DM Monthly Status Report for June 2019


(Michelle Poland) #1

The DM monthly status report covering March’s activities has been posted to DocuShare, collection-773. For convenience, the High-level Summary is pasted below. Direct link to the full report (pdf): http://ls.st/r95

High-level Summary

Community Interactions, Meetings and Workshops

Face-to-face meetings of the DM System Science Team and the DM Leadership Team and were held at NCSA in Urbana-Champaign, IL during the week of 3 June. Highlights of these wide-ranging meetings included:

A review of the current status of LSST’s scientific measurement algorithms;

Discussion of the relevance of possible “in-kind contributions” to Data Management;

Status updates on middleware development;

A review of the current status of DM’s level 3 milestones.

A very successful Community Broker Workshop was held in Seattle from 19 to 21 June. This meeting brought together members of the LSST Project with developers of potential community alert brokers and key scientific stakeholders to discuss technical and scientific details of the planned broker infrastructure.

A draft paper describing atmospheric characterization based on NASA MERRA-2 data, and how it might be applied to LSST, was submitted to the LSST Publication Board.

Technical Progress

The authentication system was integrated with both Portal and Notebook aspects of the LSST Science Platform.

Middleware development made important progress, with the integration of a registry backend based on the Oracle database system — essential to deploy the “Generation 3” middleware at the LSST Data Facility — and the development of a prototype Batch Production Service workflow system. Conversion of existing code to the new system continued apace, with all processing tasks required for our regular integration tests based on Hyper Suprime-Cam data now available.

DM started work with Quansight, an external contractor, to develop new “drill down” QA tooling. We anticipate this being an efficient and effective way to make rapid progress based on specialist commercial data visualization skills, and expect to present an initial version of the tool at the Project and Community Workshop in August.

The Science Pipelines team completed a “sprint” to better understand the false-positive rate seen in image differences. This has presented a number of promising opportunities for reducing the number of artifacts incorrectly identified as sources. Major progress was also made in improving the single-CCD astrometric fitting system to properly take account of known camera distortion models.

Hardware deployment continues, with computer systems to support initial testing of the Commissioning Camera (ComCam) being shipped from NCSA to Tucson. On the Santiago to Boca Raton Spectrum link, the network hardware facing the Angola Cables system in Boca Raton is connected and ready for activation.

A major fiber cut on the La Serena to Santiago network linked affected service on 18, 21 and 23 June, with a total outage of 41 hours. During this period, the backup links worked correctly: no traffic was lost.

Finally, the System Architecture Team released documentation on the DM software release process (DMTN-106) and software environment handling (DMTN-110).