"Seegrid will be due for a migration to confluence on the 1st of August. Any update on or after the 1st of August will NOT be migrated"

VGL Roadmap (post-Nectar Project)

The NeCTAR funded VGL project concluded on the 1 September 2013. The CSIRO and its collaborators are committed to maintaining the VGL and hosting of it on NeCTAR infrastructure out till Dec 2014. In addition to this, various groups (including csiro) have a vested interest in pursuing activities that utilise the VGL for its originally intended purpose and "repurposing" it for other science domains. Below outlines so of the outstanding issues, features, functions that related to the VGL project and the underlying

VGL Outstanding issues (20 Sept 2013)

  • More data and More computational codes
  • plus - other bugs in internal JIRA system

Generic VL Infrastructure

  • Marketplace - codes/toolboxes and templates for codes
  • Utilisation of other cloud providers - Commercial (Amazon) and CSIRO Cloud
  • Cloud Cluster support - parallel processes in the cloud
  • Replicate VGL in something else - use the VGL infrastructure via another client application
  • Data selection info needs improvement at a map level
  • Data template selection/streamlining the workflow process -> at present users select data and then a process. There is a expectation that users know what data is required for what process, perhaps some smarts can be applied to suggest a process based on the type of data selected, or; select process first and then show data that is relevant to that process?
  • Make available any data (ie: allow data other than geophysical to be included in the VGL)
  • Provenance -
    • move to own service
    • show prov records on a map
  • AAA - need service support for authorisation
  • Github repository (migrate from svn to github)

Other requirements for Virtual Laboratories

  • Data management
    • Upload of new input data that becomes available to all users
    • Capture of output metadata sufficient to allow use of output data as input
  • Workflow
    • Orchestration and execution planning of computation on multiple resources to chain inputs and outputs
    • Parameter sweeps and similar automatic job initiation
  • AAA across distributed storage and compute resources
    • Credential storage and forwarding
Topic revision: r2 - 18 Sep 2013, BenCaradocDavies
 

Current license: All material on this collaboration platform is licensed under a Creative Commons Attribution 3.0 Australia Licence (CC BY 3.0).