User:Classivertsen/sandbox
 | It has been suggested that this page be merged into Grid computing. (Discuss) Proposed since September 2013. |
 |
Distributed applications or dApps for short, is a term for an application running under an operating system that substantially communicate "inter-application" between devices rather than communicate with other instances of the application in other devices rather than communicating through a data center or central server. "Distributed applications" (dApps) and "Applications" (Apps) differ only in the methodology in which they communicate with other applications of the same type. A dApps typically refer to applications running on a smartphone but may equally be an application running in Windows, Mac OS, Ubuntu or any other operating system. A dApp may equally run in a web browser as a HTML application or in a non mobile web browser or as a native Mobile application.
Distributed application using the "DEMAND computing" model
[edit]The methodology of developing distributed applications is by using the DEMAND computing model, which is an acronym for "Distributed Extensible Mobile Application Network and Data" computing. With this methodology it is no longer necessary to have a central data center to implement the functions of the application or to store data. With DEMAND computing, the processing workload is distributed across any number of nodes that make up part of the network. For specific processing tasks required by a device, a number of available nodes are requested through the DEMAND API and the Identified Available Cluster (IAC) will then be assigned the processing task.
DEMAND computing should not be confused by Grid_computing, which is a network (or grid) of the same type of devices. In the DEMAND computing model, the computing nodes are not individual compute nodes of same or similary type as in Grid_computing, but rather a partial process that runs as part of a Distributed application than runs in a variety of devices and Operating systems.
Datacenter vs. distributed communication method
[edit]The typical and traditional way applications on devices, such as smartphone communicate with either each other or a central server is to send and receive packets of data to and from a data center. With the distribute application communication method, each application communicates with instances of the same application on other devices directly.
"The end of the data center?
[edit]With the emergence of dApps some have speculated that this method of distributing data processing across a multitude of devices without relying or requiring a central data center would mean that the current data center of cloud computing model would eventually disappear. Although the emergence of the DEMAND computing model is likely to be "the next big thing" after cloud computing it is not likely that the emergence of dApps or the emergence of services using the DEMAND computing model would have a significant impact on the marketshare of the cloud and data center services in the near future. It should also be noted that this technology is far for as fault tolerant as that of a data center with redundancy and data spread over multiple geographical locations.
Services suitable for dApps
[edit]The type of applications and services best suitable for dApps using the DEMAND computing model are those with a large or quickly growing subscriber base where the access to any specific data is not mission critical. Examples of such applications would be social media applications, Messaging Applications, Internet or on-line dating applications, blogging applications, file sharing applications, torrent applications and location based services.
Distributed application icons
[edit]It would be possible for most applications and services to adopt both an "application to datacenter" model as well as a "distributed application" model. For most services it would be possible to services it would be The typical icon of a distributed application is that of a circle with three network node connections extending from it.