- Compact style
- Indico style
- Indico style - inline minutes
- Indico style - numbered
- Indico style - numbered + minutes
- Indico Weeks View
The SPI project has experienced important changes in operations, infrastructure and architecture terms in the last three years. The large amount of packages and platforms that the LCG stack supports, the set of new compilers and architectures expected by experiments and the new communities entering the project, together with the role that new file and storage systems are playing define the frame that drives the project to a new stage. Beyond the operational nature of the Librarian and Integration bi-weekly meeting that we have with the experiments, we need to review the current SPI project, its weak areas, the points not covered for our users, their expectations and the future directions the project should take in the next three years time.
The central idea of this workshop is therefore, to review the status of the project as it is right now, to establish a common strategy that will cope with the requirements of the LHC users and communities and to identify areas of interest for the rest of LHC communities that still are not using the LCG stacks.
Objectives to review
Usage of the current LCG nightlies and releases
Deployment on the experiments build stacks
Usage of the releases within their experimental models and by their user communities
Distribution of the software: Packaging infrastructures
Distribution of LCG releases in Docker containers
Review of the packages currently included in the LCG stacks and identification of weak points in our structure
Validation of the LCG releases
Setup of validation tests within LCGtest
File and storage systems
Usage of CVMFS/EOS for experimental builds/releases at T0
Distribution of CVMFS at the Grid sites (T1, T2...)
Relevance of docker distributions for non CVMFS aware sites
Role of sft.cern.ch CVMFS area: distribution to all Grid sites
Build systems used by experiments
Configuration of the build nodes with a common perspective
Jenkins; common points
Maintenance of old releases: Production against development LCG releases
Role of AFS if any
Future Objectives
Expectations of the experiments for the next 2-3 years
Evolution of the LCG stack in terms of new packages/versions
New compilers, OS and architectures (clang, arm, mac...)
Evolution of the experiments build systems in the next 2-3 years
New communities: BE and beyond
Entry ports to new projects
Communication forums if needed
Is LIM enough?, reorganization of the meeting.
The role of the new packaging systems within the experiments plans in the context of HSF