Indico Page

Slowly, but surely, we're moving information from this page and into indico: https://indico.cern.ch/conferenceDisplay.py?confId=165687.

This page is a scratch-pad: consider the indico link authoritative.

Logistics

Doodle poll for attendance: http://www.doodle.com/bdxer4hw2znmct64

At NIKHEF, starting at 10AM on Tuesday.

(Meeting rooms arranged by Jeff Templon)

Tuesday:

• all day: Storage Management / Data Management TEG meeting H.331 @ Nikhef. Estimate 50 people

Wednesday:

• all day: Storage Management / Data Management TEG meeting H.331 @ Nikhef. Estimate 50 people

We will organize a dinner Tuesday night; to get more accurate count, we will be asking participants to email the co-chairs Tuesday morning if they'd like to come.

Meeting Goals

Aim for agreement on material that can be used to compose:

1. Assessment of current situation.
2. Needs for (and possible changes in) next 2 to 5 years;

Scheduling

It seems like this motivates a open (but focussed) discussion rather than a series of talks. We will try to divide into eight 90-120 minute sessions over the two days.

Sessions could be under each of the topics we have identified so far.

Joint sessions (perhaps in time order):

1. Discussion of document to produce.
2. DM3: Federations; + DM2 : Data placement ?
3. SM4: SRM and Clouds
4. DM4: WAN Protools / DM5 : FTS
5. SM6 / DM6 : Security and Access Control
6. DM10 : Management of Catalogues
7. SM3: Separation of Disk and Tape
8. SM7: Management and operation of storage at sites (monitoring, performance).
   • Monitoring.
   • Storage I/O / SM5 : LAN Protocols
   • SM2: Requirements for future storage and how it will evolve. What kind of disks can be expected? What directions do we see industry going?
9. Organization and writing assignments for the Storage Management report.

Things for the document(s), but maybe not enough for a session?

1. DM1: Review of Demonstrators from Last Amsterdam Workshop
2. DM7: ROOT, PROOF, POOL

Perhaps apart from SM7, its seems like people from both groups will be interested in both? Particularly, experiments may be hard-pressed to have all their representatives join both.

Decisions:

• Decision point: Could we make all sessions joint and use the alternate room as a break-out for group discussions? Sounds like there is agreement "yes"
• Would like to go in the direction of
• We have formal input from the experiments. This will serve as a mechanism to get input from the sites and middleware; more of a starting point, slightly less formal.
  • The F2F will serve as a way to "anchor" the input from the sites and middleware. For those that haven't been able to contribute yet,this is their chance to start participating.

In terms of Pre-meeting work:

1. Do we need to agree the topics we have listed in a meeting? I am conscious we never really did that?
   • From BB: Would be useful to link back to the master topic list.
2. Each discussion will need a "leader" and an "enforcer": the leader will be responsible for kicking-off the discussion, and the enforcer will be responsible for keeping it on topic.
   • Each discussion will need an introduction - ideally capturing what we have already learnt from experiment answers and twiki responses.
   • Intros could be done by the chair - split among the 4 of us it need only be 2 each. Alternatively we could ask some particular people to digest and uncover a possible view as to:
     • The description of the topic itself (as they see it);
     • Its state of play
     • Where things need to go
     • What might go wrong.
     • and identifying what unanswered questions stand in the way of us.
3. We need to setup a "framework" for each session - needs to stay on topic, free to do what format they want (discussion-based, pen-and-paper, presentation, etc).