer and product mapping to current vendors. The product mapping may not have covered all the products, please write to me if any of the significant products are missed out.
Traditional Enterprise Structured Data Layer
This has been in existence all along in the enterprises, with databases like Oracle, SQL Server, DB2 consists of them. This layer will continue to exist in the new enterprise architecture also, due to way several of the existing applications have been architected and also due to the robust and proven scalability capabilities of this layer.
Extended Enterprise Real Time Big Data/NO SQL
While the term ‘Big Data' may be used in different context in different enterprises, the following definition holds good to describe this layer.
Data storage defined in order of PETA BYTES, EXA BYTES and much higher in volume to the current storage limits in enterprises which is TERA BYTES.
Generally it is considered as Unstructured data and not really falling the under the relational database design which the enterprises have been used to:
- Data Generated using unconventional methods outside of data entry like, RFID, Sensor networks, etc.
- Data is time sensitive and consists of data collected with relevance to the time zones
- Data is stored in no traditional SQL format (like columnar format)
- Data is textual in nature
Some of the products that can be mapped to this layer are:
- Oracle NOSQL Database / Big Data Appliance
- IBM InfoSphere Big Data Platform which includes Hadoop
- Analytical databases like EXASOL which support Big Data Analytics
- ParAccel Analytical Database
- SYBASE IQ Columnar Database
- HP Vertica Big Data Analytics Platform
Extended Enterprise Rich Media Unstructured Data Layer
With 80% of enterprise data consists of unstructured data like office documents, emails, images, audio, video, engineering drawing files, etc., this layer manages the storage of those unstructured information. Products like Documentum, IBM FileNet, Microsoft Sharepoint comes under this category from product mapping perspective.
Context Aware Content Layer
This layer provides automatic understanding of unstructured data. Text Analytics automatically identifies and tags key attributes and entities within unstructured content by crawling almost any content source and identifying key words and phrases. There are many automated content understanding options like:
- Automatic Linking
- Automatic Categorization
- Automatic Clustering
Can be achieved through this layer. Important products that can be mapped to this layer are Oracle ENDECA Platform, HP's Autonomy Platform, IBM Content Analytics and Documentum Content Intelligence Services.
Data Virtualization Layer
With data from disparate sources at stake, there needs a common integration layer for them and that is provided by Data Virtualization Layer. Data Virtualization is the new concept that allows enterprises to access their information contained in disparate data sources in a seamless way. Product mapping can be done with Composite Software, Denodo Platform and Informatica data virtualization solution.
Enterprise Search Layer
While the data virtualization layer provides a common layer to access all the disparate data sources, we still needed robust searching capabilities on top of it and hence this layer is important. Some of the attributes of this layer are :
- Keyword based search
- Auto Correction
- Thesaurus expansion
- Relevance Ranking
This layer works closely with the context aware content layer. Products like Microsoft FAST Search Engine, Google Search Engine will fall under this category.
Transaction Processing Layer
The transactional updates to the underlying data sources will go through this layer. The main advantage being, Data Quality, Master Data Management and other integrity related checks can be performed in this layer. Most data virtualization platforms provide this feature.
On-Premise Packaged Applications Layer
During the last decade majority of the business processes for large enterprises were taken care by the packaged applications. While SaaS has taken root in most enterprises, these packaged applications will continue to exist in the new enterprise also. SAP, Oracle, PeopleSoft and similar packages are products that represent this layer.
On-Premise Legacy Applications Layer
While the packaged applications like SAP have been used for standard and proven business process, most enterprises have specific inventory management or order processing needs and hence over the period there is a huge set of applications written in platforms like .net, Java EE, Cobol etc.. have been created, due to the way these applications have been hard coded with the enterprise needs these cannot be totally replaced with SaaS model. Also some of the core intellectual property related business logic do exist in this layer which prevent replacing them.
Saas / BpaaS Layer
Though expected as part of enterprise transformation, we are seeing a much faster adoption of Business Process As A Service AKA Industry SaaS , model into large and medium sized enterprises.
We are seeing that most enterprises looking for Cloud Based Solutions for the new business process areas, while looking for options to migrate On premise packaged software into their Cloud equivalent. Workday, SuccessFactors,AdaptivePlanning, SalesForce, Netsuite are just a partial list of players in this layer.
Enterprise Integration Layer
With new enterprises adopting hybrid model of running certain business processes On Premise and some on the cloud, it requires strong integration layer to ensure that the end to end processing is taken care. The Integration platform could be On Premise like TIBCO, Oracle Fusion , WebSphere ESB etc... and also the new set of Integration Platform As A Service (IPaaS) like Dell Boomi, Mule, Informatica IPaaS are also becoming part of new enterprise.
Platform as a Service Layer
More and more enterprises started realizing the maintenance and upgrade issues related to their platforms and started adopting platform as a service for hosting new applications. The success of Windows Azure, vFabric, OpenShift, IBM PaaS, Oracle PaaS all point to a large adoption of PaaS in the enterprise. This layer will be part of a new enterprise application landscape.
Traditional On Premise Platform Layer
Certain ad hoc self-service mechanism like self-service reports, user generated reports are served using platforms like Cognos Query, Business Objects Adhoc Reporting, etc. These kinds of platform can continue to exist in the new enterprise too, even though it is highly desirable to move them to PaaS layer.
Enterprise Mashup Platform Layer
Enterprise Mashup Platforms enable rapid assembling of multiple information source and present the output to multiple UI devices, without custom coding. These platform achieve them with a combination of:
- A server side, runtime repository
- Visual tools for creating mashups for client side applications
- An enterprise app store for clients to find the required apps and use them in various platforms and devices
- An industry standard specification to build applications and widgets
Platforms like JackBe Presto, IBM Mashup Center form part of this layer.
Mobile Enterprise Application Platform Layer
A mobile enterprise application platform (MEAP) is a comprehensive suite of products and services that enable development of mobile applications. There are multiple types of enterprise mobile applications:
- Native Applications: That are developed for specific platform and device and installed on the device.
- Web Applications: Typical enterprise web application that can be accessed by the web browser available in the handheld device
- Hybrid Applications: As per the name, it combines both native and web applications. Hybrid Mobile Applications are essentially mobile websites running within an application that has a single web view control.
This layer is very important in the new enterprise to avoid custom development of mobile applications in native platform, rather to create hybrid applications. Platforms like, Convertigo, WorkLight, Syclo form part of this layer.
Enterprise Access Layer
As the final layer naturally this layer consumes the information from rest of the layers. While the old enterprise has got Windows, Web based clients, the new enterprise has got various kind of Mobile platforms, Portals along with the Windows and web based clients.
While the complexity of new enterprise architecture has increased, but it has come up with new possibilities for enterprises, for example, now an enterprise can serve his customers, with ease:
- A particular business process to an end-user iOS or Andriod device
- While the business process is a mashup between an existing on-premise service along with SaaS-based service
Yet the complete process can be done in a seamless way with the support of various layers in the New Enterprise Reference Architecture.
- Java EE 7 and Cloud Computing
- Cloud Computing Reference Architecture – Review of the Big Three
- Windows Azure vs VMware vFabric
- PaaS: .NET vs Java EE
- Using Amazon Elastic MapReduce in the Automotive Industry
- Five Factors to Influence Cloud Adoption – The Pros and Cons
- Dynamic Scaling and Elasticity - Windows Azure vs Amazon EC2
- Cloud Analytics - The Big Four Offerings
- Enterprise Java EE PaaS - OpenShift vs Google App Engine for Java
- Challenges and Solutions for the Health Care Industry in Cloud Computing