HSF Weekly Meeting #112, October 12, 2017
Present: Graeme Stewart, Benedikt Hegner, Dario Menasce, Liz Sexton-Kennedy, Ian Bird, Rob Kutschke, Michel Jouvin, Guilherme Amadio, Pere Mato, Mike Sokoloff, Pete Elmer, David Lange, John Harvey, Daniel Elvira, Marko Petric
This has changed from being 2 parallel workshops to a single workshop, with both parallel and plenary sessions for the two communities. Draft announcement prepared. Workshop will run from Monday 26 March at 9am until Thursday 29 March at noon.
A straw man agenda:
- Monday morning: common welcome session with WLCG/HSF
- Monday morning/afternoon: short presentations by working groups
- Tuesday morning: common session with WLCG/HSF
- Tuesday afternoon: available for work group sessions
- Wednesday morning: common session with WLCG/HSF
- Wednesday afternoon: available for work group sessions
- Thursday morning: Plenary summary session for whole workshop
- Workshop will close at noon.
A common WLCG/HSF group will finalise the agenda, but everyone can/should suggest topics we should be working on, especially coming out of our CWP working groups. Note - we are easily able to have several HSF focused sessions running in parallel (there are several rooms available). A proper draft agenda should be prepared by the end of November.
This approach/suggestion was discussed and well received at GDB yesterday.
Performance and Costs WG
WLCG Working Group on System Performance and Cost Model has been set up (see GDB yesterday). Has been going on for a while, convened by Markus Schulz. Idea is to present more in March, so people need to do some work now. It was discussed in the GDB that this group should actually operate at the HSF level.
General agreement that this is a good idea - we should draw in ideas from other communities and having it under HSF broadens the scope. Dario and Rob will look into participating.
Interested people should get in touch with Markus.
Benedikt reported they would like to start a project for upgrading reconstruction software and will remain in contact with the HSF.
General matters and Editorial Board
- Graeme, Michel, John and Pete are hard at work assembling the synthesis paper
- For the moment working in a shadow document as the process of assembling material and doing at least one consistency check is not useful for others.
- They hope very much to have a first public version in place for the meeting next week.
- For each of the WG areas they are extracting:
- Scope and Challenges
- Current Practice
- Research and Development Program
- ~3 Year: What would we like to understand to make decisions for the ATLAS and CMS Computing TDRs in ~2020? Which should explain how we might actually do and afford HL-LHC computing.
- ~5 year: What elements needed for HL-LHC could we deploy already in LHC Run 3 to gain experience and what might these elements look like?
- Discussion: timescales are not completely rigid, more for the road map milestones, though these are useful for the LHCC. 1 year goals are still things we need to do - don’t remove them from the individual papers! They will be merged in the 3 year timescales as needed for the synthesis paper (but that paper should not have sections that go out of date as quickly as a year).
- So it’s very useful if document(s) address those points directly, otherwise this has to be “translated”.
- We should identify papers needing more feedback and publicise this to the community.
- For papers in good shape, they should go to arXiv by 31 October (indeed, put them in now if possible).
- Torre confirmed that ATLAS is happy with the executive summary.
- Added author list, references still TODO.
Software Trigger and Event Reconstruction
- Some cleanup of comments needed - nothing major.
Data Access, Organisation and Management
- Draft finalized. Waiting feedback from Andy to distribute it more widely.
Data and Software Preservation
- Draft in pretty good shape: a lot of work by Mike. Distributed more widely to DP community.
Data Analysis and Interpretation
- Waiting input form K. Cramner about data interpretation
- Question from OLI: should we go without it?
Workflow and Resource Management
- Kaushik is making the first draft now, then passing it around other contributors. Draft available by next week?
Event/Data Processing Frameworks
- 6 page summary almost ready.
- One point about workflow tool and framework interplay to add to the summary document draft.
- The “full” document will be kept as a notebook, but won’t form part of the CWP.
Careers, Staffing and Training
- Waiting for review by others (comments, suggestions and criticisms are more than welcome). Document not very long, just 9 pages, so no executive summary needed. Status of the document should change from “still writing” to “ready for review”.
- Michel read the document and will provide some feedback.
Facilities and Distributed Computing
- No change, people continuing to add comments.
- David added comments, which have not been replied to yet.
- Still being reviewed by MCNet.
- Contacted HepMC authors. Andy Buckley seems fine with LGPL. For v2 Lynn Garren will have to look for an old email by the original author giving a generic OK to future maintainers to do what they would like to do.
- Meeting today of Belle II TF: unfortunately Michel could not attend it. Belle II interested by HSF input when we converged on the work with LHC experiments. More regular meetings in the future.
- Packaging meeting will be Wednesday 18 October at 1500 CERN time (agenda) and updates from anyone working on packaging solutions are welcome (contact Graeme to get Indico access).
- Community calendar contribution guide is a PR. We agreed to create a calendar editors Google Group, in addition to having startup team members able to add events.