NT1/SG-Meeting-2017-08-17

From neicext
Jump to navigation Jump to search

SG meeting 2017-08-17, 09:00 CEST, Oslo headquarter

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

Agenda

  • PM status report including WLCG availability charts for July
  • NeIC announcements (Outcome Research Council of Norway application: might need additional information on milestones and KPIs; HR Policy; NeIC Culture Book)
  • New NORDUNet contract for networking 2018+2019.
    • Decision: Presence in the second network POP at CERN
    • Decision: Go ahead to make a new contract using this cost
  • 24/7 progress report and next steps.
  • Continued work on DP checklist document.
  • Board reporting preview.
    • Cover document
    • project summary report
    • new MoU
  • Regular website info check
  • Self assessment
  • AOB

Minutes

PM status report including WLCG availability charts for July

Progress on new networking contract, updated cost breakdown.

Reliability goal looking very good.

Problem to be solved: Where and how to put tape storage in Norway and Finland (Money and location and maintenance)

Vincent Garonne as new staff onboard.

AP on maswan: Complement the pledges in the graphs with actual user requirements (“target” number).

User requirement target is split up by country based on the authors’ contribution.

ALICE sharply increased tape storage requirement last year.

We provide more than what is pledged for for ATLAS CPU, but not in terms of tape. There we are systematically below and no current concrete plans to get back up again (we stand at 47% of target pledged for ALICE tape)

What is needed: Clear statements on future (tape storage providing) expectations and then Resource proficient exchange document to be created and agreed upon by the NeIPs.

Sidenote: Two archive sites chosen in Sweden, Finland, and Norway will also need tape for EISCAT_3D.

Generally data preservation policies, will aim for a cheap long-term archiving of data (which is best provided via tape).

NeIC announcements (Outcome Research Council of Norway application: might need additional information on milestones and KPIs; HR Policy; NeIC Culture Book)

Response from RCN application was discussed.

https://neic.no/news/2017/07/01/neic-reveals-its-new-human-resource-policy/

https://wiki.neic.no/wiki/NeIC_culture


New NORDUNet contract for networking 2018+2019.

Summary: SG discussed the cost breakdown document that will be attached to the contract. NORDUnet is supposed to charge the actual cost.

Currently no redundancy at CERN site, but CERN is now building a second network routing solution, this would be an opportunity to reduce an obvious single point of failure bottleneck. (Number given is just an estimate, this is not a one-time cost).

Non OPN network cost has gone up, with the increased amount of traffic (even if cheaper per MB). Traffic increase will go up with the amount of data and amount of sites.

“NORDUnet representation” is an unfortunate name if it includes labour costs (it seems it does, at ~ 0.25% FTE)

Decision: Added redundancy at CERN site seems to be a good idea worth following, as long as the total yearly cost stays under 30 000 EUR. “NORDUnet representation” item needs to be clarified in more detail.


Decision: Presence in the second network POP at CERN

Decision: Go ahead with presence in the second network POP given that the cost does not increase too much from estimated 25000 EUR/year

Decision: Go ahead to make a new contract using this cost

Decision: The SG sees no more room for improvement and acknowledges to proceed with the contract binding the current estimated total cost (363 550EUR per year including estimates). Aiming at deadline: End of September to leave ample time for all parties to check and sign.

24/7 progress report and next steps

Formal agreement everywhere besides for one person. Time to start on contract templates.

Basis: external experts service contracts but on-call.

We have to have a calendar that can be used for billing. Auditable, including monitoring.

Helpful: Additional log of nagios configuration at midnight.

Agreement has already be made that the night belongs to the day before: sufficient to just add a date.

Check with NordForsk on any specific requirements.

https://wiki.neic.no/int/NT1_24/7_service

AP on maswan to sit together with Kine and work on the contract template (deadline: end of September)

Board reporting preview.

Mattias would like to attend in person.

Cover document

NT1 update document for the board.

project summary report

Regular project summary report document for the board.


new MoU

AP on Oxana: put old versions of the MoU in the shared NT1 SG folder


Regular website info check

  • Fix management weekly meetings link
  • Change” A number of additional services are necessary for a Grid center operations.”->”A number of additional services are necessary for WLCG center operations.
  • Computing missing
  • Accounting link no longer working
  • AP on maswan to fix those.

Continued work on DP checklist document

Focus now first on: Overall Risk analysis Activity Plan missing MoU


Plan for risk analysis: start after finishing NORDUnet contract, get all types of risks, should include many different levels (PM; PO, SG, NLCG committee as RG and staff and funding agencies) Could be bound to CERN MoU, the activity plan then only lists the risks and how ot best address them.

Decision: start with work on risk analysis after finishing the NORDUnet contract.

Self assessment

Not enough time, otherwise doing good.

AOB

Next meeting: again morning of next NLCG meeting