1 of 20

Institutional Support for Experimental Systems Research

Carlos Maltzahn (UC Santa Cruz)

Computational I/O Stack Workshop, 8/17/2023

2 of 20

Skyhook Data Management – A vision

  • Originally: embedding analytics engine into Ceph

3 of 20

Design Patterns for Offloading: Separate Paths

Key challenges:

  • Map and align data path abstractions to data processing abstractions
  • Near-data serialization of query results

4 of 20

More Design Patterns

  • Separate paths for data and analytics
  • Record batch-aware protocols
  • Opportunistic partial evaluations
  • Hardware accelerator surfacing
  • Offloaded data shuffling
  • Asynchronous data processing
  • …

5 of 20

Skyhook Data Management – A vision

  • Originally: embedding analytics engine into Ceph
    • Incubator project at Center for Research in Open Source Software
  • A community platform for the Computational I/O Stack
    • Funders, institutions, and individuals working collaboratively to develop software and infrastructure to enable industry-wide adoption of a computational I/O stack
    • Common mission
    • Common website, channels for news and updates

6 of 20

Mission

Cultivate an ecosystem in which the open source software for the computational I/O stack can be developed, distributed, and sustained. This open source software must reduce barriers of adoption and meet the current and future challenges of the computational I/O stack, and the solutions should leverage the existing expertise outside storage and network I/O communities.

​

Adapted from pangeo.io/about.html

7 of 20

Goals

  • Foster collaboration around the open source data science, storage, and networking systems ecosystem
  • Support the development with system- and domain-specific computational I/O stack packages
  • Reduce barriers of adoption of open source software for the computational I/O stack

Adapted from pangeo.io/about.html

8 of 20

The Problem

Missing: �software engineering and community management staff

9 of 20

The Problem

  • Working on NSF POSE Phase 2 proposal (2 years, $1.5m)
  • Raising funding through CROSS v2 sponsorships

10 of 20

  • Bridges gap between student research & open source projects
  • Amplifies translation and impact of university research
  • Initiated by a $2M Sage Weil gift
  • Sustained by membership fees

​

​

10

Education

Research

Incubator

UC Santa Cruz

Faculty & Students

Industry Advisory Board

Talent, Projects, Technologies

Teach students on how to productively engage in open source communities

Fund high-impact research with plausible path to successful open source projects

Incubate developer communities around research prototypes

cross.ucsc.edu est. 2015

V1

11 of 20

We thank Sage Weil for his generous $2M gift to support CROSS, and we thank out current and past industry sponsors for their $2.4M in membership fees since 2015.

V1

12 of 20

V1

Industry Advisory Board*

* Accumulative membership 2015 - 2022

13 of 20

James Davis (Professor of Computer Science and Engineering), �Founding Director of UCSC Center for Entrepreneurship, sits on advisory councils for startups and nonprofits.

�Doug Cutting (Founder, Apache Lucene & Hadoop),�Founder of numerous groundbreaking open-source projects, including Lucene, Nutch, Avro, Hadoop.

�Nithya Ruff (Head of Open Source Program Office, Amazon & Chair, Linux Foundation Board of Directors)

​

13

Advisory Committee

Karen Sandler (Executive Director, Software Freedom Conservancy), �Former executive director of GNOME Foundation, former general counsel of Software Freedom Law Center

Nissa Strottman (VP, Products and Operations, VISA), �Years of experience in legal and M&A in the open-source software space.

Sage Weil (Chief Architect of Ceph, Red Hat), �Creator of WebRing, founder of DreamHost, founder and CTO of Inktank, recipient of O’Reilly Open Source Award

V1

14 of 20

Twice a year:

    • Call for proposals
    • IAB meeting

Once a year:

    • 2-day Research Symposium

Weekly research meetings

V1

15 of 20

Successes

V1

Raised $2.4M in industry funding

Stands out among OS efforts world-wide

Encouraged open source practices

Supported 18 research fellows and 5 incubator fellows

50+ publications, 6 awards

Award-winning open source course

Served as blueprint for OSPO UC Santa Cruz

16 of 20

OSPO UC Santa Cruz

  • 🚀 Mission: create partnerships to
    • Promote open source literacy
    • Enable students & scientists
    • Connect scientists with stakeholders

​

  • 📺 Programs:
    • Open Source Research Experience
    • Open Source Incubator Fellowship
    • Open Source Education

​

  • 📖 Lessons learned at:�

​

  • 💰 Funding:
    • Alfred P. Sloan Foundation
    • National Science Foundation

​

  • 🏠 ospo.ucsc.edu

Est. 2015

17 of 20

Introducing

V2

V2

18 of 20

  • Bridges gap between student research & open source projects
  • Amplifies translation and impact of university research
  • Initiated by a $2M Sage Weil gift
  • Sustained by gifts

​

​

18

Education

Research

Incubator

UC Santa Cruz

Faculty, Students,

& Staff

Stake-holders

Talent, Projects, Technologies

Teach students on how to productively engage in open source communities

Fund high-impact research with plausible path to successful open source projects

Incubate developer communities around research prototypes

cross.ucsc.edu est. 2015

V2

19 of 20

V2

CROSS v2 selects incubator projects and employ software engineers and community architects *staff*

Incubator fellowships for postdocs at OSPOs

Candidate incubator projects can be proposed by students, faculty, research labs, research centers

CROSS v2 funding model based on gifts instead of membership fees

CROSS v2 governance based on �Open Source Best Practices

20 of 20

Sponsorships

  • Support projects “to be selected” jointly donor and CROSS advisory board
  • Gift Letter acknowledges open source licensing
  • Suggested funding levels for meaningful ROI
  • Open meetings, no NDAs

​