Minutes of Storage Phone conference
			     26 Apr 2006

Present:
	Lancaster: Brian
	Durham: Mark
	RAL Tier 1: Derek
	Edinburgh: Greig
	RAL Storage: Owen, Jiri, Jens (chair+mins)

Apologies:
	Glasgow: Graeme



0. Review of actions (see below)

1. Deliverable Status

   Jens reported that he hadn't had enough time to make enough
   progress with the Level 1 but had noted some subjects and left the
   report with Jiri.  The Level 1 is a major project deliverable (and
   is usually software integrating with an EGEE or LCG software
   release but this time it's a report) whereas Level 2s are normal
   progress deliverables (usually reports - this time the AHM papers
   are associated L2s).

2. dCache bugs and issues-for-DESY list

   We want SRM 2.1 for dCache!

   Do Tier 2s need xroot?  Alice, and perhaps Atlas, want xroot at the
   Tier 1s.  Furthermore xroot should be in the next release of dCache
   (1.6.7), and then we can test it.  ACTIONs.

   Performance during simultaneous reading and writing...?  Tier 1
   seems to have experienced problems in this area, with multiple
   writes preventing reads (of files going to tape), which means that
   the cache fills up.

   We agreed that it is worth testing this first at the Tier 2s.  The
   scenario is that a Tier 2 is getting data from a remote site (Tier
   1 or another Tier 2) and at the same time sending data to another
   Tier 2.  Or each other.  Edinburgh and Glasgow should be ideal for
   this test (hint).

3. PPS status, and plans for testing

   Edinburgh's dCache is not yet in PPS.  But both Birmingham and
   Glasgow have DPMs in PPS:
   https://wiki.gridpp.ac.uk/wiki/Category:UKI_Testzone

   Jens would like to run independent storage regression tests, and
   would follow up with Dave Kant who built the testing framework.
   ACTION.

4. Storage metrics review

   Metrics are important to show the world in general, and the PMB in
   particular, that we are making progress.

   We have the storage available (nearly 220TB) as a metric, but we
   also used to have number of sites with SEs deployed.  The latter is
   of course useless because now everyone has one.

   We brainstormed a bit: number or volume of site transfers, uptime,
   numbers of gets and puts, status of site upgrades.  They have to be
   something we can measure, but also something that is useful, and
   doesn't depend on random experiment usage fluctuations.

   Will need to follow up when we've thunk a bit more.  ACTION.

5. AOB

   Owen mentioned the FTS bug where FTS appears to transfer files in
   the order it was told to transfer them, as opposed to when they
   become ready for transfer.  Worth raising as a bug.


41	10/08/2005	Agree licence with DESY	Jens	Open		

No news but Jens is doing related boringstuff at the moment.

53	12/10/2005	Find reasoanable % for SE uptime for SC4	Jeremy	Open		Reassigned. Follow up with GDB et al

No news, but also not concerned at the moment.

54	02/11/2005	Report on performance/scalability with pools on WNs	Paul	Open		Progress. Will put in wiki. NB: dcache pools

No news.

86	08/02/2006	Extend monitoring to do sites per VO and VOs per site	Greig	Open		

No progress since last meeting, but generally progress.

91	22/02/2006	Add an SRM 2.1 tests page to the wiki	Jiri	Open		See also #75

Being done as we speak.  Write.  Read.  Speak.
https://wiki.gridpp.ac.uk/wiki/DPM_SRMv2_Status

102	19/04/2006	Find out du load on Tier 1 dCache	Derek	Open		

Done - Derek circulated some information to the mailing list.
http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0604&L=gridpp-storage&T=0&P=6236

103	19/04/2006	Send dCache shopping list items to Owen/Jens, then prioritise	ALL	Open		

Ongoing, but Owen's trip has now been postponed.