XSEDE Management and Governance John Towns XSEDE Principal

  • Slides: 14
Download presentation
XSEDE Management and Governance John Towns XSEDE Principal Investigator jtowns@ncsa. illinois. edu

XSEDE Management and Governance John Towns XSEDE Principal Investigator jtowns@ncsa. illinois. edu

Best Known Practices/Lessons Learned • Best known practice • central leadership with delegated authority

Best Known Practices/Lessons Learned • Best known practice • central leadership with delegated authority proven to be very effective • complemented by a strong collaborative environment among partners • Lessons learned • crisp vision/mission/goals help drive best organizational performance • shared vision has encouraged growing mutual respect and aligned self-interest • we want to see each other succeed • use of metrics critical in driving organizational improvement • both drive improvement of service to the community • Much of this documented in the Tera. Grid Final Report and XSEDE 1 Final Report • Tera. Grid Final Report: http: //hdl. handle. net/2142/43874 • XSEDE 1 Final Report: http: //hdl. handle. net/2142/96220 2

Org Chart: XSEDE 2. 0 3

Org Chart: XSEDE 2. 0 3

Organizational Alignment with Strategic Goals Deepen and Extend Use Advance the Ecosystem Sustain the

Organizational Alignment with Strategic Goals Deepen and Extend Use Advance the Ecosystem Sustain the Ecosystem 4 • • Deepen use (for existing communities) Extend use (to new communities) • • Prepare the current and next generation (workforce development) Raise awareness of the value of advanced digital services WBS 2. 1 CEE • • Create and open and evolving infrastructure Enhance the array of technical expertise and support services WBS 2. 3 XCI • • Provide reliable, efficient, and secure infrastructure Provide excellent user support services • • Effective and productive virtual organization Innovative virtual organization WBS 2. 2 ECSS WBS 2. 4 Operations WBS 2. 5 RAS WBS 2. 6 Program Office

Balanced governance model • Strong central management provides rapid response to issues and opportunities

Balanced governance model • Strong central management provides rapid response to issues and opportunities • Delegation and decentralization of decision-making authority • Openness to genuine stakeholder participation • stakeholder engagement, advisory committees • Professional project management practices • formal risk management and change control 5

Elements of Governance • Project PI and co-PIs • ultimately responsible to NSF for

Elements of Governance • Project PI and co-PIs • ultimately responsible to NSF for project progress and performance • Senior Management Team (SMT) • all WBS Level 2 Directors • external representation • chair of the User Advisory Committee • chair of the Service Providers Forum • ex-officio, non-voting members • Senior Project Manager • NSF Program Officer • Advisory bodies • XSEDE Advisory Board • User Advisory Committee • Service Providers Forum 6 External Advisors. SPF) (XAB, UAC, Directors (Hart, Lifka, . Payne, Peterson) Co-PIs (Blood, . Gaither, Sinkovits) PI. (Towns)

XSEDE Advisory Board (XAB) • Recommend strategic directions • Guidance to help XSEDE achieve

XSEDE Advisory Board (XAB) • Recommend strategic directions • Guidance to help XSEDE achieve maximum impact • Assist with requirements gathering and prioritization • Assist with annual planning process • review Program Plan • Lisa Arafune, Coalition for Academic Scientific Computation (CASC) • Ken Bloom, University of Nebraska, Lincoln • Randy Bryant, Carnegie Mellon Univ • Thomas Cheatham, Univ of Utah • Toni Collis, Appentra • Rudolph Eigenmann, University of Delaware • Clifford Jacobs, independent • Albert Lazzarini, Deputy Director, LIGO Laboratory • Karin Remington, NIH/Computationality, LLC (Chair) • Ani Ural, Villanova University • Service Provider Forum Representatives: • • • Ruth Marinshaw, Stanford University (SF Forum Chair) David Hancock, Indiana University Jonathan Anderson, CU Boulder • User Advisory Committee Representative: • 7 Emre Brookes, University of Texas Health Science Center in San Antonio (chair)

User Advisory Committee (UAC) • Represent the “user’s voice” to XSEDE management • Review

User Advisory Committee (UAC) • Represent the “user’s voice” to XSEDE management • Review quarterly and annual reports • Review ESCC proposals • make recommendations • Review annual User Survey Results • make recommendations • UAC Chair is a member of the XAB and a voting member of the Senior Management Team 8 • • • • Rommie Amaro, UCSD Richard Braatz, MIT Emre Brookes, UT San Antonio (chair) Manuela Campanelli, Rochester Institute of Technology Tom Cheatham, Utah Diego Donzis, Texas A&M Mark Jack, Florida A&M Balint Joo, Jefferson Labs Mark Miller, UCSD Benoit Roux, U. of Chicago Deirdre Shoemaker, Georgia Tech Xuguang Wang, Oklahoma University Kraig Winters, SIO/UCSD Frank Wuerthwein, UCSD

Service Provider Forum (SPF) • Forum for Service Providers that deliver services via XSEDE

Service Provider Forum (SPF) • Forum for Service Providers that deliver services via XSEDE • Service Providers input into XSEDE’s management • Venue via which XSEDE management may communicate effectively with SPs • SP Forum Chair is a member of the XAB and a voting member of the SMT • Two additional members of the SPF are members of the XAB 9 • 5 Level 1 Service Providers • Stampede 2, Comet, Wrangler, Jetstream, Bridges • 9 Level 2 Service Providers • NCAR, OSG, Blue Waters, Super. MIC, Rosen Center (Purdue), Stanford Research Computing Center, Beacon, Science Gateways Community Institute • 26 Level 3 Service Providers https: //www. xsede. org/about/governance/spf

Effective Stakeholder Engagement • Collection of stakeholder needs: • surveys, ticket mining, usability studies,

Effective Stakeholder Engagement • Collection of stakeholder needs: • surveys, ticket mining, usability studies, interviews, … • include various reports from NAS, NSF, etc… • Development of Use Cases • capturing community needs • Prioritization of identified needs and derived requirements • User Requirements Evaluation and Prioritization (UREP) Working Group • broad participation across integration, deployment, operations, users, and Service Providers • Assessing plans and deployments • through a variety of stakeholder-focused efforts • e. g. , stakeholder involvement in reviews, usability studies, … • Representation in the management of XSEDE • Service Providers Forum Chair +2, User Advisory Committee Chair 10

Consult with NSF Decision Making in XSEDE NO PI (Towns) Can make decision? Our

Consult with NSF Decision Making in XSEDE NO PI (Towns) Can make decision? Our philosophy PI & co-PIs – strong central management to (Towns, Blood, Gaither, Sinkovits) provide rapid response to issues and opportunities – delegation and decentralization Senior NO Management of decision-making authority Team Can make YES NO Can make decision? YES • Communicate to rest of project as necessary Very few decisions • Execute PCR happen here YES • Communicate to rest of project as necessary • Execute PCR • User Advisory Committee and SP Forum participation in decision-making process • Communicate to rest of project as necessary • Execute PCR • Within scope, schedule, budget of Level 2 WBS • Communicate to rest of project as necessary • Execute PCR decision? NO WBS Level 2 Director Can make decision? NO WBS Level 3 Manager Can make decision? Most decisions happen here 11 YES • Within scope, schedule, budget of Level 3 WBS • Communicate to rest of project as necessary • Execute PCR if appropriate PCR = Project Change Request

Benefits/Challenges of Large Partnership • Benefits • complements the capabilities of the major partners

Benefits/Challenges of Large Partnership • Benefits • complements the capabilities of the major partners • leverages particular strengths of smaller institutions • diversity of institutions brings strength to the partnership • diversity of viewpoint critical to this! • Challenges • collaboration across distributed team is challenging but XSEDE has done well in developing mechanisms to support communications • reporting is complicated and requires significant project staff effort • NSF financial reporting requirements are onerous 12

We Have Reached a Good State • Trust has been built—it is the cornerstone

We Have Reached a Good State • Trust has been built—it is the cornerstone to our successful organization • L 2 Directors and L 3 Managers are a cohesive group • L 2 Directors taking a stronger role in leadership of the project • a good sign of organizational maturity • Learned to compartmentalize roles • frequent community competitions pit team members against one another • Effectively managed significant challenges • transition of leadership: loss of co-PIs, L 2 Directors, L 3 Managers, senior technical staff • All indicators of a resilient, high performing organization 13

Panel Discussion • Are there improvements to the governance model? • Are there better

Panel Discussion • Are there improvements to the governance model? • Are there better mechanisms for engaging stakeholders? 14