Fermilab Computing Division

CS Document 6568-v1

The LSST Science Platform: Why Kubernetes and JupyterLab are our new best friends (and maybe yours too)

Document #:
CS-doc-6568-v1
Document type:
Presentation
Submitted by:
Tim Skirvin
Updated by:
Tim Skirvin
Document Created:
29 Oct 2018, 10:00
Contents Revised:
29 Oct 2018, 10:00
Metadata Revised:
29 Oct 2018, 10:00
Viewable by:
  • Public document
Modifiable by:

Quick Links:
Latest Version

Abstract:
The Large Synoptic Survey Telescope is a massive optical survey that will produce half an Exabyte of image data. Even if we only cared about the resulting catalogues, the 7 trillion row database naturally pushes us to a “bring your code to the data” regime. But astronomical data exploration is an iterative process that is ill suited to traditional batch computing models for most of its lifecycle. Here I will discuss how the architecture and implementation of the Notebook Aspect of the LSST Science Platform addresses the challenges of providing a typically interactive experience on a datacenter-side service. This talk will be of interest to those building and deploying science-facing data services, as well as those who are interested in a preview of the LSST user experience.
Files in Document:
Authors:
DocDB Home ]  [ Search ] [ Authors ] [ Events ] [ Topics ]

DocDB Version 8.8.9, contact Document Database Administrators