Sylvie (SR), Stuart (SP), Alex (AdM), Maxime (Max), Rolf (RK)
George Christian, Catherine Jones, Brian Ritchie, Chris Prosser
ICAT Server (SP & RK):
- SP working on search replacement component
- first job to document API nearly done
- next job: write tests
AdM: Why index on Datafiles at all?
SR: Users need to be able to find files by name.
- RK: ICAT Schema changes nearly done. Only need to add tests.
- RK: ICAT test suite does not run – hard-coded to Steve Fisher’s computer. See #205 and #206.
Released 126.96.36.199 in August: bugfix release for DLS. Adds configuration to show/hide the Search tab. As indexing is disabled in DLS, Search doesn’t work, so DLS asked for the Search tab to be removed. In this release, Search is removed by default; this meant that it could be installed (and Search removed) without having to change topcat.json.
Whether a similar change is required to the topcat master (2.4.x) depends on whether or not a solution is found to the DLS indexing problem before DLS is ready to upgrade topcat to 2.4.x.
Nothing (interesting) to report. Octopus work is being re-planned; may be more interested in DAaaS than in the IJP. Current Octopus development VMs are due to be retired in October, I am working on creating replacements in the new Cloud service – assuming they will still be wanted!
The Scarf team plan to take over the HTCondor-Slurm work; I’m more than happy, provided that the IJP can use whatever they do!
IDS (RK) and Python ICAT (RK) – Nothing to report.
- New staff starting
- October: Service managers
- November: Developer
- No one from STFC is attending NoBugs 2018
- Consultants will start to look at the DLS infrastructure (including their use of ICAT) from October.
- Archive has now reached 15 Petabytes!
- A new directory structure – by village – has been implemented.
- (SP: this had no real impact on ICAT)
- Now using ActiveMQ in the ingest workflow.
- The replacement for TopCat (eDataPortal – built using React) is waiting to be put into production.
- It includes an eLogBook tab which receives notifications from beamlines. Built using MongoDB.
- Action: AdM to demo it at the next meeting.
HZB (RK) – Nothing to report
INFRAEOSC-5 ExPanDS (RK)
Interesting proposal. Work package 3 is on metadata catalogues.
Action: SP to discuss STFC’s involvement with Brian Matthews.
AdM: What do other sites do when an investigation wants to have 2 datasets with the same name, eg. ‘Test’?
SR: DLS users cannot choose a dataset name – but they can create a folder which becomes the dataset name.
RK: HZB has 2 instruments connected to ICAT. For each scan/run, a new scan/run number is created which becomes the dataset name.
(Note from SP: the conclusion was that other sites did not have this problem as either a) users could not choose a dataset name (HZB) or b) users were forced to create a unique dataset name (DLS)).
AdM: Could we store the file count and size in the Dataset and Investigation?
AdM: At ESRF, all files in a Dataset are ingested at the same time and not updated so the calculation is easy.
SR: suggested this to Steve Fisher but he was against the idea.
RK: we may have to think about who/what has write access to investigation parameters. At HZB, the ingester does not have this permission. Retrieving the file count and size of a Dataset are implemented in the IDS.
|113.2||Organise a meeting for the discussion of a Topcat replacement||Catherine Jones|
|119.1||Review and update instructions for Apache proxy configuration.||SP and RK|
|120.1||Alex to demonstrate ESRF’s replacement for TopCat.||Alex de Maria|
|120.2||Stuart to speak to Brian Matthews about ExPanDS||Stuart Pullinger|
Next meeting scheduled for Thursday 25th October 3pm UK time.