NT1/SG-Meeting-2020-10-07

NT1 meeting 2020-10-07, 10:00 - 12:00

Online: link see nt1-sg chat

Invited: Mattias Wadenstein, HPC2N; Oxana Smirnova, HEP, LU; Michaela Barth, PDC-HPC

Agenda

• Presence and approval of agenda
• Staff changes
• KPI update
• Review funding level process
• EGI membership and potential consequences of Sweden leaving - update
• Onboarding new communities
• RCN application input
• Open actionpoints
• AOB

Minutes (approved)

Presence and approval of agenda

Present: Mattias Wadenstein, HPC2N; Oxana Smirnova, HEP, LU; Michaela Barth, PDC-HPC

Decision: Agenda approved.

Staff changes

The new dCache developer started work on October 1st and is currently familiarizing herself with code and the project.

Draft for activity manager position announcement set up by Michaela and commented by Mattias. Split up between Technical lead and activity management might not be clear, could be up to 50-30 in either direction. Also there are some synergies in the roles.

Michaela still has to discuss the current draft with NLCG committee members and XT.

KPI update

Current KPIs as discussed in NT1/SG-Meeting-2018-11-29.

Stakeholder engagement would be a performance indicator on the national funding and partner commitment, and not on us. What if a partner would not get the grant, shall they be red for four years? No, they can adjust the pledges accordingly in the next year, we measure pledge fulfillment. Still we would like to see clear attribution of shortfalls per country.

Proposed new KPIs:

Stakeholder contributions (NT1): Resources pledged by the partners ({ALICE,ATLAS} {Disk,CPU,Tape}) > 80% of experiment requirements (yellow: miss target on 1-2 resource types, red: miss on more than 2). Pledge fulfilment per partner > 95% (yellow: miss for 1-2 partners, red: miss for more than that). Steering and reference groups representation for one year back per partner (yellow: below 85%, red: below 75%). Since the partners set their pledge levels themselves, those promises should be realistic and their fulfillment should be dependable.

Reliable Services (NT1): green: ATLAS & ALICE Availability & Reliability >98% (yellow: one or both are >95% but <98%, red: one or both <95%).

Provided service level (NT1): Resources are delivered swiftly and fully for researcher usage upon contribution.

Cost-efficiency (NT1): Total spending within 80-120% of budget. Accumulated partner commitments >90%. Budget is matching recommended levels.

Provided service level KPI needs more work on how to quantify it.

• AP Michaela: to discuss suggested KPIs within NeIC XT.
• AP all: SG present updated KPI ideas to NLCG committee.

Review funding level process

https://wiki.neic.no/wiki/File:2020-09-24_Board_meeting_minutes_online.pdf

20-33 AP: A process for reviewing the level of funding for NT1 should be presented in a future board meeting.

https://wiki.neic.no/int/XT_meeting_2020-09-15#NT1_risk_profile

Different alternatives:

• Internal evaluation: on XT level: Tomasz, Rob, Johanna, Vilma?
• External evaluation: budget?
• Combination thereof

At the last external evaluation the cost for the external reviewer was 3 PMs.

It is important that we take care of potential conflicts of interest. Conflict of interest could also be negative (someone wants to have the money for other NeIC activities)

Could there be a regular review group to set the funding for the next x years as part of a regular process? It would be good if WLCG could come up with reference numbers to relate to, surveys with staffing has been done in the past on GDB/WLCG operations level.

How could this look like? Group of 3-4 Tier-1 heads to form an expert panel.

Argument: It would be really hard with no background knowledge involved to do an evaluation within a reasonable time frame/time involvement. External review has to come with a budget, no one will be willing to do this voluntarily.

What could an internal review look like: Internal review: NLCG committee coming with recommendations, + interviews of staff + interview with people suitable for an expert panel.

SGs summary: The SG recommends that the NT1’s long-term sustainable role should be the focus of any evaluation. NT1 would like to see a regular process that sets the funding level for several years in one go.

EGI membership

This question has already been discussed in a previous SG meeting.

DK left EGI in 2014, NO in 2017 FI left last year, SE was the only country left in EGI.

Swedish RC investigated (through SNIC) what is delivered for EGI membership fee.

It was clear for the particle physics community that usage value was way below EGI membership fee. Yet, negotiating the membership fee was not an option. Particle physics community alone can not cover the costs. Other communities we know are looking into an EGI affiliate membership (cheaper fee, but still expensive) to use (for instance) EGI CheckIn for AAI (which is not relevant within NT1).

WLCG sites’ usage of EGI services, such as APEL or GGUS, while not being EGI members, will be discussed in the coming WLCG Overview Board (Ian Bearden of NBI is the Nordic representative). EGI security team and incident tracking is very useful, but this seems to get transferred to EOSC.

No Nordic members left means that the EGI structure NGI_NDGF would be dismantled. This leaves orphan sites, like the CMS Tier-2s in Finland and Estonia.

Summary: We leave the discussion to the WLCG Overview Board.

Onboarding new communities

New draft for mini project on onboarding new communities uploaded in the NT1 Google drive.

Discussion: Needs some kind of collaboration agreement, NLCG committee needs to agree for impact on existing service delivery or usage of NLCG resources. Investigate and record what the community needs, check later if they are happy with the level of service provision. Results should be monitored and checked. Software development might be needed to interface own software. Given project cost estimate given should be seen as a minimum. Who should cover which part of the cost? Will a community be willing to pay until they see the value? What is in-kind?

Transferral should be defined in a CA.

EISCAT-3D pilot: NT1 service offering has been drafted containing a list of concrete services including cost estimates. Executive Summary and some future vision sections are still to be written. Mattias will complete the document and send it to EISCAT recipients. Mattias will report to have sent this to the relevant decision makers on Monday within EISCAT-3D meeting. (This corresponds to preparatory work for MileStone4 in the project plan). Once they say yes to (parts of) the offer, we can write a collaboration agreement and start the recruitment process.

AP: Michaela and Mattias to read through updated project plan and provide feedback (until 22nd of October)

AP Mattias: keep following-up with EISCAT-3D.

RCN application input

LoI/LoS could come from:

• Computing Coordinators for ATLAS and ALICE
• NLCG committee chair
• WLCG project leader

AP: Mattias to send all contact details to Johanna and ask her to ask for LoS.