Speaker |
Title |
Time |
Abstract |
Material |
Vicente Navarro ESA |
ESA Datalabs, an Open Digital Platform for Innovation and Collaboration in Space Science |
12' |
Nowadays, the sheer amount of data collected from space-borne and ground-based sensors, is changing dramatically past approaches towards data processing and storage. In the era of Big Data, Machine Learning and seamless collaboration, data transfers are being replaced by algorithm transfers and computation close to the data. Through a Software as a Service (SaaS) delivery model, ESA Datalabs leverages technologies such as, Docker, Kubernetes, JupyterLab and Data Science packages to implement a Digital Platform for innovation in Space Science. |
pdf |
Mario Juric University of Washington |
HiPSCat: extending HiPS to for highly scalable large-scale catalog analyses |
12' |
HiPSCat: Motivated by Rubin, we're developing a partitioned dataset format enabling distributed analysis, joins, and cross-matching of catalogs at PB+ scales. The aim is to enable analytics tools allowing users to perform joint, full-catalog, analyses on science platforms. We're looking for feedback and collaboration opportunities. |
pdf |
Robert Nikutta NOIRLab |
NOIRLab's Astro Data Lab science platform |
12' |
In six years of operations Data Lab has attracted over 2500 users. It hosts over 120TB of catalogs and 2+PB of images near compute resources. We are now also serving massive spectroscopic surveys including the imminent DESI EDR. R&D is under way to migrate some services to the cloud, beginning with the Jupyter notebook server. This talk will report on progress and challenges ahead. |
pdf |
Brian Major CADC |
CADC science platform |
12' |
The CANFAR Science Platform allows users to build, publish, and execute docker containers both interactively and programmatically. This has been a success, but challenges lay ahead through participation SKA SRCNet and the Vera C Rubic Observatory, where data and processing are geographically distributed and controlled by different institutes. We require mechanisms to allow this network of heterogeneous science platforms to work together. |
pdf |
Gregory Dubois-Felsmann Rubin observatory |
Rubin observatory and the IVOA |
5' |
Abstract TDB |
pdf |
Dave Morris Edinburgh University |
Gaia Data Mining platform |
5'' |
We are developing a data mining platform using Zeppelin and Spark to analyse data from Gaia. Our deployment provides local low-latency access to the Gaia data partitioned and formatted as Parquet files. We are interested in how we could describe and register and publish the Parquet formatted data and the service associated with it in the IVOA, enabling others to use our data and/or service. |
pdf |
All |
Discussion |
30' |
Panel discussion covering the topics raised in the talks, with emphasis on the IVOA's role going forward. What does the IVOA need to do next ? |
|
Moderator: