This is the second release of the new Taverna 2.4 Server. We welcome feedback on both the things that are there and the things that are not.

Key Features

• Runs arbitrary Taverna 2 workflows
• Based on Taverna 2.4
• REST and SOAP interfaces
• All functionality available through both interfaces
• Manages files for workflows
• Make files, read files, delete files
• Create subdirectories, list directory contents
• Can download a whole directory structure as a ZIP
• Tidies up when workflow runs expire
• Expiry time fully configurable
• Can force immediate deletion of a workflow run
• Security
• Encrypted communication supported
• Multiple users
• Users isolated from each other (via sudo)
• Workflows isolated from server
• Users may grant access to other users
• Specify credentials for workflows to access back-end services
• Notification framework
• Inform users when workflows terminate
• Many protocols: Atom feed, email, SMS, Twitter, Jabber
• Management interface
• Administrative access to all server’s tunable parameters
• Authenticated web interface, JMX
• Usage monitoring/accounting
• General quality improvements
• Improved robustness
• State can persist over (limited) server restarts
• Improved installation
• Self-contained server package
• Support for transfer of large data files

Planned Future Features

• Support for Cluster Deployment
• Access to Provenance
• Support for Interacting with Workflows

Specific Issues Addressed in This Release

See http://www.mygrid.org.uk/dev/issues/browse/TAVSERV

The funding is flexible, and can be used to pay for travel to conferences, setting up and running workshops, and starting new collaborations. In return for funding, Fellows keep the Software Sustainability Institute up to date with the latest news from their discipline. More information about the Fellowship is available on the Fellowship page.

The Software Sustainability Institute is a national facility for cultivating world-class research through software. It comprises experts from the universities of Edinburgh, Manchester, Oxford and Southampton.

To install the patch

1. Start Taverna 2.4. It is best to install the patch when no workflows are being edited or run.
2. Click “Advanced” on the top menu and select “Updates and plugins”; the Updates and plugins dialog will appear.
3. Click “Find updates” within the dialog. You should now see that updates are available for four parts of Taverna. You need to install all four updates for Taverna to be patched correctly. Do *not* restart Taverna until all four updates are installed.
4. Choose “External tool service” and then click “update” on the right. The Taverna user interface may be unresponsive while the update is downloaded. Do not quit Taverna.
5. When the update has finished, Taverna will show a dialog warning that the update will not take effect until Taverna is restarted. Do not restart Taverna yet. “External tool service” should now show that it is version 2.4.1
6. Update “Service catalogue”, update “Services” and then update “Workbench” in a similar manner
7. Restart Taverna.

Note that the title bar will still say Taverna 2.4.0

The patch includes:

• A fix for a problem where the correct connection would not be made within the diagram when connecting without ports being shown
• A change to the details for text constants so that HTML and large data will not cause problems

Here are a number of new features in the Workbench:

• Text Constant service allowing easier entry of large textual values
• Greater control over what is return from REST service
• Ability to use all the example values for a workflow run with one button
• Easier annotation of services and control over their list handling

Taverna 2.4.0 also includes the Command Line Tool that allows you to execute workflows outside of the Workbench. The Command Line Tool is available either as a stand-alone download or bundled with the Workbench.

Tutorials and examples

Tutorials for Taverna are available under the Taverna 2.x documentation.

Example workflows can be found under the Starter pack tab in the myExperiment perspective of the Workbench. They can also be downloaded from the myExperiment Web site at http://www.myexperiment.org/packs/254.

Known issues

Although Taverna 2.4.0 contains many improvements there are some known issues with it. These are detailed in known-issues.txt in the download bundle. An updated and more detailed list can be found in our issue tracker Jira at:
http://www.mygrid.org.uk/dev/issues/secure/IssueNavigator.jspa?mode=hide&requestId=10260.

System requirements

Details of Taverna’s system requirements are available on-line.

Licence and source code

Taverna is distributed under the LGPL 2.1 licence.

The source code for Taverna can be retrieved from our Subversion repository. We also recommend you to sign up to the taverna-hackers mailing list if you are a developer.

Reporting problems

If you find any problems that are not listed in known-issues.txt, then please send them to the taverna-users or taverna-hackers mailing lists. Our support team is monitoring the mailing lists and will try to respond to questions as soon as possible.

If you prefer not to subscribe to our mailing lists, you can send your questions directly to support@mygrid.org.uk and a member of our support team will respond to your inquiries.

This release is a feature-complete version of the Taverna 2 Server that is provided as a basis for deployments of server-ized Taverna in a multi-user environment. This version of the Server is based on the Taverna 2.4 codebase. Detailed documentation for the Server is available on-line.

In addition to its improved performance, this release supports a number of new key features:

• Based on Taverna 2.4
• Multiple users, with strong separation between them.
• Limited persistence over service restarts, depending on exact deployment.
• Workflow run introspection capabilities; clients can ask the server what inputs they should supply and what outputs were provided.
• Workflow run termination notifications through multiple mechanisms (RSS feed, email, SMS, twitter, etc. depending on deployment).
• Security, both of access to the service and access by the workflow runs to other services.
• Administrative REST interface including resource accounting
• Streaming of large files both for download and upload.

This is in addition to the features supported by the previous version of Taverna 2 Server:

• Upload and Execution of arbitrary Taverna 2 workflows
• Access to Workflow’s Interim Output Files; no need to wait for the workflow to finish if the results are available sooner
• Safe File Management for handling results; workflows cannot interfere with each others files
• Simple mechanism for Removal of Expired Workflows
• Support for both RESTful and SOAP APIs, for easier tooling
• JMX-based Management API

A Ruby Gem for interfacing to the Server via its REST interface is available. A Virtual Machine and Demonstrator Interface will be released soon.

Specific Issues Addressed in This Release

• TAVSERV-224 Ability to pass configuration parameters to the command line
  executor
• TAVSERV-226 Odd webapp authentication behaviour!
• TAVSERV-228 Wrong title and missing documentation link on index page

Future improvements

There remain a number of known-missing features; notably these include access to:

• Provenance Database
• Workflow Execution Model

Installation

Download Taverna Server 2.4 war and follow the installation instructions.

• Taverna Server 2.4 war
  • MD5: 02ca855484dee31737264a8d80f8dcbc
• Taverna Server 2.4 war source code
  • MD5: ab3f73a2e584aa1caf8a41fe12f1b522
• Taverna Server 2.4 full distribution
  • MD5: 149bf8620071af931ae0893bd2c1928b

System Requirements

Operating system

The Server can currently be installed on Linux and MAC OS.

Java

You will need a Java 6 installation. You can download Java from Sun Microsystems.

WebApp container

You will need a suitable servlet container.

The Server has been developed using Tomcat 6.0.26 as the servlet container, but other versions of Tomcat are known to work (back to at least 6.0.20) and other containers may also function correctly as no Tomcat-specific APIs are used in the deployable code. We welcome feedback on which containers work, as well as on how to configure them (if they are not Tomcat versions).

This release is a feature-complete version of the Taverna 2 Server that is provided as a basis for deployments of server-ized Taverna in a multi-user environment. This version of the Server is based on the Taverna 2.3 codebase. Detailed documentation for the Server is available on-line.

In addition to its improved performance, this release supports a number of new key features:

• Based on Taverna 2.3
• Multiple users, with strong separation between them.
• Limited persistence over service restarts, depending on exact deployment.
• Workflow run introspection capabilities; clients can ask the server what inputs they should supply and what outputs were provided.
• Workflow run termination notifications through multiple mechanisms (RSS feed, email, SMS, twitter, etc. depending on deployment).
• Security, both of access to the service and access by the workflow runs to other services.
• Administrative REST interface including resource accounting
• Streaming of large files both for download and upload.

This is in addition to the features supported by the previous version of Taverna 2 Server:

• Upload and Execution of arbitrary Taverna 2 workflows
• Access to Workflow’s Interim Output Files; no need to wait for the workflow to finish if the results are available sooner
• Safe File Management for handling results; workflows cannot interfere with each others files
• Simple mechanism for Removal of Expired Workflows
• Support for both RESTful and SOAP APIs, for easier tooling
• JMX-based Management API

There remain a number of known-missing features; notably these include access to:

• Provenance Database
• Workflow Execution Model

A Ruby Gem for interfacing to the Server via its REST interface is available. A Virtual Machine and Demonstrator Interface will be released soon.

Installation

Download Taverna Server 2.3 war and follow the installation instructions.

• Taverna Server 2.3 war
  • MD5: 4ace507cbbc7f7224bf7ad6419a25598
• Taverna Server 2.3 war source code
  • MD5: b43aff472c915d407928bf9311f15a7d

System Requirements

Operating system

The Server can currently be installed on Linux and MAC OS.

Java

You will need a Java 6 installation. You can download Java from Sun Microsystems.

WebApp container

You will need a suitable servlet container.

The Server has been developed using Tomcat 6.0.26 as the servlet container, but other versions of Tomcat are known to work (back to at least 6.0.20) and other containers may also function correctly as no Tomcat-specific APIs are used in the deployable code. We welcome feedback on which containers work, as well as on how to configure them (if they are not Tomcat versions).

SCUFL2 is the proposed new mechanism for specifying Taverna workflows, and forms the basis for our current development work on Taverna 3. SCUFL2 adopts Linked Data technology and preservation methodologies to create a platform-independent workflow language that can be inspected, modified, created and executed.

SCUFL2 comes with a Java API that can be used for programmatic access to read and write SCUFL2 workflow bundles. A workflow bundle is a structured ZIP file with the workflow definitions included as RDF/XML documents.

The workflow structure is defined using an OWL ontology and annotated with URIs so that third parties can form semantic statements about any component of a SCUFL2 workflow, for example to state that a particular service produces outputs of a certain type, or that a data link was added by a specific researcher.

Version 0.9.2 of the SCUFL2 is the first public beta release of the Java API for working with SCUFL2 workflow bundles. The main features are:

• Standalone Java library, no external dependencies (optional dependency on JAXB)
• Import Taverna 2 workflows (.t2flow)
• Import/export SCUFL2 workflow bundles (.wfbundle)
• API for inspecting, creating and modifying workflow structures (JavaDoc)

The API has also experimental support for importing Taverna 1 workflows (SCUFL 1) and performing structural validation of a workflow. Note that the API does not provide a way to execute the workflow (Taverna 3 platform will do this) and does not currently implement export to t2flow.

The team have provided a set of example SCUFL2 programs to show how the API can be used. Here is an extract from ServiceTypes.java, which shows which service types (WSDL, Beanshell, etc) have been used in a workflow:

Set types = new LinkedHashSet();
WorkflowBundleIO io = new WorkflowBundleIO();
File file = new File(filepath);
WorkflowBundle wfBundle = io.readBundle(file, null);
for (Profile profile : wfBundle.getProfiles()) {
  for (Activity activity : profile.getActivities()) {
    types.add(activity.getConfigurableType().toASCIIString());
  }
}

Note that this is a beta release of the API. Based on your feedback, we might be have to do minor changes to some methods or the Workflow Bundle file format for the 1.0 release, but will try to keep such changes to a minimum.

The myGrid team welcome developers to try SCUFL2 and contact us on the taverna-hackers mailing list with feedback, suggestions, bug reports, and most importantly, how they use or would like to use SCUFL2.

Quick links:

• http://www.mygrid.org.uk/dev/wiki/display/developer/SCUFL2
• http://www.mygrid.org.uk/dev/wiki/display/developer/SCUFL2+API
• http://taverna.googlecode.com/svn/scufl2/ or
  https://github.com/mygrid/scufl2
• https://github.com/mygrid/scufl2-examples
• http://mygrid.github.com/scufl2/api/0.9/