ICAT Steering Group Meeting #5
21st November 2017, 3pm Rutherford Appleton Lab, STFC
- Andy Gotz (AG), ESRF (Chair)
- Suart Pullinger (SP), STFC, Scientific Computing
- Tom Griffin (TG), STFC, ISIS
- Shelly Ren (SR), SNS
- Rolf Krahl (RK), HZB
- Brian Matthews (BM), STFC, Scientific Computing
- Alun Ashton (AA), Diamond
- Gordon Brown (GB), STFC, Scientific Computing
- Catherine Jones (CJ), STFC, Scientific Computing (Secretary)
- Milan Prica (MP), ELETTRA
Welcome & Agree agenda
AG welcomed those present. The proposed agenda was agreed.
Minutes of the previous meeting
These were agreed. They are available from https://icatproject.org/collaboration/communication/steering-group/meeting-04/
2.1 Review of actions
|2.4||Check with CLF/LSF that they are correctly represented||BM||Completed||They were happy for representation through STFC Facilities Programme Manager (GB)|
|4.1||Let the steering group know if he is going to remain a member.||BM||Completed||BM stays as a non-voting member to represent corporate memory|
|4.2||Update the TOR to make it explicit that the steering group secretary does not have a vote||CMJ||Ongoing||CMJ taken this action and will amend ToRs|
|4.3||Provide a summary page for the icatproject website preferably containing a link to TopCAT||AG, SDGR, SR, MT||Superseded and thus closed||Work on new website/documentation means this is no longer relevant|
|4.4||Provide a link to the TopCAT videos from icatproject||SP||Ongoing||Action transferred to SP and will be done as part of new website|
|4.5||Provide feedback to SF on how to improve the website||AG, SR||Completed||Feedback being incorporated in new website work|
|4.6||Provide some examples of metadata that can be stored in ICAT to make available on the website||AG||Ongoing||Will be done as part of new website|
- SC Membership review
No new members planned.
- GitHub Repositories review
The move to GitHub was considered a success and was working well. There was a discussion about how to distinguish projects in contrib and corel. It was agreed that projects in core should be used by more than one site and are actively maintained. Moves from contrib to core shall be recommended by the developer(s) and approved by the Steering Committee.
Active maintenance includes working on high priority bugs and formal developer workflows.
ACTION 5.1: Produce contrib rules/guidelines (SP)
ACTION 5.2 Review the settings for projects currently in Contrib/Official and propose changes (SP & CJ)
- Component Review
It was agreed that it is in the interests of the community to have an active developers collaboration. For components which are only used & developed at one site, then they make decisions based on listening to the community (collaboration meetings). For components that have more than one contributor, then there needs to be a more formal process with the SC responsible for any resolution of disagreements.
ACTION 5.3 Document the process for collaboration (SP & CJ)
|Authentication||All sites need to implement this for their own needs.
There are example/reference ones in the ICAT project. These are considered to be part of the core ICAT project
|STFC – SP|
|STFC is currently the sole site developing this||STFC – SP|
|IDS client, plug-in & server||Noted concerns from Rolf’s boss regarding liability but agreed covered by Apache licence||HZB – RK
Secondary STFC – SP
|IJP||Noted that this should probably move to contrib following review||n/a|
|TopCat||ESRF keen to collaborate here||STFC – BR
Secondary ESRF – Alex
There was then a discussion about eLog notebooks and the links to ICAT.
ACTION 5.4 Report back to SC on investigations into possible tools & approaches (AG)
- Website review
The SC noted the plans SP had put forward for refreshing the website. Noted that using Markdown means it would be possible to port the revised website to another provider.
Agreed that other sites would have the opportunity to comment once the site is ready & content moved
ACTION 5.5 STFC to ask for feedback once website is ready for launch (SP)
- Documentation review
The SC noted the good start with the Documentathon running simultaneously with the SC meeting. It was suggested that the monthly collaboration meeting would be a good place to review progress.
It was acknowledged that this is a significant piece of work.
It was suggested that there should be more time in the next face to face meeting for a documentation sprint
ACTION 5.6 Consider further documentation sprints on the agenda of the next meeting (RK & SP)
- Review of ICAT resources
AG started the discussion on this topic with a reminder that to survive we need to grow.
ACTION 5.7 Contact the CSNS to re-engage with them (GB)
ESRF are able to contribute 1.5 FTE to the development.
DLS 1.2 (more operational than development)
ISIS 1 FTE (split 50-50 operation/development
HRZ 1 FTE
STFC 3-4 FTE (although currently there are vacancies following staff movements)
SR noted that there is a steep learning curve for using ICAT and suggested using Docker for deployment. The issue with the metadata being collected changing causes issues with using ICAT.
There was a discussion about EC requirements to hold data for at least 10 years and to comply with the FAIR principles. AG noted that EOSC is mainly about data.
Points discussed in the meeting included:
- the observation that the ICAT approach to parameters is not flexible enough
- The schema might benefit from a review, especially around datafiles in the RDBMS and consideration of scaling issues and new technology
- Further investigation of ONCat (see section x) and SCICat
- Need to signal high priority of blocking issues that need to be addressed
It was agreed that SP would consolidate the suggestions made by Steve Fisher and the general tickets into a new roadmap.
ACTION 5.8 Share access to ONCat with the collaboration (SR)
ACTION 5.9 Share high priority or blocking issues (via tickets) (All sites)
ACTION 5.10 Establish a new roadmap (SP)
- New projects
SR updated the SC on a new project ORNL are working on called ONCat. They have been working on data management/reduction with the HFIR (https://neutrons.ornl.gov/hfir ). This has 12 instruments and 30 years of data. There is a challenge to use ICAT for the metadata as the approach to parameters is too flat a structure and there is not enough consistency in parameters between instruments. The metadata parameters change between run cycles as well as instruments. It is difficult to get metadata out of ICAT. They are looking at MongoDB & elasticsearch and gained flexibility and performance using this approach & technology and they have developed ONCAT (Oakridge Neutron Catalogue). This has an API and there are several different clients. They have ingested all experimental data. This enables google-like searches and thumbnail displays.
AG asked if it was going to replace ICAT. SR responded that they are both running and ONCat is not in production.
SR reported that there was also US based work on materials data facility (U Chicago) and Argonne National Lab ran a hackathon on commonality in data management.
PSI are using similar technology in SCICat for the ESS and Max IV (scicatproject.org)
DLS have additional resource through the INSTRUCT project for prototyping, for similar requirements as the ICR discussion in the main meeting. They have also submitted to SRI and Mark Basham will be attending the conference/meeting in Japan.
RK noted that the next RDA meeting is in Berlin in March.
- Next meeting
The next steering group meeting will be in Berlin, colocated with the RDA meeting and the workshop on Scientific Data Management for Photon and Neutron Facilities. The workshop will be on 19 and 20 March at HZB, see  for details. RK has scheduled the SC meeting for Tuesday afternoon after the workshop. (The workshop is planned to end on 20 March 15:00, a room for the SC from 14:00 – 17:00.)
Noted that the next NoBugs meeting is in the autumn in the US
There is an expectation that sites will send deputies if main representative not able to attend. AA apologised in advance for not being able to attend the next meeting
ACTION 5.11 Ensure there is remote participation available for the Berlin SC meeting (RK)
|5.1||Produce contrib rules/guidelines||SP|
|5.2||Review the settings for projects currently in Contrib/Official and propose changes||SP & CJ|
|5.3||Document the process for collaboration||SP & CJ|
|5.4||Report back to SC on investigations into possible tools & approaches for eLog books||AG|
|5.5||STFC to ask for feedback once website is ready for launch||SP|
|5.6||Consider further documentation sprints on the agenda of the next meeting||RK & SP|
|5.7||Contact the CSNS to re-engage with them||GB|
|5.8||Share access to ONCat with the collaboration||SR|
|5.9||Share high priority or blocking issues (via tickets)||All sites|
|5.10||Establish a new roadmap||SP|
|5.11||Ensure there is remote participation available for the Berlin SC meeting||RK|