ICAT Collaboration Meeting 31st August 2023
- Introductions (if needed)
- ICALEPCS paper
- Extra meetings being arranged
- Java version discussion
- Usual components and site updates
Chris for Allan/Gabriel
- Deadline early October
- Andy plans to create paper on Overleaf
- Suggestions for draft sections/topics
- Summary of sites (facilities) and their status
- ICALEPS link:
- Intention is to send update on the lastest ICAT developments
- Andy will be chairing conference, only author planned to attend the conference
- Meeting in a week to confirm sections
- Organise now? Next Thursday same time?
- Alejandra can't make it but not everyone needs to be there - meet anyway
- Organise now? Next Thursday same time?
- Looking like 14th of September is most popular date?
- Wait until Rolf is back to organise the meeting
- Wait for feedback from HZB e-logbook contributor (Marcus?) - final decision on date next week
- Probably back from the 12th
- Andy: Interest from other communities - is this an internal meeeting? Yes
https://doodle.com/meeting/participate/id/bqZ4v8Gd Allan to orgnaise. Deadline for filling form for next Monday - send reminder then.
- Sirius has a problem with some beamline which want sample tracking (SAXS, powder diffraction and tomography beamlines)
- ESRF have a WIP tracking solution which joins the ispyb and ICAT Samples
- Good time for discussion for if (Sirius) can collaborate on this
- Andy: see no problem with supporting multiple techniques
- Allan: is this feasible? does it make sense to have information for the different domains in the same interface
- Alex: when the samples are sent to the synchotron, we want to know where the sample is and also describe it. Parameters are determined by the acquisition panel - (temperate etc.) also processing parameters (what has been done). This is more than just "sample tracking" in terms of location.
- Allan: Sirius has two beamlines that do want to know T and other physical parameters. Special case at Sirius for safety of the sample - user needs to fill in a form to assure the safety of the sample at Sirius. Want a way of extracting this information. What is the pipeline from user sending the sample to then sending the form - how are other synchotrons doing this?
- Alex: Have safety concerns as well. Have a diagram showing the process.
- Alejandra: we are also interested in how people are doing this and want to know if the description of samples can be standardised and shared.
- Kirsty: there is interest/progress along these lines at DLS as well (but Kirsty is not the person to ask about it)
- Might be worth waiting for Rolf as he raised this.
- Discussion ongoing via the mailing list
- Pros/cons of running 11/17 in production?
- Rolf wants to see "results" - i.e. a conclusion on what to use
- ESRF have no plans - using Java 8 and if it is not necessary do not want to change
- Alan: Payara supports 11, 17, 21, move to 17?
- Allan: has concerns about security of 11
- Still working on supporting crystallography with ICAT
- Goal is to have one beamline simultaneously on ICAT and ispyb
- Support for running workflows and the results ending up in ICAT
- Workflow could mean anything. Could run on the edge (beamline) or the cluster
- Alejandra: what is the script? Anything you want, python script etc.
- Support multiple workflow engines:
- Don't like CWL as is not feature complete - does not support loops, for example
- Marjolaine is working on a new access method to the data (have IDS and MongoDB for small files) but want larger file support - "/online"
- Will be archived but also accessible 10 -> 100 MB in size
- "online" directory treated differently than "gallery" for example
- No real updates, still down from the cyber attack, no realistic timeline for when things will be running again - open time frame
- Rolf has announced snapshot releases for several minor components.
- ANdy: security of ICAT being considered?
- ICAT has nothing to do with the cyber attack. To prevent fututre attacks the network infrastructure of the institute external facing tools are being discussed. ICAT will not be online until this is resolved.
ALBA: No-one present
- Almost finished K8s infrastructure but no update to give
- In November want to have ICAT running in production - good time as a nation attempting data standardisation - two techniques
- Andy: installing ICAT+? Docker compose is the thing working for them, working through the icatproject documentation
- Sesame have ICAT running
ISIS: No-one present
DLS: Kirsty and Chris
- Yesterday moved ingest to K8s - "baby steps"
- Hopeful improve stability and make more improvements in the future
- Have been running ICAT6 (Payara6 etc.) for a month
- Had minor performance issues with DB, created new indices to improve READs
- Also upgrading servers for ingesting to tape to improve throughput 60TB a day currently
- Want to duplicate ingest path to double the throughput
Snapshot releases mentioned in email from Rolf
authn.oidc & icat.oaipmh
Gabriel wants to join mailing list - but this requires linking the business and personal accounts
- Viktor knows a way of subscribing without using GMail
Some emails to the mailing list get flagged as spam/moderation
- Requires AGB to manually approve emails which get caught (unclear what triggers this)
Next meeting on the 28th of September. Alejandra unable to chair.