Uploaded image for project: 'SAFe Program'
  1. SAFe Program
  2. SP-4323

Implementation plan for scientific software distribution for SRCNet v0.1

Change Owns to Parent OfsSet start and due date...
    XporterXMLWordPrintable

Details

    • SRCnet
    • Hide

      The goal of this feature is to create a plan for implementing a distribution pipeline for scientific software (such as LOFAR software) for the SRCNet v0.1.

      The benefits of this pipeline would be:

      • A collection of software that is consistently available to end-users on all systems 
      • Less effort for local sites to make scientific software available on their systems 
      • Quality control of scientific software 
      • Publication of software metadata to support software discovery and research reproducibility
      Show
      The goal of this feature is to create a plan for implementing a distribution pipeline for scientific software (such as LOFAR software) for the SRCNet v0.1. The benefits of this pipeline would be: A collection of software that is consistently available to end-users on all systems  Less effort for local sites to make scientific software available on their systems  Quality control of scientific software  Publication of software metadata to support software discovery and research reproducibility
    • Hide

      AC1: Evaluate tooling for generating software metadata, present findings for (at least) codemeta and Escape OSSR to the ART
      AC2: A decision is made (by Jesus or achitecture forum, in consultation with the software quality architect) what scientific software distribution is required for v0.1<- this may include the example workflows repo
      AC3: Create an implementation plan for scientific software distribution for SRCNet v0.1 and present this plan to the ART.

      Show
      AC1: Evaluate tooling for generating software metadata, present findings for (at least) codemeta and Escape OSSR to the ART AC2: A decision is made (by Jesus or achitecture forum, in consultation with the software quality architect) what scientific software distribution is required for v0.1<- this may include the example workflows repo AC3:  Create an implementation plan for scientific software distribution for SRCNet v0.1 and present this plan to the ART.
    • 1.5
    • 1.5
    • 0
    • Team_TANGERINE
    • Sprint 3
    • Show
      AC 1: https://confluence.skatelescope.org/display/SRCSC/%5BTAN-261%5D+Notes+on+software+metadata AC 2:   https://confluence.skatelescope.org/display/SRCSC/%5BTAN-260%5D+List+of+software+for+release+and+distribution+in+SRCNet+v0.1 AC 3: https://confluence.skatelescope.org/display/SRCSC/%5BTAN-262%5D+Technical+implementation+for+the+release+and+distribution+of+containers
    • 24.1
    • Stories Completed, Accepted by FO
    • PI24 - UNCOVERED

    • SRC23-PB SRCNet0.1 science-platform-services software-sharing

    Description

      In PI-22 (feature SP-4193), we have made recommendations for how scientific software (such as LOFAR pipielines) should be distributed throughout the SRCNet in a process we have, for the time being, termed "CI/CD/CM" (Continuous Integration/Continuous Deployment/Continuous Management). See [TAN-220] Recommendations for CI/CD/CM for Software Distribution on Confluence.

      The preliminary conclusion of the recommendations is that containers should be the format in which software is distributed, and container registries combined with CVMFS should be used as the distribution technology.
      For software compilation, we recommend creating an EESSI-like environment in which the containers are built on a variety of hardware architectures.

      A diagram of the recommended pipeline (still work in progress):

      In this feature, we want to create an implementation plan for SRCNet v0.1 in which these recommendations are worked out to a technical degree that they can be implemented in PI-24 and PI–25. 

      In addition, the plan should specify how to deal with the publication of software metadata, which so far has not been addressed conclusively. Concretely, we would like to test whether codemeta as used in Escape OSSR would be sufficient for astronomy software.

      We will also follow up with EESSI/UGent to get a status update on the software discovery mechanism  they are developing.

      Attachments

        Issue Links

          Structure

            Activity

              People

                Jesus.Salgado Salgado, Jesus
                T.Kok Kok, Tim
                Votes:
                0 Vote for this issue
                Watchers:
                2 Start watching this issue

                Feature Progress

                  Story Point Burn-up: (100.00%)

                  Feature Estimate: 1.5

                  IssuesStory Points
                  To Do00.0
                  In Progress   00.0
                  Complete711.0
                  Total711.0

                  Dates

                    Created:
                    Updated:
                    Resolved:

                    Structure Helper Panel