Today, a broad variety of Internet-based applications such as Flickr and Facebook are now available. They allow users to maintain profiles, participate in communities, create and post content, and perform a wide variety of other tasks. Some applications even mashup business logic and services from different sources. Tim O’Reilly, a well-known Silicon Valley investor and blogger, has defined this new phenomenon as follows:
Web 2.0 is the business revolution in the computer industry caused by the move to the internet as platform, and an attempt to understand the rules for success on that new platform. Chief among those rules is this: Build applications that harness network effects to get better the more people use them. (This is what I've elsewhere called "harnessing collective intelligence.")
A basic level of standardization is often the first step and is an important requirement for more powerful data integration services to emerge. In other words, data portability is the precursor to data integration. Today, new organizations and committees aimed at standardizing representation of data on the Internet are already beginning to make data portability a reality.
New Standards Emerge
Dataportability.org is one of the more interesting groups driving data portability today. Originally, this organization focused on putting users in control of their own data, as it accumulated across various Internet-based services. However, the organization now has a broader goal, and champions overall standards for Web data portability. Let’s take a quick look at some of them.
OpenID. OpenID aims to be a free and simple way to use a single digital identity across the Web, and has been adopted by Yahoo, AOL, Google, Microsoft, MySpace, Orange, France Telecom and many other providers. More than half a billion people now have an OpenID identifier, which should further encourage active adoption of this standard.
OAuth. OAuth is defined as a simple way to publish and interact with protected data on the Internet.
RSS. RSS stands for Really Simple Syndication. It’s a dialect of XML that is often used to share content on the Internet, signified by the orange RSS icon you’ll spot on numerous sites.
CMIS for content integration and SSE for file sharing.
OPML. OPML stands for Outline Processor Markup Language. It’s closely related to RSS, and is often used to share RSS subscription lists amongst readers and aggregators, which typically allow users to export and import RSS subscriptions and Web bookmarks as an OPML file.
A Data OASIS
Dataportability.org is just one of many organizations promoting standardization of data on the Internet. OASIS (Organization for the Advancement of Structured Information Standards) is another prominent player and has sponsored many other standards, most notably CMIS or Content Management Interoperability Services.
As content and services providers continue to embrace the Web as their platform, the same data portability and integration problems that emerged in the enterprise will appear again. New Internet-based tools and services that can connect disparate data islands and providing transparent and secure access to Internet data regardless of where it is stored will be essential to making Web content and services truly accessible - and standardizing data-related format and protocols will be a critical first step.
Interesting new standards such as OpenID, OAuth, CMIS and others have already begun to emerge. These standards are a huge step forward, but much more must be done to improve these standards and ensure they are supported by new and existing Internet-based services. Until this is accomplished, data integration in the Internet will remain a big issue. However, a big problem is usually a big opportunity in disguise, and many companies will undoubtedly rise to the challenge and fill this important need.
Tuesday, September 29, 2009
Tuesday, September 15, 2009
Integration at Home - SOA
Overview on Home Integration
It all started when Microsoft came out with Universal Plug and Play (UPnP) capabilities to allow users to dynamically plug-in new devices such as mouse, keyboards, external storage, etc. As broadband started being widely adopted and people started having more than one PC at home, the network/wireless hubs also became popular within households. This required the device manufactures to adopt a standard for dynamic discovery, control, event notification, presentation, etc. As the PCs and all their peripherals were already supporting UPnP, it became the choice by default. UPnP is the architecture for peep-to-peer networking of PCs, networked appliances and wireless devices. It is a distributed, open architecture based on TCP/IP, UDP and HTTP.
Next came all the gaming consoles with internet connections and, as the networks hubs (both wired and wireless) were the gateways to the internet, these devices also started supporting UPnP. After the first major release of the gaming consoles, the subsequent releases started adding additional multimedia capabilities. These same capabilities were also available on PCs (and Macs) the storage requirements for households increased exponentially to store all the picture, music and video, personal documents, etc. Today these storage devices also come with multimedia and Web servers to present the content on demand to the user. Once again, the device manufactures leveraged UPnP not only to discover the devices and services but to control access to remote drives, files and control the audio and video play.
At the same time, the mobile phone devices had implemented Bluetooth a low-power, short-range wireless personal area network for exchanging information between devices such as mobile phones, laptops, PCs, digital cameras and video game consoles. As this was short range, most of the Bluetooth devices were the clients interacting with a server such as PCs and game consoles sending commands. These servers would, in turn, leverage UPnP if they needed to access or control remote content.
Until recently, the home automation technology was very expensive and not affordable by most of the households. With the introduction of ZigBee and Wibree low cost chips, the home automation industry has taken off. Home automation is not just for environment control such as light and temperature control, it has also expanded to control audio and video devices – all based on the IP networks. Basically the chip technology allows for control from the server to the client such as light switch, garage door and temperature controller. The servers leverage UPnP to control access to remote services, those not directly connected to the server the user is interacting with.
Basically UPnP is the underlying protocol that provides peer-to-peer networking capabilities within the household. However, due to some limitations of this protocol, Microsoft along with some of its partners proposed a new device profile Web services (DPWS) standard to replace UPnP. As this is not currently backward compatible with UPnP, all vendors are expected to support both UPnP and DPWS until there is convergence.
Independent of the protocol, the home integration is now real which unifies the PCs, gaming consoles and home automation from a console on any device.
Peer-to-Peer Network Overview
The concept of the P2P network is pretty straight forward. Whenever a peer joins a network, it multicasts its availability which triggers interaction between it and the other peers to acknowledge its existence and maintain the peer map. In addition, the peers may exchange device description, control information such as commands available. Once the peer network is established, the clients on any of these devices could control any remote device/resource. All these remote devices/resources could also be referred to as services available on the peer network.
One of the major limitations today is that these protocols used at home only support access from within the local network, i.e., the user cannot access these services from outside the house. The vendors are now actively working on expanding the standard to support remote secure access to these personal home integration services.
Relevance to the Enterprise
With all the major enterprise software vendors backing the services component architecture (SCA), there is now a single standard for all the services and their dependencies. The SCA defines only the metadata to describe the service and its dependencies with most of the vendors initially leveraging this in their design/development tools to enable architects model the service assembly. However, this does not yet address the IT operations administration nightmare of reconfiguring the impacted services whenever the deployment model changes.
The vendors are expected to address this administration overhead in their subsequent issues by developing/incorporating the SCA container. In this approach, the development teams focus on developing the services and invoking external services with the right interfaces but not focus on the bindings. In addition, IT operations no longer need to reconfigure all the services. The SCA container dynamically discovers all the relevant services using the similar approach used for home integration. The UDDI standard is not sufficient to enable dynamic discovery and some of the current Open Source SCA container projects have started leveraging peer-to-peer technologies such as Jini and JXTA. This dynamic discovery shall make it simpler and easier for IT operations to manage all the service dependencies and shall also reduce the need of an enterprise service bus.
Similar to social networking site, which drove the innovation of Web 2.0 solutions in the consumer space and now impacting the enterprises, the integration-at-home approach should be adopted by IT operations to manage and monitor all the service dependencies in production. As WS-Discovery is leveraged by DPWS for dynamic discovery of services at home, it would make sense for the ISVs to simplify the convergence between personal and work services for the user.
reference: informationmanagement.com
It all started when Microsoft came out with Universal Plug and Play (UPnP) capabilities to allow users to dynamically plug-in new devices such as mouse, keyboards, external storage, etc. As broadband started being widely adopted and people started having more than one PC at home, the network/wireless hubs also became popular within households. This required the device manufactures to adopt a standard for dynamic discovery, control, event notification, presentation, etc. As the PCs and all their peripherals were already supporting UPnP, it became the choice by default. UPnP is the architecture for peep-to-peer networking of PCs, networked appliances and wireless devices. It is a distributed, open architecture based on TCP/IP, UDP and HTTP.
Next came all the gaming consoles with internet connections and, as the networks hubs (both wired and wireless) were the gateways to the internet, these devices also started supporting UPnP. After the first major release of the gaming consoles, the subsequent releases started adding additional multimedia capabilities. These same capabilities were also available on PCs (and Macs) the storage requirements for households increased exponentially to store all the picture, music and video, personal documents, etc. Today these storage devices also come with multimedia and Web servers to present the content on demand to the user. Once again, the device manufactures leveraged UPnP not only to discover the devices and services but to control access to remote drives, files and control the audio and video play.
At the same time, the mobile phone devices had implemented Bluetooth a low-power, short-range wireless personal area network for exchanging information between devices such as mobile phones, laptops, PCs, digital cameras and video game consoles. As this was short range, most of the Bluetooth devices were the clients interacting with a server such as PCs and game consoles sending commands. These servers would, in turn, leverage UPnP if they needed to access or control remote content.
Until recently, the home automation technology was very expensive and not affordable by most of the households. With the introduction of ZigBee and Wibree low cost chips, the home automation industry has taken off. Home automation is not just for environment control such as light and temperature control, it has also expanded to control audio and video devices – all based on the IP networks. Basically the chip technology allows for control from the server to the client such as light switch, garage door and temperature controller. The servers leverage UPnP to control access to remote services, those not directly connected to the server the user is interacting with.
Basically UPnP is the underlying protocol that provides peer-to-peer networking capabilities within the household. However, due to some limitations of this protocol, Microsoft along with some of its partners proposed a new device profile Web services (DPWS) standard to replace UPnP. As this is not currently backward compatible with UPnP, all vendors are expected to support both UPnP and DPWS until there is convergence.
Independent of the protocol, the home integration is now real which unifies the PCs, gaming consoles and home automation from a console on any device.
Peer-to-Peer Network Overview
The concept of the P2P network is pretty straight forward. Whenever a peer joins a network, it multicasts its availability which triggers interaction between it and the other peers to acknowledge its existence and maintain the peer map. In addition, the peers may exchange device description, control information such as commands available. Once the peer network is established, the clients on any of these devices could control any remote device/resource. All these remote devices/resources could also be referred to as services available on the peer network.
One of the major limitations today is that these protocols used at home only support access from within the local network, i.e., the user cannot access these services from outside the house. The vendors are now actively working on expanding the standard to support remote secure access to these personal home integration services.
Relevance to the Enterprise
With all the major enterprise software vendors backing the services component architecture (SCA), there is now a single standard for all the services and their dependencies. The SCA defines only the metadata to describe the service and its dependencies with most of the vendors initially leveraging this in their design/development tools to enable architects model the service assembly. However, this does not yet address the IT operations administration nightmare of reconfiguring the impacted services whenever the deployment model changes.
The vendors are expected to address this administration overhead in their subsequent issues by developing/incorporating the SCA container. In this approach, the development teams focus on developing the services and invoking external services with the right interfaces but not focus on the bindings. In addition, IT operations no longer need to reconfigure all the services. The SCA container dynamically discovers all the relevant services using the similar approach used for home integration. The UDDI standard is not sufficient to enable dynamic discovery and some of the current Open Source SCA container projects have started leveraging peer-to-peer technologies such as Jini and JXTA. This dynamic discovery shall make it simpler and easier for IT operations to manage all the service dependencies and shall also reduce the need of an enterprise service bus.
Similar to social networking site, which drove the innovation of Web 2.0 solutions in the consumer space and now impacting the enterprises, the integration-at-home approach should be adopted by IT operations to manage and monitor all the service dependencies in production. As WS-Discovery is leveraged by DPWS for dynamic discovery of services at home, it would make sense for the ISVs to simplify the convergence between personal and work services for the user.
reference: informationmanagement.com
Tuesday, September 1, 2009
Presentation for week three: System Integration Models
Here was our first presentation where we talked about the general topics and styles of System Integration. We also talked about a lot of topics that will be discussed later by other teams. If you are on one of those teams and we stole your thunder, we would just like to say: "Our bad, beer is on us."
The primary topics discussed are the main Integration Models, and the styles in which you can implement them. They are:
1. Presentation Integration
- Very simple and straight forward.
2. Data Integration
- Shared Database
- Maintain Data Copies
- File Transfer
3. Functional Integration
- Distributed Object Integration
- Message-Oriented Middleware Integration
- Service-Oriented Integration
Had a lot of great discussion and interesting talking points. Thanks everyone!
Here is a link to the presentation, hosting courtesy of our professor: http://gzheng.cis.gsu.edu/teaching/cis8020/files/sysintteam03.ppt
Subscribe to:
Posts (Atom)