As described for the Taverna command line utility, sometimes it may be impractical to run Taverna workflows within the Taverna workbench. Examples of such scenarios are when running a workflow that may take a long time to complete, or requires greater resources than is available on a typical desktop or laptop computer.
If you are working alone, and want a quick simple solution to remote execution then the Taverna Command Line utility would most likely be the preferred option. If you are wanting to create a long term solution that is to be used collaboratively as part of a wider team then the Taverna Server is going to be more suitable.
The Taverna server is a more sophisticated solution to running Taverna workflows remotely than the Taverna command line utility. In particular it allows you to share your workflow execution and results with others, and also provides an Application Programming Interface (API) to allow you to integrate Taverna workflow execution into you own application or pipeline. The interface supports both a Representation State Transfer (REST) and Simple Object Access Protocol (SOAP) style of interface, and utility libraries are available for the Ruby and Java programming languages.
Its important to note that the Taverna server itself doesn’t provide a web User Interface other than for administration. However, there is a demonstrator Taverna Portal that provides a good starting point and example for creating a full web interface.
The server is available as a Web Application Archive (WAR file), designed to be used with the Apache Tomcat Servlet engine. Additional technical details about the server can be found within the Taverna User Guide
The Taverna server has been used within the Amazon Elastic Cloud for processing the results of Next Generation Sequencing. In collaboration with the University of Liverpool’s Animal Sciences & Physiology Research group, in particular Dr Harry Noyes, Taverna scientific workflows were combined with the computing power from the Amazon cloud to create a powerful next generation sequencing application for whole genome Single Nucleotide Polymorphism (SNP) analysis.
Through a Web Portal, the application allows scientists to upload their input data, fire off a number of parallel cloud instances for the analysis, whilst monitoring progress and collecting results. Each cloud instance was created using the Taverna server.
More details on this piece of work is described on the Taverna website.