dCache Tier I meeting April 27, 2010

[part of a series of meetings]

Present

dCache.org(Patrick,Tigran,Antje,Paul), IN2P3(), Sara(), Triumf(), BNL(), NDGF(Gerd), PIC(Gerard), GridKa(Doris), Fermi(), CERN()

Agenda

(see box on the other side)

Site reports

FZK

Nothing to report.

PIC

Things are OK.

1.9.5-17 everything is OK

2 GIB 4 GiB out.

SRM database recovery procedure done. Using procedure from Dmitry. Seems to be working. System is working after the procedure. Waiting for feedback from end-users about success.

Possible cause for the problems: client does PUT-done issued before GridFTP is completed.

NDGF

Mostly fine at NDGF.

Have an issue: have a file where the tape-hardware is too slow to do pool-2-pool at the same time as flushing.

Tried using the flush controller; flushing phase and a xxx phase.

What is the status of the flush controller in dCache? Is the central flush-controller controlling: just for flushing, not for staging. Not been supported for some time: originally added for Jon, but wasn't adopted. Not used at DESY.

Modulular: can plug in your own modules.

To be discussed during dev. meeting.

Triumf

Two tickets:

1.9.5-17 working fine.

To see

Could you continue to monitor the issue : 4% issue.

Certificate CAs disappearing.

WebDAV

Thinking about doing tests with WebDAV: 1.9.5. Want to upgrade to 1.9.6.

Can we use this with PNFS? Yes.

dcap++ =

No release date yet.

Support tickets for discussion

[Items are added here automagically]

DTNM

Proposed: same time, next week.